ROBOT JOINT TORQUE CONTROL SYSTEM AND LOAD COMPENSATION METHOD THEREFOR

Abstract

A robot joint torque control system and a load compensation method therefor are provided, which relate to the technical field of robot joint motion control. A mathematical model of the robot joint torque control system is established first. Equivalent transformation is performed on a system functional block diagram thereof, and then it can be seen that load parameters have a great influence on joint torque output. A load compensation controller is designed to effectively eliminate the influence of the load parameters on an output torque of the joint. The system is equivalent to an inertial element on the basis of the compensation, and then a PD controller parameter is adjusted to increase an open-loop gain of the system, so as to increase a system bandwidth and increase a response speed of the joint torque control system, thereby improving performance of the joint torque control system.

Description

TECHNICAL FIELD

The present disclosure belongs to the technical field of robot joint motion control, and specifically, relates to a robot joint torque control system and a load compensation method therefor.

BACKGROUND

In recent years, robots have been more and more widely used in all walks of life, so that the research of robots has become a hot spot. Common drive modes for robots include motor drive, hydraulic drive, and pneumatic drive. Motor drive has the advantages of compact structure, simple control, high transmission efficiency, and high control accuracy, and is the most common drive mode in the field of robots at present. A joint of a robot is usually formed by a servo motor, a gear reducer, an encoder, and so on, and uses a typical position control system. However, a large rigid impact is caused if the position control system is still used when a foot of the robot touches the ground or an obstacle, which causes instability in operations such as walking and running of the robot. At this time, the joint must switch to a torque control system, so as to effectively mitigate the impact and improve the overall stability of the robot.

Currently, the following several torque control solutions exist: {circle around (1)} a series elastic actuator (SEA): elastic components are connected in series between a drive end and a joint load end, having significant advantages such as low output impedance and impact resistance, but having a low output bandwidth and a small output torque. {circle around (2)} MIT Cheetah uses a quasi direct drive, uses a solution of high-torque motor+small-reduction-ratio reducer, and uses internal current loop feedback to realize low-cost force control, and is currently only used in small quadruped robots. {circle around (3)} A joint drive unit uses a control solution of high-torque servo motor+large-reduction-ratio reducer+torque sensor, and realizes large torque output, thereby satisfying high dynamic tasks such as fast walking, running, and jumping of heavy robots.

In order to realize high dynamic tasks of robots, the control solution of high-torque servo motor+large-reduction-ratio reducer+torque sensor must be used. Currently, robot joint torque control systems often use control methods such as PD control, quantitative feedback control, and robust control. PD control is currently one of the most common control methods in engineering projects, in which adjustment is made based on an error input to a control system, without the need to model a controlled object, and meanwhile, simple debugging and easy implementation are realized. However, when PD control is used alone, in order to achieve better control performance, control parameters are often set to be excessively large, which easily causes system instability. Such a method still needs to be used together with other compensation control methods, so as to achieve better control performance. Moreover, although various advanced control methods such as quantitative feedback control and robust control can achieve a desirable control effect, the control methods have complex models, and the research work thereof is mostly simulation analysis or confined to laboratories.

SUMMARY

In view of the deficiencies in the prior art, the present disclosure provides a robot joint torque control system and a load compensation method therefor, which eliminate the influence of load parameters on actual torque output of the system and improve performance of the joint torque control system.

The present disclosure achieves the aforementioned technical objective by the following technical means.

A load compensation method for a robot joint torque control system includes accumulating a current i_f whose joint-end velocity is compensated for by a load compensation controller and an original current loop input instruction i_r, and then using the accumulated current as a current loop input instruction of a joint torque control system to effectively compensate for an influence of load parameters on an output torque of the joint torque control system.

Further, the load compensation controller is

$G_f\left(s\right)=\frac{n{K}_v}{K_a},$

wherein n is a reduction ratio of a gear reducer, K_v is a velocity feedback coefficient of a servo motor, and K_a is a current loop control parameter.

