Device and method for pressure control of electric injection molding machine

Abstract

{Problem} The exact method with small time-lag of detecting injection pressure for controlling pressure in an electric-motor driven injection molding machine without using a pressure detector has been asked for because the pressure detector is very expensive, necessitates troublesome works for mounting, an electric protection against noise and the works for zero-point and span adjustings and causes a complicate mechanical structure.

Description

TECHNICAL FIELD

This invention is concerning an apparatus and a method for controlling pressure in an electric-motor driven injection molding machine.

BACKGROUND

AC servomotors are becoming used for middle-sized injection molding machines heretofore driven by hydraulic actuators (clamping force >3.5 MN) that have high precision, quick response and higher power which are obtained by performance improvements of permanent magnets and cost reductions.

An injection molding machine consists of a plasticizier in which resin pellets are melted by friction heat generated by plasticizing screw revolution and stored at the end of a barrel, an injector in which an amount of melted polymer is injected into a metal mold at a given velocity and a given dwell pressure is applied, and a clamper in which the metal mold is clamped and opened, all using AC servomotors drive system. FIG. 2 shows a view of an injection molding mechanism using an AC servomotor.

On an injection machine base which is fixed on the ground, a movable base is located which moves on a linear slider and both the bases are not shown in FIG. 2. All parts except a metal mold 1 shown in FIG. 2 are mounted on the movable base. By sliding the movable base, the top of a barrel 2 is clamped on the metal mold 1 and vice versa the top of the barrel 2 is separated from the metal mold 1. FIG. 2 shows a mode in which the top of the barrel 2 is clamped on the metal mold 1 before melted polymer being injected into the metal mold 1.

On the movable base, a barrel 2, a servomotor 3, a reduction gear 4, a ball screw 5 and a bearing 6 are fixed. A nut 7 of the ball screw 5, a moving part 8, a screw 9 and a pressure detector 10 such as a load cell consist of an integral structure. The moving part 8 is mounted on a linear slider 11 so that the integral structure is moved back and forth by the movement of the nut 7 of the ball screw 5.

Rotation of the servomotor 3 is transferred to the ball screw 5 which magnifies linear force through the reduction gear 4 and rotation of the ball screw 5 is converted to a linear motion of the nut 7 of the ball screw 5 and a linear motion of the screw 9 and pressure application to melted polymer are realized through the moving part 8. Position of the screw 9 is detected by a rotary encoder 12 mounted on the servomotor 3. Pressure applied to the melted polymer at the end of the barrel 2 is detected by the pressure detector 10 mounted between the nut 7 and the moving part 8. A cavity 13 in the metal mold 1 is filled up with melted polymer by a movement of the screw 9.

Mold good manufacturing consists of injection and dwell pressure application. In the injection process, polymer melt must be injected into the cavity 13 as fast as possible so that temperatures of polymer in the cavity become homogeneous. However, as excessive injection velocity brings about excessive polymer pressure and mold defects, polymer pressure in the injection process is constrained under a given pressure limit pattern. In the pressure application process following the injection process, a given pressure pattern is applied for each given duration at the polymer in the cavity during cooling in order to supply a deficiency due to polymer shrinkage. Therefore, the following two requirements are given to the injection velocity pattern and the pressure application pattern.

• (1) In the injection process, a given injection velocity pattern is realized and at the same time injection pressure is constrained under a given pressure limit pattern in terms of mold good quality.
• (2) In the pressure application process, a given pressure pattern is realized and at the same time injection velocity is constrained under a given velocity limit pattern in terms of safety operation.

In the injection process (time 0˜t₁) shown in FIG. 3(a), injection velocity control is carried out by giving injection velocity command shown in FIG. 3(b) to realize a given injection velocity pattern. However, injection pressure has to be controlled lower than a given pressure limit pattern shown in FIG. 3(c). Vertical scales 100% shown in FIGS. 3(b) and (c) indicate maximum values of injection velocity and injection pressure, respectively.

In the pressure application process (time t₁˜t₂) shown in FIG. 3(a), pressure application control is carried out by giving pressure application command shown in FIG. 3(c) to realize a given pressure application pattern. However, injection velocity has to be controlled lower than a given injection velocity limit pattern shown in FIG. 3(b).

FIG. 4 shows a block diagram of a controller which realizes the above two requirements (1) and (2) (paragraph {0008}) (patent literature PTL 1). The controller consists of an injection controller 20 and a motor controller (servoamplifier) 40.

The injection controller 20 executes a control algorithm at a constant time interval Δt and a discrete-time control is used. The injection controller 20 consists of an injection velocity setting device 21, a transducer 22, a pulse generator 23, an analog/digital (A/D) converter 25, an injection pressure setting device 26, a subtractor 27, a pressure controller 28, and a digital/analog (D/A) converter 29. The pressure detector 10 is connected to the injection controller 20.

The injection velocity setting device 21 feeds a time sequence of injection velocity command V_i* to the transducer 22. The transducer 22 calculates screw displacement command Δx_v* for the screw 9 which has to move during the time interval Δt by the following equation (1).{Math. 1}Δx_v=V_i*Δt  (1)

The command Δx_v* is fed to the pulse generator 23.

The pulse generator 23 feeds a pulse train 24 corresponding to the command Δx_v*. The pulse train 24 is fed to a pulse counter 41 in the motor controller 40.

The pressure detector 10 feeds an injection pressure signal P_i to the injection controller 20 through the A/D converter 25. The A/D converter 25 feeds the pressure signal P_i to the subtractor 27.

The injection pressure setting device 26 feeds a time sequence of injection pressure command P_i* to the subtractor 27. The subtractor 27 calculates a pressure control deviation ΔP_i by the following equation (2).{Math. 2}ΔP_i=P_i*−P_i  (2)

The control deviation ΔP_i is fed to the pressure controller 28.

The pressure controller 28 calculates a motor current demand i_p* from ΔP_i by using PID (Proportional+Integral+Derivative) control algorithm and feeds the demand i_p* to the motor controller 40 through the D/A converter 29.

Next, the motor controller 40 is explained. The motor controller 40 consists of pulse counters 41 and 44, an A/D converter 42, a comparator 43, subtractors 45 and 48, a position controller 46, a differentiator 47, a velocity controller 49 and a PWM (Pulse Width Modulation) device 50. The motor controller 40 is connected to the servomotor 3 equipped with the rotary encoder 12.

In the motor controller 40 the demand from the injection controller 20 is fed to the comparator 43 through the A/D converter 42.

The pulse counter 41 accumulates the pulse train 24 from the injection controller 20 and obtains screw position demand x* and feeds the demand x* to the subtractor 45. The pulse counter 44 accumulates the pulse train from the rotary encoder 12 and obtains actual screw position x and feeds the position signal x to the subtractor 45.

The subtractor 45 calculates a position control deviation (x*−x) and feeds the position deviation to the position controller 46. The position controller 46 calculates velocity demand v* by the following equation (3) and feeds the demand v* to the subtractor 48.{Math. 3}v*=K_p(x*−x)  (3)

where K_p is a proportional gain of the position controller 46.

The rotary encoder 12 feeds a pulse train to the differentiator 47 and to the pulse counter 44. The differentiator 47 detects actual screw velocity v and feeds the velocity signal v to the subtractor 48.

The subtractor 48 calculates a velocity control deviation (v*−v) and feeds the velocity deviation to the velocity controller 49. The velocity controller 49 calculates a motor current demand i_v* by the following equation (4) and feeds the demand i_v* to the comparator 43.

$\begin{array}{cc}\left\{\mathrm{Math}.4\right\}& \\ i_v^{*}=K_{P\upsilon }\left(v^{*}-v\right)+\frac{K_{Pv}}{T_{Iv}}\int \left(v^{*}-v\right)dt& \left(4\right)\end{array}$

where K_Pv and T_Iv are a proportional gain and an integral time constant of the velocity controller 49, respectively. In the motor controller 40 a position control loop has a minor loop of velocity control.

The comparator 43 to which both motor current demands i_v* and i_p* from the velocity controller 49 and the pressure controller 28, respectively, are fed, selects a lower current demand i* of i_v* and I_p* and feeds the lower demand i* to the PWM device 50. The PWM device 50 applies three-phase voltage to the servomotor 3 so that the servomotor 3 is driven by the motor current i*. The comparator 43 restricts motor current demand i_v* decided by injection velocity control loop to motor current demand i_p* decided by pressure control loop.

Next, it can be shown by using FIG. 3 that the above described two requirements (1) and (2) (paragraph {0008}) are realized by the comparator 43.

In FIG. 3, transfer time t₁ from injection to pressure application is specified by an operator, so the time t₁ should coincide with the time at which the cavity is filled up with polymer melt, but it is difficult for an operator to set the time t₁ at the exact time. Firstly the finishing time t₁ of injection process is supposed to be set by an operator before the time at which the cavity 13 is filled up with polymer melt actually. When the time reaches t₁ and pressure application process starts, actual pressure P_i is lower than a set value P_i* because the cavity is not yet filled and motor current demand i_p* fed by the pressure controller 28 increases so that pressure P_i is increased to the set value P_i*.

If demand i_p* is selected as a final motor current demand, injection velocity increases rapidly because the cavity 13 is still filling. Actual velocity could exceed the velocity limit shown in FIG. 3(b). However, even if demand i_p* exceeds i_v* when pressure application starts, the comparator 43 always selects a lower demand of i_p* and i_v* and selects the lower demand i_v* as a final motor current demand and limits velocity. That is, by the comparator 43 pressure application control is transferred to velocity limit control and the above described requirement (2) (paragraph {0008}) is always satisfied.

