Like reference numbers and designations in the various drawings indicate like elements.
Throughout this description, embodiments and variations are described for the purpose of illustrating uses and implementations of the inventive concepts. The illustrative description should be understood as presenting examples of the inventive concept, rather than as limiting the scope of the concept as disclosed herein.
A block diagram of a PWM valve model 200 is shown in
When the input current I is provided to a coil of a PWM valve (e.g., the coil 106 of the prior art PWM valve 100,
An armature force produced by the valve dynamics element 210 is operatively transmitted to a fluid control element 216 (e.g., the ball 104,
A feedback line 222 conveys a hydraulic force 208 to the combined force element 204. The magnitude of the hydraulic force 208 is proportional to the product of the control pressure Pc and an effective area of the fluid control element 216. In one example, the effective area of the fluid control element 216 may be an area of the ball 104 (
From the complex behavior of a PWM valve in response to a drive current I described above, persons skilled in the art of hydraulic design will understand that, for purposes of designing a feedback loop, the PWM valve cannot be modeled as a DC gain or linear transfer function. Due to the above-described complex behavior, traditional proportional control methods cannot be readily implemented for PWM valve-based feedback control systems without adverse outcomes such as control instability and degraded response speed. The present teachings overcome these limitations of the prior art feedback control methods for PWM valves.
As previously noted, although a normally closed PWM valve has been used in the description set forth above for exemplary purposes, the present methods and apparatus apply equally to normally open PWM valves.
In some embodiments, the controller 304 may be implemented using a digital processing unit, such as an embedded digital controller, microprocessor, microcontroller or other computer processing unit (CPU). In other embodiments analog signal processing electronics may be employed. In
The PWM valve 308 responds to the drive current I received from the PWM driver 306. The PWM valve 308 receives an input fluid from the fluid supply 310, generates a return fluid to the fluid return element 312, and provides the control fluid to the load 314 at a specified control pressure Pc, responsive to the drive current I. As noted above, the PWM valve 308 operates in a binary mode, and the control pressure Pc is switched between high and low values corresponding to Ps and Pr, respectively, at a rate and duty cycle responsive to the drive current I and the duty cycle control signal output by the controller 304. In some embodiments, the load 314 responds only to an average value of the control pressure Pc. For example, if the pulse rate is 100 Hz, and the load 314 comprises a clutch mechanism in an automatic transmission, the clutch mechanism will respond over a time period corresponding to many pulse cycles.
In the embodiments described herein, an exemplary controller type (i.e., a proportional and integral (PI) feedback controller) is described. However, those skilled in the arts of feedback control systems will understand that the teachings herein may be practiced using many types of controllers including, without limitation, the following: all types of proportional controllers; all types of PI controllers; all types of proportional, integral and derivative (PID) controllers; controllers having any type of compensator design and/or gain-scheduling; and all types or combinations of discrete, analog, and event-driven controllers.
The control pressure Pc output by the PWM valve 308 varies between Ps and Pr at a selected pulse frequency, Fpwm. The frequency of Fpwm may equal 100 Hz. Frequencies of several Hz to several hundreds of Hz are also commonly used). For most applications, the load 314 responds only at frequencies of much less than Fpwm, and is therefore sensitive only to an average value of Pc averaged over a plurality of cycles of Fpwm. Therefore, it is desirable to control the average value of Pc. The averaging filter 406 facilitates this requirement by providing an average value of the pressure feedback signal to the input processor 402.
A maximum frequency of interest for a control system response may be designated as “Fsystem”. The averaging filter 406 should have a well-defined cut-off frequency above Fsystem and below Fpwm. Further, the averaging filter should not cause a delay or phase shift sufficient to cause the feedback control system 400 to become unstable. The averaging filter 406 may be implemented using a variety of methods well known to persons skilled in the averaging filter design arts. These methods may include, without limitation, analog filter circuits and digital signal processing methods that provide an average of a digitized signal.
For each and all of the embodiments described herein, it should be understood that one or more averaging filters may be included according to the present teachings, even wherein the averaging filter or filters are not explicitly illustrated in the figures.
The input processor 402 performs processing functions such as signal scaling and A/D conversion. The input processor 402 may also compute an error signal based on the pressure command signal and the pressure feedback signal. The error signal may be input to the controller 304. In other embodiments, the input processor may output both the pressure command signal and the pressure feedback signals as separate signals to the controller 304, and the controller 304 may internally compute the error signal for feedback control processing. For simplicity, the controller 304 of the present example may comprise a proportional and integral (PI) controller. An exemplary PI controller is described in the above-noted related patent application Ser. No. 10/874,133, filed Jun. 22, 2004, titled “CLOSED-LOOP VALVE-BASED TRANSMISSION CONTROL ALGORITHM” (Eaton Ref. No. TBD). This related application is commonly owned by the assignee hereof, and is hereby fully incorporated by reference herein as though set forth in full, for its teachings on PI feedback controllers for hydraulic systems. These skilled in the art of feedback control design and manufacture will understand that the present teachings may also be applied using other feedback controllers (e.g., PID controllers).