Furthermore, an open-loop transfer function of the joint torque control system is:

$G\left(s\right)=\frac{T_p}{E_T}=\frac{{\mathrm{nK}}_t{K}_a{K}_{\mathrm{PD}}}{\mathrm{Ls}+R+K_a{K}_t}=\frac{\frac{{\mathrm{nK}}_t{K}_a{K}_{\mathrm{PD}}}{R+K_a{K}_t}}{\frac{s}{\omega }+1}=\frac{K_o}{\frac{s}{\omega }+1}$

wherein a corner frequency of an inertial element:

$\omega =\frac{R+K_a{K}_t}{L},$

an open-loop gain

$K_o=\frac{K_{\mathrm{PD}}{\mathrm{nK}}_t{K}_a}{R+K_a{K}_t},$

K_PD is a PD controller parameter, K_t is a torque coefficient of the servo motor, L is an inductance of the servo motor, and R is a resistance of the servo motor.

Furthermore, when the PD controller parameter K_PD is increased, the open-loop gain K_o of the system increases accordingly, and a cutoff frequency when the joint torque control system passes through a 0 db line increases; that is, a system bandwidth increases, thereby improving rapidity of the system.

A joint torque control system includes a joint body and a control system, wherein the joint body includes a servo motor, a gear reducer, a torque sensor, and an absolute encoder, an output shaft of the servo motor is rigidly and fixedly connected to the gear reducer, the gear reducer is rigidly connected to a joint-end load by means of the torque sensor, and the absolute encoder is mounted at the joint-end load; the control system includes a first comparator, a torque loop controller, a second comparator, and a current loop controller that are sequentially in signal connection.

In the aforementioned technical solution, the torque sensor is used for detecting an actual torque of a joint end; a current sensor is mounted inside the servo motor and used for detecting an actual output current of the servo motor; and the absolute encoder is used for detecting an actual position of the joint-end load.

In the aforementioned technical solution, in the control system, a given current and the actual output current of the servo motor are compared and subtracted, and pass through the second comparator to control current output of the servo motor and form an internal current loop; a given torque and the actual torque of the joint end are compared and subtracted, and then pass through the first comparator and the internal current loop to control joint torque output and form an external torque loop.

The beneficial effects of the present disclosure are: in the present disclosure, first, mathematical modeling is performed on a robot joint torque control system, and equivalent transformation is performed on a system control functional block diagram thereof to obtain by analysis that load parameters have a certain influence on actual torque output; then, a load compensation controller is designed to eliminate the influence of the load parameters on the actual torque output of the system; the system is equivalent to an inertial element on the basis of the compensation, and a PD controller parameter is adjusted to increase an open-loop gain of the system, so as to increase a system bandwidth, thereby improving performance of the joint torque control system. The compensation method of the present disclosure does not need to build a complex mathematical model, is easy to apply and is especially suitable for use in engineering projects.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a diagram of a robot joint torque control system in the present disclosure.

FIG. 2 is a functional block diagram of the robot joint torque control system in the present disclosure.

FIG. 3 is a block diagram of the joint torque control system in the present disclosure after primary transformation.

FIG. 4 is a block diagram of the joint torque control system in the present disclosure after secondary transformation.

FIG. 5 is a block diagram of the joint torque control system in the present disclosure with a load compensation controller added thereto.

FIG. 6 is a functional block diagram of the joint torque control system in the present disclosure after equivalent transformation of FIG. 5.

FIG. 7 is an open-loop Bode diagram of the joint torque control system in the present disclosure.

DETAILED DESCRIPTION OF THE EMBODIMENTS

The present disclosure is further illustrated below with reference to the accompanying drawings and specific embodiments, but the protection scope of the present disclosure is not limited thereto.

As shown in FIG. 1, a robot joint torque control system is formed by a joint body and a control system. The joint body is formed by a servo motor, a gear reducer, a torque sensor, and an absolute encoder, an output shaft of the servo motor is rigidly and fixedly connected to the gear reducer, the gear reducer is rigidly connected to a joint-end load by means of the torque sensor, and the torque sensor is used for detecting an actual torque of a joint end; a current sensor is mounted inside the servo motor and used for detecting an actual output current of the servo motor; and the absolute encoder is mounted at the joint-end load and used for detecting an actual position of the joint-end load. The control system is formed by a first comparator, a torque loop controller, a second comparator, and a loop current controller. A given current and the actual output current of the servo motor are compared and subtracted, and then pass through the second comparator to control current output of the servo motor and form an internal current loop; a given torque and the actual torque of the joint end are compared and subtracted, and then pass through the first comparator and the internal current loop to control joint torque output and form an external torque loop.