Secondly the time t₁ is supposed to be set by an operator after the time at which the cavity 13 is filled up actually. Even when the cavity is filled up, injection process continues and injection velocity control is carried out. But actual screw speed slows as the cavity is already filled up and so motor current demand i_v* fed by the velocity controller 49 is increased to maintain the injection velocity.

If demand i_v* is selected as a final motor current demand, injection pressure increases rapidly because filling is completed and so actual pressure may exceed the pressure limit shown in FIG. 3(c). However, even if demand i_v* exceeds i_p* in injection process when filling is completed, the comparator 43 always selects a lower demand of i_v* and i_p* and selects the lower demand i_p* as a final motor current demand and limits pressure in injection.

That is, by the comparator 43 injection velocity control is transferred to pressure limit control and the above described requirement (1) (paragraph {0008}) is always satisfied.

The object of the comparator 43 is to constrain an excessive pressure variation or an excessive velocity variation generated by a mutual transfer between the injection velocity control and the injection pressure control and a minimum selector (a low selector) which selects a smaller signal of inputted two signals found in patent literatures PTL 2 and PTL 3 has the same object as the above comparator 43.

In the controller shown in FIG. 4, the pressure detector 10 is absolutely necessary. Patent literatures PTL 4˜PTL 12 are applications of the apparatus and method for pressure control of injection molding machines without using the pressure detectors.

In patent literature PTL 4 for hydraulic actuator driven injection machines, polymer characteristics formula which gives the relational expression among polymer pressure, polymer temperature and polymer specific volume is used and the required polymer pressure is calculated by inputting measured polymer temperature and polymer specific volume which is decided from the desired value of mold good weight. Then by using initial temperatures of metal mold and polymer at the start of pressure application process and the above required polymer pressure, the required set value of pressure application is derived through an approximate expression. The pressure application set value is fed to the hydraulic servovalve amplifier as the voltage command converted and the set value of applied pressure is realized by the hydraulic pressure of hydraulic cylinder piston.

In patent literature PTL 5, in order to detect polymer pressure in the cavity the pressure is applied to a plunger which moves back and forth in the cavity and is connected with a ball screw mechanism whose nut is rotated by a servomotor. In injection and pressure application process the servomotor holds the position of the plunger to which the polymer pressure is applied and the servomotor current is detected by a current transducer and the detected current is converted to the polymer pressure in the cavity. The position of the plunger is detected by a rotary encoder equipped with the servomotor.

In patent literature PTL 6, in order to detect polymer pressure in the cavity a disturbance observer is used for a servomotor drive system which moves a plunger back and forth in the cavity. In injection and pressure application process the pressure is applied to the plunger and the servomotor drive system holds the position of the plunger. Then the disturbance observer estimates a load torque of the servomotor by using a motor speed signal and a motor torque command signal. The pressure in the cavity is obtained from the estimated load torque. The arithmetic expressions of the observer are shown in the literature. The method by which pressure in the cavity is obtained directly by using detected servomotor current or motor torque command, is also shown in the literature.

In patent literature PTL 7, firstly a function which estimates a polymer pressure in the cavity by using injection screw drive force and injection velocity, is decided. In the actual control actions, feature size of mold good, polymer data, real-time data of screw drive force and injection velocity are fed to the above function and the real-time estimated pressure in the cavity is obtained. Injection velocity is controlled by deviation of the estimated pressure from the reference value. The procedures of obtaining the above exact function are shown in the literature.

In patent literature PTL 8, the pressure control apparatus is realized in which an observer for a servomotor drive system estimates a polymer pressure and the estimated pressure is used as a detected signal for the pressure control. The observer is fed by a motor speed in an injection process and the total friction resistance in an injection mechanism and outputs the estimates of motor speed and polymer pressure. The observer is applied to the following two models.

• (1) A servomotor drives an injection screw through a linear motion converter such as a ball screw only.
• (2) A servomotor drives an injection screw through a belt pulley reduction gear and a linear motion converter.

In the observer model (1), the friction resistance consists of a dynamic friction resistance and a static friction resistance over an injection mechanism. In the observer model (2), the friction resistance consists of a dynamic friction resistance only which is defined as a sum of a velocity dependent component and a load dependent component.

In the observer model (1) the polymer pressure is assumed to be constant. In the observer model (2) the observer outputs not only the estimates of motor speed and polymer pressure but also the estimates of pulley speed at load side, belt tension and force applied to polymer melt by a screw. It is assumed that the belt is elastic and the time rate of change in polymer pressure is proportional to pulley speed at load side, to pulley acceleration and to force applied to polymer by a screw. The force applied to polymer melt by a screw is assumed to be constant.

In patent literature PTL 9, an injection velocity and pressure control apparatus is realized in which an observer for a servomotor drive system estimates the load torque generated by polymer pressure. The observer is fed by motor speed signal and motor current command signal and outputs the estimates of motor speed, load torque and position difference between a motor shaft and a load side shaft. The model of the observer consists of the servomotor drive system which moves a screw through a belt pulley reduction gear. The pressure value converted from the estimated load torque obtained by the observer is used as a detected pressure signal.

In patent literature PTL 10, the observer for the model (2) in PTL 6 (paragraph {0041}) is used and the pressure control method is invented. The method uses a motor speed and the estimates of pulley speed at load side, belt tension and polymer pressure as feedback signals of state variables. These estimates are obtained by the observer. Another control method is invented, in which the servomotor torque command is decided by the above four state variables.

In patent literature PTL 11, a method is invented, in which the estimate of load torque applied by polymer pressure is obtained by using an inverse model of a transfer function whose inputs are a motor generated torque and a load torque and output is a motor speed. The inverse model is fed by a motor speed and a motor generated torque and derives the estimate of load torque. The polymer pressure is obtained from the estimate of the load torque. The inverse model requires the high-order differentiation.

In patent literature PTL 12, a method is invented, in which a polymer pressure is estimated by using a motion equation of an integral structure consisted of a moving member and a screw and by using measured values of a motor rotation speed and a motor torque. The motion equation of the integral structure converted to the motor axis is shown by the following equation (5). The motor shaft and the ball screw shaft are coupled through a belt pulley reduction gear.

$\begin{array}{cc}\left\{\mathrm{Math}.5\right\}& \\ J_{\mathrm{TOT}}\frac{d\omega }{dt}=T_2-\frac{l}{2\pi }\frac{1}{e_S{e}_B}\frac{N_{\mathrm{MP}}}{N_{\mathrm{SP}}}\left(A_{\mathrm{BARREL}}{P}_{\mathrm{MELT}}+F_{\mathrm{LOSS}}\right)-T_U& \left(5\right)\end{array}$

where t: Time variable, J_TOT: Reduced total moment of inertia at motor axis, ω: Angular velocity of motor, T₂: Motor torque, l: Ball screw lead, e_S: Ball screw efficiency, e_B: Belt pulley efficiency, N_MP: Pulley diameter at motor side, N_SP: Pulley diameter at ball screw side, A_BARREL: Barrel section area, P_MELT: Polymer pressure, F_LOSS: Total friction force acting on the integral structure due to a friction at a ball screw mechanism and due to a friction between a barrel surface and a screw, T_U: Torque loss due to a friction force at a support rail of the integral structure. Angular velocity of motor ω and motor torque T₂ are measured. Angular acceleration of motor α−dω/dt is obtained by a numerical time differential operation for an angular velocity of motor ω. If F_LOSS, T_U etc. are known, P_MELT is obtained by using the following equation (6).

$\begin{array}{cc}\left\{\mathrm{Math}.6\right\}& \\ P_{\mathrm{MELT}}=\frac{1}{A_{\mathrm{BARREL}}}\left\{\frac{2\pi }{l}\frac{N_{\mathrm{SP}}}{N_{\mathrm{MP}}}{e}_S{e}_B\left(T_2-J_{\mathrm{TOT}}\alpha -T_U\right)-F_{\mathrm{LOSS}}\right\}& \left(6\right)\end{array}$

In the estimation method of polymer pressure by equation (6), an error of angular acceleration of motor a due to the numerical differential operation and errors of F_LOSS and T_U bring about the error of the estimated polymer pressure.

The common object of inventions described in patent literatures PTL 4˜PTL 12 which detect polymer pressure without using a pressure detector is to avoid the following disadvantages.

• (1) A highly reliable pressure detector is very expensive under high pressure circumstances.
• (2) Mounting a pressure detector in the cavity or the barrel nozzle part necessitates the troublesome works and the working cost becomes considerable.
• (3) Mounting a load cell in an injection shafting alignment from a servomotor to a screw complicates the mechanical structure and degrades the mechanical stiffness of the structure.
• (4) A load cell which uses strain gauges as a detection device necessitates an electric protection against noise for weak analog signals. Moreover the works for zero-point and span adjustings of a signal amplifier are necessary (patent literature PTL 13).

Patent literatures PTL 14 and PTL 15 are inventions concerning the pressure control at electric-motor driven injection molding machines and both necessitate pressure detectors. In patent literature PTL 14, the concept of virtual screw velocity ω₁ is introduced in equation (1) of the description of PTL 14 and equation (1) is based on the point of view which the pressure control is conducted by screw position control. An exact control method is realized by using virtual velocity ω₁ as a parameter which compensates the pressure loss due to the nonlinear loss which results in the difference between the pressure corresponding to motor generated torque and pressure set value. The disturbance observer outputs the estimate of virtual velocity ω₁ so that the difference between the pressure detected by a load cell and the estimated pressure becomes zero by using the error between detected pressure and the estimated pressure. Patent literature PTL 15 is a prior application of PTL 14 and it is different from PTL 14 in the observer structure.