In one embodiment, the controller 304 provides a duty cycle control signal to the PWM driver 306 and to a limit logic processor 408. The limit logic processor 408 provides protection from the following effects: “integral windup” (“integral windup” is a problem occurring for feedback controllers having integrators, as is well known to persons skilled in the art of feedback control systems); limit cycle output (wherein the duty cycle approaches the limits at which the maximum and minimum values of Pc occur); and other behaviors detrimental to feedback control performance. The output of the limit logic processor 408 is fed back to the controller 304 to provide additional control to the duty cycle control signal as described below in more detail. In other embodiments, the functions of the limit logic processor 408 may be performed by equivalent limit logic elements (not shown) included in the controller 304.
The limit logic processor 408 may include the following: a) exception logic wherein the error signal and the duty cycle signal values are evaluated to prevent integral windup; b) integral trim logic to reduce the quantitative effect of the integration process on the duty cycle control signal (e.g., by reducing the gain of the integration process); c) limit logic to prevent the duty cycle control signal from overshooting and undershooting selected limits; and d) Boolean logic based on activation or deactivation of limit logic processes. The functions performed by the limit logic processor have the following advantageous effects: a) to maintain overall stability of the controller under all operating conditions; b) to meet performance specifications of the closed-loop pressure control system; c) to achieve control pressure tracking or regulation under all load conditions and environments; d) to provide a general control system that will work with different types of PWM values; and e) to achieve desired system performance independent of sensor variations, component variations, and other system or operating variables.
In one embodiment, the following limit logic examples may be employed:
IF (DutyCycle>DCmax) AND (ErrorSignal>0), THEN Set Integration Process Output=Constant Value.
If the DutyCycle (i.e., the pulse width divided by the pulse rate of the duty cycle control signal) exceeds a selected duty cycle maximum (DCmax, which may comprise the duty cycle at which Pc reaches a maximum value, as described in more detail below), and the ErrorSignal (i.e., an error signal derived from the pressure command signal and the pressure feedback signal) is greater than zero, then an integration process performed by the controller 304 (
IF (DutyCycle<DCmin) AND (ErrorSignal<0), THEN Set Integration Process Output=Constant Value.
If the DutyCycle is less than a selected duty cycle minimum (DCmin, which may comprise the duty cycle at which Pc reaches a minimum value, as described in more detail below), and the ErrorSignal is less than zero, then the integration process performed by the controller 304 (
For each and all of the embodiments described herein, it should be understood that one or more limit logic processors may be included according to the present teachings, even wherein the limit logic processor or processors are not explicitly illustrated in the figures.
One known nonlinearity in a hydraulic fluid system is flow nonlinearity, which relates a flow variable of the fluid to the square root of a pressure variable of the fluid. An example of a procedure for deriving a control law α(t) based, in part, on flow nonlinearity is indicated by the equations shown in the following paragraph.
EQUATION 1, below, shows the relation between a fluid flow Q(t), such as the fluid flowing from the control port 114 of PWM valve 100 (
wherein ρ is the fluid density, α(t) is a nonlinear control law, Pr is the return pressure, and Kservo is the forward path gain. Based on EQUATION 1, an EQUATION 2 shown below may be derived.