According to a functional diagram of the robot joint torque control system, mathematical modeling on the robot torque control system is as follows:

The servo motor uses a current closed loop, and has an output voltage U_c being:

U _c =K _a(i_r−K_Ii_a)  (1)

where K_a is a current loop control parameter, i_r is a current loop input instruction, K_I is a circuit loop feedback coefficient of the servo motor, and i_a is an actual current value; the current loop input instruction i_r is obtained by multiplying a joint torque error E_T and a PD controller parameter K_PD, and the joint torque error E_T is obtained by subtracting a given torque T_r of the joint and an actual torque T_p of the joint end.

A voltage balance equation of the servo motor is:

$\begin{array}{cc}{U}_c=E+L\frac{{\mathrm{di}}_a}{\mathrm{dt}}+{\mathrm{Ri}}_a& \left(2\right)\end{array}$

where E is a counter electromotive force of the servo motor, and

$E={\mathrm{nK}}_v\frac{d\theta }{\mathrm{dt}},$

n is a reduction ratio of the gear reducer, K_v is a velocity feedback coefficient of the servo motor, θ is an actual position of the joint end,

$\frac{d\theta }{\mathrm{dt}}$

is an actual velocity of the joint end, L is an inductance of the servo motor, and R is a resistance of the servo motor.

An output torque of the servo motor is:

T _m =i _a K _t  (3)

where K_t is a torque coefficient of the servo motor.

Since the torque sensor has large rigidity, the elastic deformation thereof is neglected, and then a relationship between outputs of the motor output end and the actual torque of the joint end is:

T _p =nT _m  (4)

where T_m is the output torque of the servo motor, and T_p is the actual torque of the joint end.

A torque balance equation of the joint end is:

$\begin{array}{cc}{T}_p=J_m\frac{d{\theta }^2}{\mathrm{dt}}+B_m\frac{d\theta }{\mathrm{dt}}+G& \left(5\right)\end{array}$

where J_m is a total moment of inertia converted to the joint-end load, and is a sum of converted moments of inertia of the load, the torque sensor, the gear reducer, the servo motor, and other components, B_m is a total viscous damping coefficient converted to the joint-end load, and is a sum of converted viscous damping coefficients of the load, the gear reducer, and the servo motor, and G is a gravity term of the joint end; regarding the gravity term G of the joint end, the influence of the gravity term on the system output is eliminated by a method such as gravity compensation (in the prior art), and the influence of the gravity term is not considered in the present disclosure.

Equations (1) to (5) are combined and subjected to Laplace transformation to establish a functional block diagram of the robot joint torque control system, as shown in FIG. 2.

Equivalent transformation is performed on the block diagram of FIG. 2, a node A is moved backward (across K_a) to a node B to obtain a block diagram of the joint torque control system after primary transformation, as shown in FIG. 3.

Since the closed loop 1 in FIG. 3 is a typical closed-loop feedback loop, arrangement is made according to simplified rules for closed-loop feedback to obtain a block diagram of the joint torque control system after secondary transformation, as shown in FIG. 4.

It can be seen from FIG. 4 that load parameters J_m and B_m have a certain influence on actual torque output of the system through a loop A (the load parameters reduce the actual torque output of the system), and the larger the load parameters, the larger the disturbing influence on the actual output torque of the system.

In order to eliminate the influence of the load parameters on the joint torque control system, a load compensation controller is designed, a current i_f whose signal value of a joint-end velocity (which may be obtained after differential processing is performed on an actual position signal θ of the joint end) is compensated for by the load compensation controller is selected, and i_f and i_r are added and then used as a current loop input instruction. The principle is as follows: at a node C, a value of {dot over (θ)} after passing through a loop C and a value of {dot over (θ)} after passing through a loop B are equal in size and opposite in sign, so that they cancel each other out, thereby eliminating the influence of the load parameters on the force control system, as shown in FIG. 5.