CITATION LIST

Patent Literature

• PTL 1: Patent No. 3787627
• PTL 2: Patent 6-55599
• PTL 3: Patent 2000-202875
• PTL 4: Patent 5-77298
• PTL 5: Patent 6-856
• PTL 6: Patent 7-299849
• PTL 7: Patent 9-277325
• PTL 8: Patent WO2005/028181
• PTL 9: Patent 2006-142659
• PTL 10: Patent 2006-256067
• PTL 11: Patent 2008-265052
• PTL 12: U.S. Pat. No. 6,695,994
• PTL 13: Patent 2003-211514
• PTL 14: Patent 10-244571
• PTL 15: Patent 10-44206

Non Patent Literature

• NPL 1: H. K. Khalil, Nonlinear Systems, 14.5 High-Gain Observers, Prentice-Hall, (2002), pp. 610-625
• NPL 2: B. D. O. Anderson and J. B. Moore, Optimal Control, Linear Quadratic Methods, 7.2 Deterministic Estimator Design, Prentice-Hall, (1990), pp. 168-178
• NPL 3: A. M. Dabroom and H. K. Khalil, Discrete-time implementation of high-gain observers for numerical differentiation, Int. J. Control, Vol. 72, No. 17, (1999), pp. 1523-1537
• NPL 4: A. M. Dabroom and H. K. Khalil, Output Feedback Sampled-Data Control of Nonlinear Systems Using High-Gain Observers, IEEE Trans. Automat. Contr., Vol. 46, No. 11, (2001), pp. 1712-1725

SUMMARY OF INVENTION

Technical Problem

The problem that starts being solved is to realize a pressure control apparatus and a pressure control method of electric-motor driven injection molding machines which satisfy the above two requirements (1) and (2) (paragraph {0008}) without using a pressure detector in order to avoid the four disadvantages described in Background Art (paragraph {0051}) resulted by using a pressure detector.

Solution to Problem

Mold good manufacturing consists of injection and dwell pressure application. In the injection process injection pressure has to be constrained under a given pressure limit pattern by the requirement (1) (paragraph {0008}) and so it is necessary to detect the actual pressure without time-lag. In the pressure application process a given pressure pattern has to be realized by the requirement (2) (paragraph {0008}) and so it is necessary to detect the applied pressure without time-lag. Therefore a pressure detecting means is required to have no time-lag.

As the error of detected pressure causes mold good defects and lack of safety operation, the exact pressure detection is required. Therefore the method of a high-gain observer (non patent literature NPL 1) is used to realize a pressure detecting means which satisfies the following two requirements (A) and (B).

(A) The detection means is high-precision.

(B) The detection means has very small time-lag.

A high-gain observer estimates all state variables by using detected variables. This is explained by using a simple mathematical model as follows. Equation (7) shows a state equation and an output equation of a simple model.

$\begin{array}{cc}\left\{\mathrm{Math}.7\right\}& \\ \begin{array}{c}{\stackrel{.}{x}}_1=x_2\\ {\stackrel{.}{x}}_2=\varphi \left(x,u\right)\\ y=x_1\\ x=\left[\begin{array}{c}{x}_1\\ x_2\end{array}\right]\end{array}\}& \left(7\right)\end{array}$ where x₁, x₂: State variables, u: Input variable, y: Output variable, φ(x, u): Nonlinear function of variables x, u. For example x₁ is position variable, x₂ is velocity variable and a is motor current variable. Output variable y and input variable a are supposed to be measurable. The high-gain observer which estimates state x is given by equation (8).

$\begin{array}{cc}\left\{\mathrm{Math}.8\right\}& \\ \begin{array}{c}{\stackrel{\stackrel{.}{^}}{x}}_1={\hat{x}}_2+H_1\left(y-{\hat{x}}_1\right)\\ {\stackrel{\stackrel{.}{^}}{x}}_2={\varphi }_0\left(\hat{x},u\right)+H_2\left(y-{\hat{x}}_1\right)\end{array}\}& \left(8\right)\end{array}$

where {circumflex over (x)}₁, {circumflex over (x)}₂: Estimates of state variables x₁, x₂, H₁, H₂: Gain constants of the high-gain observer which are larger than 1, φ₀: Nominal function of φ used in the high-gain observer computing. Estimation errors {tilde over (x)}₁, {tilde over (x)}₂ by using the high-gain observer (8) are given by equation (9) from equations (7) and (8).

$\begin{array}{cc}\left\{\mathrm{Math}.9\right\}& \\ \begin{array}{c}{\stackrel{\stackrel{.}{~}}{x}}_1=-H_1{\tilde{x}}_1+{\tilde{x}}_2\\ {\stackrel{\stackrel{.}{~}}{x}}_2=-H_2{\tilde{x}}_1+\delta \left(x,\tilde{x},u\right)\end{array}\}& \left(9\right)\\ \begin{array}{c}{\tilde{x}}_1=x_1-{\tilde{x}}_1\\ {\tilde{x}}_2=x_2-\widetilde{x_2}\\ \delta \left(x,\tilde{x},u\right)=\varphi \left(x,u\right)-{\varphi }_0\left(\tilde{x},u\right)\end{array}\}& \left(10\right)\end{array}$

where δ: Model error between the nominal model φ₀ and the true but actually unobtainable function φ. Introducing a positive parameter ε much smaller than 1, H₁, H₂ are given by equation (11).

$\begin{array}{cc}\left\{\mathrm{Math}.10\right\}& \\ H_1=\frac{K_1}{\varepsilon }{H}_2=\frac{K_2}{{\varepsilon }^2}& \left(11\right)\end{array}$

As H₁, H₂ in equation (11) are large gain constants, equation (8) is called by a high-gain observer. By using equation (11), equation (9) is rewritten as equation (12).

$\begin{array}{cc}\left\{\mathrm{Math}.11\right\}& \\ \begin{array}{c}{\stackrel{\stackrel{.}{~}}{x}}_1=-K_1\left({\tilde{x}}_1/\varepsilon \right)+{\tilde{x}}_2\\ {\stackrel{\stackrel{.}{~}}{x}}_2=-\left(K_2/\varepsilon \right)\left({\tilde{x}}_1/\varepsilon \right)+\delta \left(x,\tilde{x},u\right)\end{array}\}& \left(12\right)\end{array}$

The estimation errors {circumflex over (x)}₁, {circumflex over (x)}₂ are replaced by new variables η₁, η₂ as written in equation (13).

$\begin{array}{cc}\left\{\mathrm{Math}.12\right\}& \\ {\eta }_1=\frac{{\tilde{x}}_1}{\varepsilon }{\eta }_2={\tilde{x}}_2& \left(13\right)\end{array}$

By using equation (13), equation (12) is rewritten as equation (14).

$\begin{array}{cc}\left\{\mathrm{Math}.13\right\}& \\ \begin{array}{c}\varepsilon {\stackrel{.}{\eta }}_1=-K_1{\eta }_1+{\eta }_2\\ \varepsilon {\stackrel{.}{\eta }}_2=-K_2{\eta }_1+\varepsilon \delta \left(x,\eta ,u\right)\end{array}\}& \left(14\right)\end{array}$

As the parameter ε is much smaller than 1, the effects of model error δ on the estimation errors η₁, η₂ can be made small enough by equation (14). Thus by using the high-gain observer for a model which has injection pressure as a state variable, the above requirement (A) “High-precision detection” for a pressure detecting means (paragraph {0057}) is satisfied.

When the effects of the model error δ on the estimation errors η₁, η₂ are neglected, equation (14) is rewritten as equation (15).

$\begin{array}{cc}\left\{\mathrm{Math}.14\right\}& \\ \left[\begin{array}{c}{\stackrel{.}{\eta }}_1\\ {\stackrel{.}{\eta }}_2\end{array}\right]=\frac{1}{\varepsilon }\left[\begin{array}{cc}-K_1& 1\\ -K_2& 0\end{array}\right]\left[\begin{array}{c}{\eta }_1\\ {\eta }_2\end{array}\right]=\frac{1}{\varepsilon }A\left[\begin{array}{c}{\eta }_1\\ {\eta }_2\end{array}\right]& \left(15\right)\\ A=\left[\begin{array}{cc}-K_1& 1\\ -K_2& 0\end{array}\right]& \left(16\right)\end{array}$

When K₁, K₂ are decided so that conjugate complex eigenvalues λ₁, λ₁ of matrix A have a negative real part, that is, Re(λ₁)=Re( λ₁)<0, the estimate errors η₁, η₇₂ are given by equation (17) with initial values η₁₀, η₂₀ from equation (15).

$\begin{array}{cc}\left\{\mathrm{Math}.15\right\}& \\ \begin{array}{c}{\eta }_1\left(t\right)=\exp \left(\frac{\mathrm{Re}\left({\lambda }_1\right)}{\varepsilon }t\right)\left(C_1\left(t\right){\eta }_{10}+C_2\left(t\right){\eta }_{20}\right)\\ {\eta }_2\left(t\right)=\exp \left(\frac{\mathrm{Re}\left({\lambda }_1\right)}{\varepsilon }t\right)\left(C_3\left(t\right){\eta }_{10}+C_4\left(t\right){\eta }_{20}\right)\end{array}\}& \left(17\right)\end{array}$

where t: Time variable, C₁(t)˜C₄ (t): Sinusoidal components with constant amplitudes and constant frequency decided by K₁, K₂. As Re(λ₁)<0 and ε is much smaller than 1, equation (17) reveals that the time responses η₁(t), η₂(t) of estimation errors tend to zero rapidly. In other words, by using high-gain observer equation (8), the above requirement (B) “Detection with small time-lag” for a pressure detecting means (paragraph {0057}) can be satisfied.