At the node C:

{dot over (θ)}nK_v={dot over (θ)}_fG_f(s)K_a  (6)

Thus, the load compensation controller is:

$\begin{array}{cc}{G}_f\left(s\right)=\frac{{\mathrm{nK}}_v}{K_a}& \left(7\right)\end{array}$

The influence of the load parameters on the output torque is effectively eliminated by the load compensation controller, and equivalent transformation is performed on the whole joint torque control system to obtain a transformed joint torque control functional block diagram, as shown in FIG. 6.

An open-loop transfer function of the robot joint torque control system is:

$\begin{array}{cc}G\left(s\right)=\frac{T_p}{E_T}=\frac{{\mathrm{nK}}_t{K}_a{K}_{\mathrm{PD}}}{\mathrm{Ls}+R+K_a{K}_t}=\frac{\frac{{\mathrm{nK}}_t{K}_a{K}_{\mathrm{PD}}}{R+K_a{K}_t}}{\frac{s}{\omega }+1}=\frac{K_a}{\frac{s}{\omega }+1}& \left(8\right)\end{array}$

in the equation, a corner frequency of an inertial element:

$\omega =\frac{R+K_a{K}_t}{L},$

and an open-loop gain

$K_o=\frac{K_{\mathrm{PD}}{\mathrm{nK}}_t{K}_a}{R+K_a{K}_t}.$

The system is a typical inertial element, and when the PD controller parameter K_PD is adjusted, the open-loop gain of the system changes accordingly, and a Bode diagram thereof also changes. As shown in FIG. 7, when the open-loop gain K_o increased to K_o′; that is, K_o<K_o′, an open-loop Bode diagram of the joint torque control system moves up as a whole, so that a cutoff frequency when the system passes through a 0 db line increases; that is, ω₁<ω₂, where ω₁ is the cutoff frequency of the system when the open-loop gain is Ko, and ω₂ is the cutoff frequency of the system when the open-loop gain is Ko′; since an open-loop cutoff frequency of a Bode diagram of the open-loop transfer function is a system bandwidth, the system bandwidth increases, thereby enhancing rapidity of the system and improving control performance of the system.

The described embodiments are preferred embodiments of the present disclosure, but the present disclosure is not limited to the aforementioned embodiments. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essential content of the present disclosure shall fall within the protection scope of the present disclosure.

Claims

1. A load compensation method for a robot joint torque control system, comprising accumulating a current if whose joint-end velocity is compensated for by a load compensation controller and an original current loop input instruction ir, and then using the accumulated current as a current loop input instruction of a joint torque control system to effectively compensate for an influence of load parameters on an output torque of the joint torque control system; wherein the load compensation controller is

2. (canceled)

3. (canceled)

4. The load compensation method for the robot joint torque control system according to claim 1, wherein when the PD controller parameter KPD is increased, the open-loop gain Ko of the system increases accordingly, and a cutoff frequency when the joint torque control system passes through a 0 db line increases; that is, a system bandwidth increases, thereby improving rapidity of the system.

5. A joint torque control system according to claim 1, comprising a joint body and a control system, wherein the joint body comprises a servo motor, a gear reducer, a torque sensor, and an absolute encoder, wherein an output shaft of the servo motor is rigidly and fixedly connected to the gear reducer, the gear reducer is rigidly connected to a joint-end load by means of the torque sensor, and the absolute encoder is mounted at the joint-end load; the control system comprises a first comparator, a torque loop controller, a second comparator, and a current loop controller, wherein the first comparator, the torque loop controller, the second comparator, and the current loop controller are sequentially in signal connection.

6. The joint torque control system according to claim 5, wherein the torque sensor is used for detecting an actual torque of a joint end; a current sensor is mounted inside the servo motor and used for detecting an actual output current of the servo motor; and the absolute encoder is used for detecting an actual position of the joint-end load.

7. The joint torque control system according to claim 6, wherein in the control system, a given current and the actual output current of the servo motor are compared and subtracted, and pass through the second comparator to control current output of the servo motor and form an internal current loop; a given torque and the actual torque of the joint end are compared and subtracted, and then pass through the first comparator and the internal current loop to control joint torque output and form an external torque loop.

8. (canceled)

9. (canceled)

10. The joint torque control system according to claim 5, wherein when the PD controller parameter KPD is increased, the open-loop gain Ko of the system increases accordingly, and a cutoff frequency when the joint torque control system passes through a 0 db line increases; that is, a system bandwidth increases, thereby improving rapidity of the system.