Although estimates {circumflex over (x)}₁, {circumflex over (x)}₂ of all state variables are obtained by equation (8), it is sufficient to get only the estimate {circumflex over (x)}₂ because x₁ is detected as output y. Then the high-gain observer is given by equation (18) (non patent literature NPL 2).{Math. 16}{circumflex over ({dot over (x)}₂=−H{circumflex over (x)}₂+H{dot over (y)}+φ₀({circumflex over (x)}₂,y,u)  (18)

where H: Gain constant of the high-gain observer which is larger than 1. As time-derivative term of output y is included in the right-hand side of equation (18), equation (18) cannot be used as a computing equation by itself. But it can be shown that the high-gain observer by equation (18) satisfies the above two requirements (A) and (B) (paragraph {0057}). Equation (19) is given from the third equation in equation (7).{Math. 17}{dot over (y)}={dot over (x)}₁=x₂  (19)

Equation (20) is given by using equations (18) and (19).{Math. 18}{circumflex over ({dot over (x)}₂=H{circumflex over (x)}₂+φ₀({circumflex over (x)}₂,y,u)  (20)

By using the second equation of equation (7), equation (21) is given from equation (20).

$\begin{array}{cc}\left\{\mathrm{Math}.19\right\}& \\ {\stackrel{\stackrel{.}{~}}{x}}_2=-H{\tilde{x}}_2+\delta \left(x,{\tilde{x}}_2,y,u\right)& \left(21\right)\\ \begin{array}{c}{\tilde{x}}_2=x_2-{\hat{x}}_2\\ \delta \left(x,{\tilde{x}}_2,y,u\right)=\varphi \left(x,u\right)-{\varphi }_0\left({\hat{x}}_2,y,u\right)\end{array}\}& \left(22\right)\end{array}$

Gain constant H is given by equation (23) by introducing a positive parameter ε much smaller than 1.

$\begin{array}{cc}\left\{\mathrm{Math}.20\right\}& \\ H=\frac{K}{\varepsilon }\left(K>0\right)& \left(23\right)\end{array}$

By using equation (23), equation (21) is rewritten as equation (24).{Math. 21}ε{tilde over ({dot over (x)}₂=−K{tilde over (x)}₂+εδ(x,{tilde over (x)}₂,y,u)  (24)

As ε is much smaller than 1, the effect of model error δ on the estimation error {tilde over (x)}₂ can be made small enough from equation (24). Therefore by using the high-gain observer for a model which has injection pressure as a state variable, the above requirement (A) “High-precision detection” for a pressure detecting means (paragraph {0057}) can be satisfied.

When the effect of model error δ on the estimation error {tilde over (x)}₂ is neglected, equation (24) is rewritten as equation (25).{Math. 22}ε{tilde over ({dot over (x)}₂=−K{tilde over (x)}₂  (25)

The estimation error {tilde over (x)}₂ is given by equation (26) from equation (25).

$\begin{array}{cc}\left\{\mathrm{Math}.23\right\}& \\ {\tilde{x}}_2\left(t\right)=\exp \left(-\frac{K}{\varepsilon }t\right){\tilde{x}}_{20}& \left(26\right)\end{array}$

where {tilde over (x)}₂₀: Initial value of {tilde over (x)}₂. As ε is much smaller than 1, equation (26) reveals that the time response {tilde over (x)}₂ (t) of estimation error tends to zero rapidly. In other words, by using high-gain observer equation (18), the above requirement (B) “Detection with small time-lag” for a pressure detecting means (paragraph {0057}) can be satisfied. As in equation (18) the minimum number of state variables to be estimated are included and the measurable state variables are excluded, equation (18) is called by a reduced-order high-gain observer because the order of observer equation (18) is lower than that of observer equation (8).

Then a procedure to modify equation (18) is shown so that the time-derivative term of output y is not appeared. A new variable w is given by equation (27).{Math. 24}ŵ={circumflex over (x)}₂−H_y  (27)

By using equation (27), equation (18) is rewritten as equation (28).{Math. 25}{circumflex over ({dot over (w)}==H(ŵ+Hy)+φ₀(ŵ,y,u)  (28)

Variable ŵ is calculated by equation (28) and estimate {circumflex over (x)}₂ is obtained by equation (29).{Math. 26}{circumflex over (x)}₂=ŵ+Hy  (29)

Procedures of applying a high-gain observer for a model of electric-motor driven injection molding machines which has injection pressure as a state variable are described in detail in Example to be hereinafter described.

Advantageous Effects of Invention

By applying a high-gain observer for a model of electric-motor driven injection molding machines which has injection pressure as a state variable, a high-precision pressure detection with small time-lag becomes possible without using a pressure detector. By using the high-gain observer the two requirements (1) and (2) described in paragraph {0008} for controlling pressure of electric-motor driven injection molding machines can be satisfied and also the four disadvantages described in Background Art (paragraph {0051}) can be avoided.

BRIEF DESCRIPTION OF DRAWINGS

FIG. 1 is an explanation drawing of a working example which shows a system configuration of an apparatus and a method for pressure control of an electric-motor driven injection molding machine according to an embodiment of the present invention.

FIG. 2 is a view which shows an existing injection and pressure application mechanism of an electric-motor driven injection molding machine.

FIG. 3 is an explanation drawing which shows a time schedule of mold good manufacturing.

FIG. 4 is an explanation drawing which shows a system configuration of an existing apparatus and method for pressure control of an electric-motor driven injection molding machine.

FIG. 5 is a view which shows an injection and pressure application mechanism of an electric-motor driven injection molding machine according to an embodiment of the present invention.

FIG. 6 is an explanation drawing of a working example which shows computer simulation results of injection pressure estimation by the high-gain observer according to an embodiment of the present invention.

DESCRIPTION OF EMBODIMENT

Hereinafter, the embodiment of the present invention on the controller of electric-motor driven injection molding machines is described based on the drawings.

Example

FIG. 5 is a view which shows an injection and pressure application mechanism without using a pressure detector. As the mechanism in FIG. 5 consists of the parts with the same reference signs as in FIG. 2 except a pressure detector 10, explanations of FIG. 5 are replaced by those of FIG. 2 described in Background Art (paragraph {0004}˜{0006}).

FIG. 2 is an example of a controller of an electric-motor driven injection molding machine using a high-gain observer as an injection pressure detecting means according to an embodiment of the present invention and shows a block diagram of a system configuration for the controller. The controller consists of an injection controller 20 which contains a high-gain observer 31 and a motor controller (servoamplifier) 40.

The injection controller 20 is explained as follows. The injection controller 20 executes a control algorithm at a constant time interval Δt and feeds a discrete-time control demand to the motor controller 40. The injection controller 20 consists of an injection velocity setting device 21, a transducer 22, a pulse generator 23, an injection pressure setting device 26, a subtractor 27, a pressure controller 28, a D/A converter 29, an A/D converter 30 and a high-gain observer 31.

The injection velocity setting device 21 feeds a time sequence of injection velocity command V_i* to the transducer 22. The transducer 22 calculates screw displacement command Δx_v* for the screw 9 which has to move during the time interval Δt by the following equation (30).{Math. 27}Δx_v*=V_i*Δt  (30)

The command Δx_v* is fed to the pulse generator 23 which feeds a pulse train 24 corresponding to the command Δx_v*. The pulse train 24 is fed to a pulse counter 41 in the motor controller 40.

The injection pressure setting device 26 feeds a time sequence of injection pressure command P_i* to the subtractor 27. Motor current demand i* in the motor controller 40 is fed to the high-gain observer 31 through the A/D converter 30 in the injection controller 20. The screw velocity signal v which is fed by a differentiator 47 in the motor controller 40 is fed to the high-gain observer 31. The high-gain observer 31 executes discrete-time arithmetic expressions which are obtained from a mathematical model of an injection mechanism and outputs an estimate of injection pressure {circumflex over (P)}_i by using the input signals v and i*.

The estimate {circumflex over (P)}_i is fed to the subtractor 27. The subtractor 27 calculates a pressure control deviation ΔP_i from injection pressure command P_i* by the following equation (31).{Math. 28}ΔP_i=P_i*−P_i*  (31)

The subtractor 27 feeds ΔP_i to the pressure controller 28.

The pressure controller 28 calculates a motor current demand i_p* from ΔP_i by using PID control algorithm and feeds the demand i_p* to the motor controller 40 through the D/A converter 29.

The motor controller 40 is explained as follows. The motor controller 40 consists of pulse counters 41 and 44, an A/D converter 42, a comparator 43, subtractors 45 and 48, a position controller 46, a differentiator 47, a velocity controller 49, and a PWM device 50. The motor controller 40 is connected to the servomotor 3 equipped with the rotary encoder 12.

The motor current demand i_p* is fed to the motor controller 40 by the injection controller 20 through the A/D converter 42 and is fed to the comparator 43.

The pulse counter 41 accumulates the pulse train 24 from the injection controller 20 and obtains screw position demand x* and outputs the demand x* to the subtractor 45. The pulse counter 44 accumulates the pulse train from the rotary encoder 12 and obtains the actual screw position x and feeds the position x to the subtractor 45.

The subtractor 45 calculates a position control deviation (x*−x) by using inputted signals x* and x and feeds the position deviation to the position controller 46. The position controller 46 calculates velocity demand v* by the following equation (32) and feeds the demand v* to the subtractor 48.{Math. 29}v*=K_p(x*−x)  (32)

where K_P is a proportional gain constant of the position controller 46. The rotary encoder 12 feeds the pulse train to the differentiator 47 and to the pulse counter 44. The differentiator 47 detects an actual screw velocity v and feeds the velocity v to the subtractor 48.

The subtractor 48 calculates a velocity control deviation (v*−v) by using inputted signals v* and v and feeds the deviation to the velocity controller 49. The velocity controller 49 calculates a motor current demand i_v* by the following equation (33) and feeds the demand i_v* to the comparator 43.

$\begin{array}{cc}\left\{\mathrm{Math}.30\right\}& \\ i_{\upsilon }^{*}=K_{P_{\upsilon }}\left({\upsilon }^{*}-\upsilon \right)+\frac{K_{P_{\upsilon }}}{T_{I_{\upsilon }}}\int \left({\upsilon }^{*}-\upsilon \right)dt& \left(33\right)\end{array}$

where K_Pv and T_Iv are a proportional gain constant and an integral time constant of the velocity controller 49, respectively.

The comparator 43 to which motor current demands i_v* and i_p* are given from the velocity controller 49 and the pressure controller 28, respectively, selects the lower current demand i* of i_v* and i_p* and feeds the demand i* to the PWM device 50. The PWM device 50 applies three-phase voltage to the servomotor 3 so that the servomotor 3 is driven by the motor current demand i*.

That the above two requirements (1) and (2) (paragraph {0008}) for the controller of electric-motor driven injection molding machines are realized by the comparator 43, is already described in detail in Background Art (paragraph {0033}˜{0037}).

The high-gain observer 31 outputs an estimate {circumflex over (P)}_i of injection pressure by using screw velocity signal v and motor current demand signal i*. The mathematical model of an injection mechanism shown in FIG. 5 is derived as follows, which is necessary to design the high-gain observer 31. A motion equation of the motor 3 axis is given by equation (34).

$\begin{array}{cc}\left\{\mathrm{Math}.31\right\}& \\ \left(J_M+J_{\mathrm{G1}}\right)\frac{d{\omega }_m}{dt}=T_M-r_{1}F& \left(34\right)\end{array}$

where J_M: Moment of inertia of motor itself, J_G1: Moment of inertia of motor-side gear, ω_m: Angular velocity of motor, T_M: Motor torque, r₁: Radius of motor-side gear, F: Transmission force of reduction gear, t: Time variable. A motion equation of the ball screw 5 axis is given by equation (35).

$\begin{array}{cc}\left\{\mathrm{Math}.32\right\}& \\ \left(J_S+G_{\mathrm{G2}}\right)\frac{d{\omega }_s}{dt}=r_{2}F-T_{\alpha }& \left(35\right)\end{array}$

where J_S: Moment of inertia of ball screw axis, J_G2: Moment of inertia of load-side gear, ω_s: Angular velocity of ball screw axis, r₂: Radius of load-side gear, T_a: Ball screw drive torque. A motion equation of the moving part 8 is given by equations (36) and (37).

$\begin{array}{cc}\left\{\mathrm{Math}.33\right\}& \\ \frac{W}{g}\frac{d\upsilon }{dt}=F_a-F_L-\mu W\frac{\upsilon }{\upsilon }& \left(36\right)\\ \frac{dx}{dt}=\upsilon & \left(37\right)\end{array}$

where W: Weight of the moving part 8, g: Gravity acceleration, v: Velocity of the moving part (the screw), x: Screw position (initial position x−0), F_a: Drive force of the ball screw, F_L: Applied force by polymer to the screw, μ: Friction coefficient at the slider. A relation between ball screw drive force F_a and ball screw drive torque T_a is given by equation (38).

$\begin{array}{cc}\left\{\mathrm{Math}.34\right\}& \\ T_a=\frac{l}{2\pi }\frac{1}{\eta }{F}_a& \left(38\right)\end{array}$

where l: Ball screw lead, η: Ball screw efficiency. Equations among v, ω_s and ω_m are given by equation (39).

$\begin{array}{cc}\left\{\mathrm{Math}.35\right\}& \\ \upsilon =\frac{l}{2\pi }{\omega }_s=\frac{l}{2\pi }\frac{r_1}{r_2}{\omega }_m& \left(39\right)\end{array}$

Applied force to the screw F_L is given by equation (40).

$\begin{array}{cc}\left\{\mathrm{Math}.36\right\}& \\ F_L=A_s{P}_i+C_{\mathrm{mt}}\frac{\upsilon }{\upsilon }{\upsilon }^{\gamma }& \left(40\right)\end{array}$

where A_s: Screw section area, P_i: Injection pressure which means polymer pressure at the end of a barrel, C_mt: Friction coefficient between the screw and the barrel surface, γ: Velocity power coefficient. A dynamic equation of injection pressure P_i is given by equations (41) and (42).

$\begin{array}{cc}\left\{\mathrm{Math}.37\right\}& \\ \frac{V_i}{\beta }\frac{d{P}_i}{dt}=A_s\upsilon -Q_{\mathrm{in}}& \left(41\right)\\ V_i=V_{\mathrm{i0}}-A_{s}x& \left(42\right)\end{array}$

where V_i: Polymer volume at the end of a barrel, V_i0: Initial volume of V_i, Q_in: Injected rate of polymer, β: Bulk modulus of polymer. The characteristics of the servomotor 3 is given by equation (43).{Math. 38}T_M=K_Ti_m  (43)

where K_T: Motor torque coefficient, i_m: Motor current. By using equations (34), (35) and (39) and deleting ω_s and F, equation (44) is derived.

$\begin{array}{cc}\left\{\mathrm{Math}.39\right\}& \\ \left\{J_M+J_{\mathrm{G1}}+\left(J_S+J_{\mathrm{G2}}\right){\left(\frac{r_1}{r_2}\right)}^2\right\}\frac{d{\omega }_m}{dt}=T_M-\frac{r_1}{r_2}{T}_a& \left(44\right)\end{array}$

By using equations (36), (38), (39) and (44) and deleting T_a and F_a, equation (45) is derived.

$\begin{array}{cc}\left\{\mathrm{Math}.40\right\}& \\ J_{\mathrm{eq}}\frac{d{\omega }_m}{dt}=T_M-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\left(F_L+\mu W\frac{\upsilon }{\upsilon }\right)& \left(45\right)\\ J_{\mathrm{eq}}=J_M+J_{G1}+\left(J_S+J_{G1}\right){\left(\frac{r_1}{r_2}\right)}^2+\frac{W}{g}{\left(\frac{r_1}{r_2}\right)}^2{\left(\frac{l}{2\pi }\right)}^2\frac{1}{\eta }& \left(46\right)\end{array}$

where J_eq: Reduced total moment of inertia at motor axis. Equation (45) is the motion equation of a total injection molding mechanism converted to the motor axis. From equations (37) and (39), equation (47) is derived.

$\begin{array}{cc}\left\{\mathrm{Math}.41\right\}& \\ \frac{dx}{dt}=\frac{r_1}{r_2}\frac{l}{2\pi }{\omega }_m& \left(47\right)\end{array}$

From equations (40), (43) and (45), the motion equation of the total mechanism is given by equation (48).

$\begin{array}{cc}\left\{\mathrm{Math}.42\right\}& \\ J_{\mathrm{eq}}\frac{d{\omega }_m}{dt}=K_T{i}_m-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\left\{A_s{P}_i+C_{\mathrm{mt}}\frac{\upsilon }{\upsilon }{\upsilon }^{\gamma }+\mu W\frac{\upsilon }{\upsilon }\right\}& \left(48\right)\end{array}$

Equation (42) is rewritten as equation (49).{Math. 43}V_i=A_s(x_max−x)  (49)where x_max: Maximum screw stroke. By using equations (39) and (49), equation (41) is rewritten as equation (50).

$\begin{array}{cc}\left\{\mathrm{Math}.44\right\}& \\ \frac{A_s\left(x_{\max }-x\right)}{\beta }\frac{d{P}_i}{dt}=A_s\frac{l}{2\pi }\frac{r_1}{r_2}{\omega }_m-Q_{\mathrm{in}}& \left(50\right)\end{array}$

The variables in the above equations are made dimensionless. By using dimensionless variables, equation (47) is rewritten as equation (51).

$\begin{array}{cc}\left\{\mathrm{Math}.45\right\}& \\ \begin{array}{c}\frac{d}{dt}\left[\frac{x}{x_{\max }}\right]=\frac{l}{2\pi }\frac{r_1}{r_2}\frac{{\omega }_{\max }}{x_{\max }}\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]\\ =\frac{{\upsilon }_{\max }}{x_{\max }}\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]\end{array}& \left(51\right)\end{array}$

where ω_max: Motor rating speed, v_max: Maximum injection velocity.

By using dimensionless variables, equation (48) is rewritten as equation (52).

$\begin{array}{cc}\left\{\mathrm{Math}.46\right\}& \\ J_{\mathrm{eq}}{\omega }_{\max }\frac{d}{dt}\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]=K_T{i}_{\max }\left[\frac{i_{\max }}{i_{\max }}\right]-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}{A}_s{P}_{\max }\left[\frac{P_i}{P_{\max }}\right]-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\frac{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}\left\{C_{\mathrm{mt}}{\upsilon }_{\max }^{\gamma }{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}^{\gamma }+\mu W\right\}& \left(52\right)\end{array}$

where i_max: Motor current rating, P_max: Maximum injection pressure. In deriving equation (52), equation (53) is used.

$\begin{array}{cc}\left\{\mathrm{Math}.47\right\}& \\ \left[\frac{\upsilon }{{\upsilon }_{\max }}\right]=\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]& \left(53\right)\end{array}$

Equation (52) is rewritten as equation (54).

$\begin{array}{cc}\left\{\mathrm{Math}.48\right\}& \\ \frac{d}{dt}\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]=\left[\frac{T_{M\max }}{J_{\mathrm{eq}}{\omega }_{\max }}\right]\left[\frac{i_m}{i_{\max }}\right]-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\frac{A_s{P}_{\max }}{J_{\mathrm{eq}}{\omega }_{\max }}\left[\frac{P_i}{P_{\max }}\right]-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\frac{1}{J_{\mathrm{eq}}{\omega }_{\max }}\frac{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}\left\{C_{\mathrm{mt}}{\upsilon }_{\max }^{\gamma }{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}^{\gamma }+\mu W\right\}& \left(54\right)\end{array}$

where T_Mmax=K_Ti_max: Motor rating torque.

By using dimensionless variables, equation (50) is rewritten as equation (55).

$\begin{array}{cc}\left\{\mathrm{Math}.49\right\}& \\ \frac{1}{\beta }{A}_s{x}_{\max }{P}_{\max }\left\{1-\left[\frac{x}{x_{\max }}\right]\right\}\frac{d}{dt}\left[\frac{P_i}{P_{\max }}\right]=A_s{\upsilon }_{\max }\left\{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]-\left[\begin{array}{c}{Q}_{\mathrm{in}}\\ Q_{\max }\end{array}\right]\right\}& \left(55\right)\end{array}$

where Q_max=A_sv_max: Maximum injection rate. Equation (55) is rewritten as equation (56).

$\begin{array}{cc}\left\{\mathrm{Math}.50\right\}& \\ \frac{d}{dt}\left[\frac{P_i}{P_{\max }}\right]=\frac{\beta }{1-\left[\frac{x}{x_{\max }}\right]}\frac{{\upsilon }_{\max }}{x_{\max }{P}_{\max }}\left\{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]-\left[\frac{Q_{\mathrm{in}}}{Q_{\max }}\right]\right\}& \left(56\right)\end{array}$

In general dimensionless injection rate [Q_in/Q_max] is a function of dimensionless injection pressure [P_i/P_max] given by equation (57). The function (57) is decided by a nozzle shape of a barrel, an entrance shape of the mold, a cavity shape and a polymer characteristics.

$\begin{array}{cc}\left\{\mathrm{Math}.51\right\}& \\ \left[\begin{array}{c}{Q}_{\mathrm{in}}\\ Q_{\max }\end{array}\right]=f\left(\left[\frac{P_i}{P_{\max }}\right]\right)& \left(57\right)\end{array}$

The mathematical model necessary for designing the high-gain observer 31 is given by equations (58), (59) and (60) by using equations (51), (54), (56) and (57).

$\begin{array}{cc}\left\{\mathrm{Math}.52\right\}& \\ \frac{d}{dt}\left[\frac{x}{x_{\max }}\right]=\frac{{\upsilon }_{\max }}{x_{\max }}\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]& \left(58\right)\\ \frac{d}{dt}\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]=\left[\frac{T_{M\max }}{J_{\mathrm{eq}}{\omega }_{\max }}\right]\left[\frac{i_m}{i_{\max }}\right]-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\frac{A_s{P}_{\max }}{J_{\mathrm{eq}}{\omega }_{\max }}\left[\frac{P_i}{P_{\max }}\right]-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\frac{1}{J_{\mathrm{eq}}{\omega }_{\max }}\frac{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}\left\{C_{\mathrm{mt}}{\upsilon }_{\max }^{\gamma }{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]}^{\gamma }+\mu W\right\}& \left(59\right)\\ \frac{d}{dt}\left[\frac{P_i}{P_{\max }}\right]=\frac{\beta }{1-\left[\frac{x}{x_{\max }}\right]}\frac{{\upsilon }_{\max }}{x_{\max }{P}_{\max }}\left\{\left[\frac{{\omega }_m}{{\omega }_{\max }}\right]-f\left(\left[\frac{P_i}{P_{\max }}\right]\right)\right\}& \left(60\right)\end{array}$

When the cavity is filled up with polymer melt, equation (61) is satisfied.

$\begin{array}{cc}\left\{\mathrm{Math}.53\right\}& \\ f\left(\left[\frac{P_i}{P_{\max }}\right]\right)=0& \left(61\right)\end{array}$

The following state variables x₁, x₂ and x₃ defined by equation (62) are introduced.

$\begin{array}{cc}\left\{\mathrm{Math}.54\right\}& \\ x_1=\frac{x}{x_{\max }}{x}_2=\frac{{\omega }_m}{{\omega }_{\max }}{x}_3=\frac{P_i}{P_{\max }}& \left(62\right)\end{array}$

Input variable u defined by equation (63) is introduced. u is measurable. In the design of high-gain observer 31, the actual motor current i_m is considered to be equal to motor current demand i*.

$\begin{array}{cc}\left\{\mathrm{Math}.55\right\}& \\ u=\frac{i_m}{i_{\max }}& \left(63\right)\end{array}$

The state variable x₂ is supposed to be measurable and output variable y is defined by equation (64).{Math. 56}y=x₂  (64)

The state equation and the output equation representing equations (58), (59), (60) and (64) are given by the following equations (65)˜(68).

$\begin{array}{cc}\left\{\mathrm{Math}.57\right\}& \\ {\stackrel{.}{x}}_1={\mathrm{ax}}_2& \left(65\right)\\ {\stackrel{.}{x}}_2={\mathrm{bx}}_3+\chi \left(x_2\right)+\mathrm{cu}& \left(66\right)\\ {\stackrel{.}{x}}_3=\frac{d}{1-x_1}\left(x_2-f\left(x_3\right)\right\}=\left(x\right)& \left(67\right)\\ y=x_2& \left(68\right)\\ x=\left[\begin{array}{c}{x}_1\\ x_2\\ x_3\end{array}\right]\chi \left(x_2\right)=e\frac{x_2}{x_2}\left(h{x_2}^{\gamma }+p\right)& \left(69\right)\\ \begin{array}{cccc}a=\frac{v_{\max }}{x_{\max }{P}_{\max }}& b=-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\frac{A_s{P}_{\max }}{J_{\mathrm{eq}}{\omega }_{\max }}& c=\frac{T_{\mathrm{Mmax}}}{J_{\mathrm{eq}}{\omega }_{\max }}& \\ d=\frac{\beta v_{\max }}{x_{\max }{P}_{\max }}& e=-\frac{l}{2\pi }\frac{1}{\eta }\frac{r_1}{r_2}\frac{1}{J_{\mathrm{eq}}{\omega }_{\max }}& h=C_{\mathrm{mt}}{v}_{\max }^{\gamma }& p=\mu W\end{array}\}& \left(70\right)\end{array}$

where χ(x₂) and ψ(x) are nonlinear functions.

As the output variable y=x₂ represented by equation (68) is measurable, by using the above equation (65) the state variable x₁ is calculated by the following equation (71) and is replaced by a variable y_s. The variable y_s represents a dimensionless screw position.x₁=∫ax₂dt=a∫ydt=y_s  (71)

Therefore, the state variable x₁ is removed from the state variables and a new state equation and a new output equation are represented by the following equations (72) and (73) by using equations (66)˜(68).

$\begin{array}{cc}\left\{\mathrm{Math}.59\right\}& \\ \begin{array}{c}\left[\begin{array}{c}\stackrel{.}{x_2}\\ {\stackrel{.}{x}}_3\end{array}\right]=\left[\begin{array}{cc}0& b\\ 0& 0\end{array}\right]\left[\begin{array}{c}{x}_2\\ x_3\end{array}\right]+\left[\begin{array}{c}\chi \left(x_2\right)+\mathrm{cu}\\ \left(x_2,x_3,y_s\right)\end{array}\right]\\ =\left[\begin{array}{cc}{A}_{11}& A_{12}\\ A_{21}& A_{22}\end{array}\right]\left[\begin{array}{c}{x}_2\\ x_3\end{array}\right]+\left[\begin{array}{c}\chi \left(x_2\right)+\mathrm{cu}\\ \left(x_2,x_3,y_s\right)\end{array}\right]\end{array}& \left(72\right)\\ y=\left[10\right]\left[\begin{array}{c}{x}_2\\ x_3\end{array}\right]& \left(73\right)\\ \chi \left(x_2\right)=e\frac{x_2}{x_2}\left(h{x_2}^{\gamma }+p\right)\left(x_2,x_3,y_s\right)=\frac{d}{1-y_s}\left\{x_2-f\left(x_3\right)\right\}& \left(74\right)\end{array}$

The variable y_s in the state equation (72) is considered to be a new input variable in addition to the input variable u and the input variable y_s is given by the following equation (75) by equation (71).{Math. 60}y_s=a∫ydt  (75)

As state variable x₂ is measurable, it is not necessary to estimate state variable x₂. Therefore, the high-gain observer 31 outputs the estimate of state variable x₃ by using the measurable screw velocity signal y=x₂ and the motor current demand u. The estimate {circumflex over (x)}₃ is given by the following equation (77) (non patent literature NPL 2). Input variable y, in equation (77) is calculated in the high-gain observer 31 by using a time integration method of equation (75) applied for the measurable screw velocity y=x₂. K is a gain constant of the high-gain observer 31.{Math. 61}{circumflex over ({dot over (x)}=(A₂₂+KA₁₂){circumflex over (x)}₃−K{{dot over (y)}−A₁₁y−χ₀(y)−cu}+A₂₁y+ψ₀(ŷ,y,y_s)  (76){circumflex over ({dot over (x)}=Kb{circumflex over (x)}₃−K{{dot over (y)}−χ₀(y)−cu}+ψ₀({circumflex over (x)}₃,y,y_s)  (77)

where χ₀(y)ψ₀({circumflex over (x)}₃, y, y_s): Nominal functions of χ(y), ψ({circumflex over (x)}₃, y, y_s), respectively, used in the high-gain observer 31. Equation (77) is rewritten by equation (78).{Math. 62}{circumflex over ({dot over (x)}₃+K{dot over (y)}=Kb{circumflex over (x)}₃+K{χ₀(y)+cu}+ψ₀({circumflex over (x)}₃,y,y_s)  *78)

A new variable ŵ is introduced by the following equation (79).{Math. 63}ŵ={circumflex over (x)}₃+K_y  (79)

The estimate {circumflex over (x)}₃ is given by equations (80) and (81) by using equations (78) and (79).{Math. 64}{circumflex over ({dot over (w)}=Kb({circumflex over (ω)}−Ky)+K{χ₀(y)+cu}+ψ₀({circumflex over (ω)}y,y_s)  (80){circumflex over (x)}₃={circumflex over (ω)}−Ky  (81)

A positive parameter ε much smaller than 1 is introduced and the gain constant K is given by equation (82) and a new variable {circumflex over (η)} is introduced by equation (83).

$\begin{array}{cc}\left\{\mathrm{Math}.65\right\}& \\ K=\frac{K_1}{\varepsilon }& \left(82\right)\\ \hat{\eta }=\varepsilon \hat{w}& \left(83\right)\end{array}$

Equation (80) is rewritten as the following equation (84) by using equations (82) and (83).

$\begin{array}{cc}\left\{\mathrm{Math}.66\right\}& \\ \stackrel{.}{\hat{\eta }}=\frac{K_1}{\varepsilon }b\left(\hat{\eta }-K_{1}y\right)+K_1{\chi }_0\left(y\right)+K_1\mathrm{cu}+\varepsilon 0_{}\left(\hat{\eta },y,y_s\right)& \left(84\right)\end{array}$

The following equation (85) is given from equation (83).

$\begin{array}{cc}\left\{\mathrm{Math}.67\right\}& \\ \hat{w}=\frac{1}{\varepsilon }\hat{\eta }& \left(85\right)\end{array}$

By using equation (85), equation (81) is rewritten as the following equation (86).

$\begin{array}{cc}\left\{\mathrm{Math}.68\right\}& \\ {\hat{x}}_3=\frac{1}{\varepsilon }\hat{\eta }-\mathrm{Ky}& \left(86\right)\end{array}$

Thus the estimate of state variable {circumflex over (x)}₃ is obtained by the high-gain observer 31. From equations (75), (84) and (86), the calculation procedures are given by equations (87), (88) and (89).

(1) Calculation procedure 1{Math. 69}y_s=a∫ydt  (87)

(2) Calculation procedure 2

$\begin{array}{cc}\left\{\mathrm{Math}.70\right\}& \\ \stackrel{.}{\hat{\eta }}=\frac{K_1}{\varepsilon }b\left(\hat{\eta }-K_{1}y\right)+K_1{\chi }_0\left(y\right)+K_1\mathrm{cu}+\varepsilon 0_{}\left(\hat{\eta },y,y_s\right)& \left(88\right)\end{array}$

(3) Calculation procedure 3

$\begin{array}{cc}\left\{\mathrm{Math}.71\right\}& \\ {\hat{x}}_3=\frac{1}{\varepsilon }\left(\hat{\eta }-K_{1}y\right)& \left(89\right)\end{array}$

By the calculation procedure 1, y_s is calculated, by the calculation procedure 2, the estimate {circumflex over (η)} is obtained and by the calculation procedure 3 the estimate {circumflex over (x)}₃ is obtained.

Then it is shown that the high-gain observer 31 as the pressure detecting means satisfies the following two requirements (A) and (B) described in Solution to Problem (paragraph {0057}).

(A) The detection means is high-precision.

(B) The detection means has very small time-lag.

If the nominal functions χ₀(y) and ψ₀({circumflex over (η)}, y, y_s) in equation (88) are replaced with the true but actually unobtainable functions χ(y) and ψ(η, y, y_s), the true value η of the estimate {circumflex over (η)} may be obtained by equation (90).

$\begin{array}{cc}\left\{\mathrm{Math}.72\right\}& \\ \stackrel{.}{\eta }=\frac{K_1}{\varepsilon }b\left(\eta -K_{1}y\right)+K_1\chi \left(y\right)+K_1\mathrm{cu}+\varepsilon \left(\eta ,y,y_s\right)& \left(90\right)\end{array}$

Then the estimate error {tilde over (η)}=η−{circumflex over (η)} is obtained by equation (91) by using equations (88) and (90).

$\begin{array}{cc}\left\{\mathrm{Math}.73\right\}& \\ \varepsilon \stackrel{.}{\hat{\eta }}=K_{1}b\tilde{\eta }+\varepsilon K_1{\delta }_1\left(y\right)+{\varepsilon }^2{\delta }_2\left(\tilde{\eta },y,y_s\right)& \left(91\right)\\ \begin{array}{c}{\delta }_1\left(y\right)=\chi \left(y\right)-{\chi }_0\left(y\right)\\ {\delta }_2\left(\tilde{\eta },y,y_s\right)=\left(\eta ,y,y_s\right)-0_{}\left(\hat{\eta },y,\mathrm{ys}\right)\end{array}\}& \left(92\right)\end{array}$

As ε is much smaller than 1, the effects of model errors δ₁ and δ₂ on the estimation error {tilde over (η)} can be made small enough by equation (91). In other words, the high-gain observer 31 satisfies the above requirement (A) “High-precision detection” (paragraph {0057}) for injection pressure estimate {circumflex over (x)}₃ obtained by equations (87), (88) and (89).

When the effects of model errors δ₁ and δ₂ on the estimation error 1) are neglected in equation (91), equation (91) is rewritten as equation (93).{Math. 74}ε{dot over ({tilde over (η)}=K₁b{tilde over (η)}  (93)

The estimate error {tilde over (η)} is given by the following equation (94) from equation (93).

$\begin{array}{cc}\tilde{\eta }\left(t\right)=\exp \left(\frac{K_{1}b}{\varepsilon }t\right){\tilde{\eta }}_0\left(K_1>0b<0\right)& \left(94\right)\end{array}$

where t: Time variable, {tilde over (η)}₀: Initial value of estimate error {tilde over (η)}. As b<0 in the injection process of injection molding machines and ε is much smaller than 1, equation (94) reveals that the time response {tilde over (η)}(t) of the estimate error tends to zero rapidly. In other words, the high-gain observer 31 satisfies the above requirement (B) “Detection with small time-lag” (paragraph {0057}) for injection pressure estimate {circumflex over (x)}₃ obtained by equations (87), (88) and (89).

As the injection controller 20 executes a control algorithm at a constant time interval Δt, the arithmetic expressions (87), (88) and (89) of the high-gain observer 31 are transformed into the discrete-time arithmetic expressions (non patent literature NPL 3, NPL 4).

A new parameter a is introduced and the time interval Δt is expressed by equation (95).{Math. 76}Δt=αε  (95)

A discrete-time expression of a time integration equation (87) can be found by using the standard method of trapezoid rule and is given by the following equation (96).{Math. 77}y_s(t_k+1)=y_s(t_k)+0.5aαε{y(t_k)+y(t_k+1)}  (96)

When function values y_s(t_k), y(t_k) at a discrete-time t_k (k=0, 1, 2, . . . ) are represented by y_s(k), y(k), equation (96) is given by the following equation (97).{Math. 78}y_s(k+1)=y_s(k)+0.5aαε{y(k)+y(k+1)}  (97)

As the time interval Δt is small, the numerical time integration equation (97) of the output variable y(t) is considered to be high-precision.

Next, a discrete-time equivalent of the continuous-time equation (88) can be found by using the standard method of forward rectangular rule which gives the relation between the Laplace-transform operator s representing time-derivative operation and z-transform operator z as follows.

$\begin{array}{cc}\left\{\mathrm{Math}.79\right\}& \\ s=\frac{z-1}{\Delta t}=\frac{z-1}{\mathrm{\alpha \varepsilon }}& \left(98\right)\end{array}$

By using equation (98), equation (88) is rewritten as the following equation (99).

$\begin{array}{cc}\left\{\mathrm{Math}.80\right\}& \\ \frac{z-1}{\mathrm{\alpha \varepsilon }}\hat{\eta }=\frac{K_1}{\varepsilon }b\left(\hat{\eta }-K_{1}y\right)+K_1{\chi }_0\left(y\right)+K_1\mathrm{cu}+\varepsilon 0_{}\left(\hat{\eta },y,y_s\right)& \left(99\right)\end{array}$

The discrete-time expression of equation (99) similar to equation (97) is given by the following equation (100).

$\begin{array}{cc}\left\{\mathrm{Math}.81\right\}& \\ \hat{\eta }\left(k+1\right)-\hat{\eta }\left(k\right)=\alpha K_{1}b\left(\hat{\eta }\left(k\right)-K_{1}y\left(k\right)\right)+\mathrm{\alpha \varepsilon }{K}_1{\chi }_0\left(k\right)+\mathrm{\alpha \varepsilon }{K}_1\mathrm{cu}\left(k\right)+{\mathrm{\alpha \varepsilon }}^2{0}_{}\left(k\right)& \left(100\right)\\ {\chi }_0\left(k\right)=e\frac{y\left(k\right)}{y\left(k\right)}\left(h{y\left(k\right)}^{\gamma }+p\right)0_{}\left(k\right)=\frac{d}{1-y_s\left(k\right)}\left\{y\left(k\right)-f\left(\hat{\eta }\left(k\right),y\left(k\right)\right)\right\}& \left(101\right)\end{array}$

where {tilde over (η)}(k): Estimate {tilde over (η)}(t_k) at a discrete-time t_k, y(k), u(k), y_s(k): y(t_k), u(t_k), y_s(t_k) at a discrete-time t_k, χ₀(k), ψ₀(k): χ₀ (t_k), ψ_o(t_k) at a discrete-time t_k. χ₀ (k) is given by equation (69) and ψ₀(k) is given by equation (67). Equation (100) is rewritten as the following equation (102).{Math. 82}{tilde over (η)}(k+1)=(1+αK₁b){tilde over (η)}(k)−αK₁²by(k)+αεK₁χ₀(k)+αεK₁cu(k)+αε²ψ₀(k)  (102)

The discrete-time equivalent of equation (89) is given by equation (103).

$\begin{array}{cc}\left\{\mathrm{Math}.83\right\}& \\ {\hat{x}}_3\left(k\right)=\frac{1}{\varepsilon }\left\{\hat{\eta }\left(k\right)-K_{1}y\left(k\right)\right\}& \left(103\right)\end{array}$

The high-gain observer 31 obtains injection pressure estimate {circumflex over (x)}₃ (k) at a discrete-time t_k by executing the arithmetic expressions of equations (97), (102) and (103) at a constant time interval Δt. The high-gain observer 31 by equations (97), (102) and (103) does not estimate the measurable state variable x₂(k) (screw velocity) and estimates only necessary state variables x₃ (k) and so is called by a reduced-order high-gain observer.

FIG. 6 shows the results of computer simulation when the high-gain observer 31 is used for the pressure control of an electric-motor driven injection molding machine.

The constants of the mathematical model are as follows.

Maximum screw stroke x_max=37.2 cm

Maximum injection velocity v_max=13.2 cm/sec

Maximum injection pressure P_max=17652 N/cm²

Motor rating speed ω_max=209.4 rad/sec (2000 rpm)

The constants a, b, c and d in equations (65)˜(67) are expressed in equation (104). In this calculation it is assumed that resistance component χ(x₂)=0.

$\begin{array}{cc}\left\{\mathrm{Math}.84\right\}& \\ \begin{array}{c}a=0.3556{\sec }^{-1}\\ b=-6.474{\sec }^{-1}\\ c=3.716{\sec }^{-1}\\ d=3.626{\sec }^{-1}\end{array}\}& \left(104\right)\end{array}$

In order to realize an arbitrary injection pressure time response according to the setting value P_set(t), a hydraulic characteristics of solenoid operated proportional relief valve is used for the characteristics of function ƒ({circumflex over (P)}_i/P_max), which decides the polymer flow into the cavity according to the value of {circumflex over (P)}_ii(t)/P_set(t). The gain constant K of equation (82) used by the high-gain observer 31 is given by equation (105). The data K₁=0.003, ε=0.01 and Δt=5 msec are used.{Math. 85}K=0.3  (105)

FIG. 6 shows the control performance of injection pressure [P_i/P_max]. FIG. 6 (a) shows the time response of injection pressure [P_i(t)/P_max] when the hitherto known control system in FIG. 4 is used and the pressure detector 10 is used. When the high-gain observer 31 is supposed to be used under the control system shown in FIG. 4 and to calculate the estimate of injection pressure [{circumflex over (P)}_i/P_max] by using screw velocity signal y(t) and motor current signal u(t), FIG. 6 (b) shows the time response of the estimated injection pressure [P_i(t)/P_max]. FIG. 6 (c) shows the time response of the actual injection pressure [P_i(t)/P_max] when the control system shown in FIG. 1 is used and the estimated injection pressure [{circumflex over (P)}_i(t)/P_max] obtained by the high-gain observer 31 is fed to the subtractor 27 as a feedback signal of injection pressure, in other words, the pressure detector 10 is not used. The time response of estimated injection pressure [{circumflex over (P)}_i(t)/P_max] shown in FIG. 6 (b) agrees well with that of actual injection pressure [P_i(t)/P_max] shown in FIG. 6 (a). The time response of actual injection pressure [P_i(t)/P_max] shown in FIG. 6 (a) agrees well with that of actual injection pressure shown in FIG. 6 (c) when the high-gain observer 31 is used. The transfer from the injection process to the pressure application process is conducted at the screw position [x/x_max]=0.55 and the time t₁ shown in FIG. 3 is 2.6 seconds. Thus the high-gain observer 31 can estimate the injection pressure exactly with small time-lag. The estimate of injection pressure obtained by the high-gain observer 31 can be used to monitor the injection pressure in the injection process and can be used as a feedback signal of injection pressure in the pressure application process.

INDUSTRIAL APPLICABILITY

In the pressure control apparatus and pressure control method of electric-motor driven injection molding machines, the following four disadvantages can be avoided by using the estimated injection pressure obtained by the high-gain observer as a feedback signal of injection pressure in place of a pressure detector.

• (1) A highly reliable pressure detector is very expensive under high pressure circumstances.
• (2) Mounting a pressure detector in the cavity or the barrel nozzle part necessitates the troublesome works and the working cost becomes considerable.
• (3) Mounting a load cell in the injection shafting alignment from a servomotor to a screw complicates the mechanical structure and degrades the mechanical stiffness of the structure.
• (4) A load cell which uses strain gauges as a detection device necessitates an electric protection against noise for weak analog signals. Moreover, the works for zero-point adjusting and span adjusting of a signal amplifier are necessary (patent literature PTL 13).

As the high-gain observer can estimate the injection pressure exactly with small time-lag, the estimate of injection pressure obtained by the high-gain observer can be used to monitor the injection pressure in the injection process and can be used as a feedback signal of injection pressure in the pressure application process. Thus the high-gain observer of the present invention can be applied to the pressure control apparatus and pressure control method of electric-motor driven injection molding machines.

REFERENCE SIGNS LIST

• 1 Metal mold
• 2 Barrel
• 3 Servomotor
• 4 Reduction gear
• 5 Ball screw
• 6 Bearing
• 7 Nut
• 8 Moving part
• 9 Screw
• 10 Pressure detector
• 11 Linear slider
• 12 Rotary encoder
• 13 Cavity
• 20 Injection controller
• 21 Injection velocity setting device
• 23 Pulse generator
• 24 Pulse train
• 25 Analog/digital (A/D) converter
• 26 Injection pressure setting device
• 27 Subtractor
• 28 Pressure controller
• 29 Digital/analog (D/A) converter
• 30 Analog/digital (A/D) converter
• 31 High-gain observer
• 40 Motor controller (Servoamplifier)
• 41 Pulse counter
• 42 Analog/digital (A/D) converter
• 43 Comparator
• 44 Pulse counter
• 45 Subtractor
• 46 Position controller
• 47 Differentiator
• 48 Subtractor
• 49 Velocity controller
• 50 Pulse width modulation (PWM) device

Claims

1. A method for controlling pressure in an electric-motor driven injection molding machine having an injection and pressure application mechanism in which rotation of a servomotor is transferred to rotation of a ball screw through a reduction gear and rotation of said ball screw is converted to a linear motion of a nut of said ball screw and said nut drives a moving part and a linear motion of a screw is realized through said moving part and pressure application to a melted polymer stored at the end of a barrel and fill-up of a cavity with polymer melt are realized by a movement of said screw, comprising:deriving an estimate of injection pressure {circumflex over (x)}3 which a high-gain observer outputs by using an injection velocity signal detected by a rotary encoder mounted on a servomotor axis and a differentiator and a motor current demand signal applied to said servomotor or an actual motor current signal as inputs and by executing at a constant time interval a discrete-time expression of time integration giving a screw position signal wherein the equation (110) is used, which is obtained by applying a standard method of trapezoid rule to a continuous-time time integration equation wherein the equation (109) is used and by executing at a constant time interval discrete-time arithmetic expressions wherein the equation (112) and the equation (114) are used, which are obtained by applying a standard method of forward rectangular rule to a continuous-time calculation procedures which are derived from a mathematical model of an injection mechanism representing motion equations of said injection and pressure application mechanism and consisting of a state equation having two state variables of an injection velocity variable and an injection pressure variable and having two input variables of a motor current demand signal applied to said servomotor or an actual motor current signal and a screw position signal wherein the equation (106) is used and an output equation having one output variable of said injection velocity signal wherein the equation (107) is used; ⁢[x.2x.3]=[0b00]⁡[x2x3]+[χ⁡(x2)+cu⁢(x2,x3,ys)](106)y=[10]⁡[x2x3](107)χ⁡(x2)=e⁢x2x2⁢(h⁢x2γ+p)⁢⁢⁢(x2,x3,ys)=d1-ys⁢{x2-f⁡(x3)}(108)ys=a∫ydt  (109)where x2: Dimensionless state variable of injection velocity which is defined to be an injection velocity variable divided by maximum injection velocity, x3: Dimensionless state variable of injection pressure which is defined to be an injection pressure variable divided by maximum injection pressure, u: Dimensionless input variable of motor current demand or actual motor current variable divided by motor current rating, y: Dimensionless output variable expressing a measurable state variable x2, ys: Dimensionless input variable expressing a screw position variable divided by maximum screw position, a, b, c, d, e, h, p, γ: Constants of a mathematical model of an injection mechanism, ƒ(x3): Function of a state variable x3 determining dimensionless injection rate of polymer into a cavity which is defined to be an injection rate divided by maximum injection rate, χ(x2), ψ(x2, x3, ys): Nonlinear functions of the equation (108), t: Time variable ys(k+1)=ys(k)+0.5aαε{y(k)+y(k+1)}  (110)αε=Δt  (111)where k: Discrete variable representing a discrete-time tk (k=0, 1, 2, . . . ), ys(k): Value of input variable ys(tk) at a discrete-time tk, y(k): Value of output variable y(tk) at a discrete-time tk, Δt: Sampling period of a discrete-time high-gain observer, ε: Positive parameter much smaller than 1 used in the high-gain observer