Wherein A is a cross sectional area (e.g., the area of the control port 114 of
In one embodiment, the NLTerm expression, as a function of Pc(t) and Pref(t) may be determined by performing experimental measurements to obtain measurement data on a feedback system (e.g., the feedback control system 500 of
In another embodiment, the control law α(t) and the NLTerm may be represented by an analytic function of the variables Pc(t) and Pref(t), wherein the analytic function is determined according to an approximate fit to the measured data. In another embodiment, the control law α(t) and the NLTerm may be represented by a piece-wise linear function of the variables Pc(t) and Pref(t), wherein the piece-wise linear function is determined according an approximate fit to the measured data. Other methods for determining and/or representing the control law α(t) and the NLTerm (e.g., such as analytical derivation based on physical models of the feedback control system components) may also be used within the scope of the present teachings. As persons skilled in the arts of feedback control systems will readily understand from these teachings, the analytical square root flow nonlinearity (as shown by the square root expression in EQUATION 2) may thus be incorporated as part of the control law α(t) that is implemented by the compensator 504 and the controller 304 (
In another embodiment, external inputs may be provided to the compensator 504 by the external inputs element 502. The external inputs may be used to selectively enable signal processing (which may be linear or nonlinear) that compensates for effects such as temperature induced variation of fluid density, bulk modulus, and other hydraulic fluid properties. These external inputs may be used to selectively implement a “mode” of operation for the compensator 504. For example, an external input indicating the temperature of the hydraulic fluid may be provided from a temperature sensor (not shown). The fluid temperature input may be used in conjunction with explicit equations that determine the relation between the temperature of the fluid and the fluid properties (e.g., fluid density or bulk modulus) to derive an analytical compensation function or algorithm. The analytical function or algorithm may be selectively implemented by the compensator 504 responsive to the measured temperature of the fluid. For example, if the temperature exceeds a selected threshold, a first analytical function or algorithm may be implemented, and if the temperate does not exceed the selected threshold, a second analytical function or algorithm may be implemented. In another embodiment, a look-up table (not shown) or a gain-scheduling table (not shown) may be selected from a plurality of look-up tables or gain-scheduling tables to selectively implement a mode of operation, responsive to the external inputs. For example, the look-up table or gain-scheduling table may be selectively implemented by the compensator 504 according to the measured temperature of the fluid to effect compensation. If the temperature of the fluid is greater than a selected threshold value, a first table may be implemented, and if the temperature is less than the threshold value, a second table may be implemented. The look-up table may represent either a linear or a nonlinear signal processing function that is applied to the control variables (e.g., the pressure feedback signal, or an error signal computed from the pressure feedback signal and the pressure command signal). Thus, the compensator 504 may provide, without limitation, one or more of the following types of compensation: analytical linear compensation, analytical nonlinear compensation, empirical linear compensation, and empirical nonlinear compensation. Further, the mode of operation of the compensator 504 may be selectively modified responsive to the external inputs provided by external inputs element 502, as illustrated by the examples hereinabove.
For each and all of the embodiments described herein, it should be understood that one or more compensators may be included according to the present teachings, even wherein the compensator or compensators are not explicitly illustrated in the figures.
The controller A 304A receives input from the input processor 402 and provides output to a switching logic processor 602. The controller B 304B also receives input from the input processor 402 and provides output to the switching logic processor 602. In addition to receiving inputs from the controller A 304A and controller B 304B, the switching logic processor 602 may also optionally receive inputs from one, or any combination of the following elements: the external inputs element 502; the pressure sensor 316; a sensor A 604; a sensor B 606; and a sensor C 608. The sensor A 604 is coupled to the PWM valve 308 in order to generate a sensor signal, A, representative of a property, A, of the PWM valve 308. For example, the property A may be a position of the armature 102, or a current flowing in the coil 106 (
In one embodiment, the switching logic processor 602 determines whether the controller A 304A or the controller B 304B provides an output to the PWM driver 306 based on a fluid temperature signal input received from the sensor B 606. As is well known, the fluid temperature affects the fluid properties, such as the bulk modulus (i.e., the ratio of fluid pressure to a decrease in fluid volume), and viscosity. As fluid properties vary, the behavior of the fluid pressure Pc as a function of the PWM duty cycle also varies, and therefore properties of a feedback controller should adapt accordingly. For example, the controller gain, the compensation, the limit logic, and the pulse rate may all be modified based on the fluid temperature, in accordance with methods well known to persons skilled in the art of feedback control systems. Such modification can be effected, for example, by selecting either the controller A 304A or the controller B 305B based on the fluid temperature. In an alternative embodiment, the fluid temperature may be provided to the switching logic processor 602 via the external inputs element 502, which may receive a fluid temperature signal from an alternative temperature sensor (not shown). Exemplary logic for controller selection may be representative as follows.
In another embodiment, the switching logic processor 602 determines whether the controller A 304A or the controller B 305B provides output to the PWM driver 306. This determination is based on a signal provided by the external inputs element 502 that is representative of other operating conditions of the feedback control system 600. If the feedback control system 600 is part of an automatic transmission control system, for example, information regarding a current gear number (which may affect the properties of the load 314) may be received from the external inputs element 502, and used to determine which controller is selected (i.e., controller A 304A or controller B 304B).
The switching logic processor 602 may use a combination of inputs to determine which controller is used, as shown by EXAMPLE LOGIC 2 as set forth below.
In another embodiment, a pressure signal output by the pressure sensor 316 may be used by the switching logic processor 602 to determine whether the controller A 304A or the controller B 305B is used to provide output to the PWM driver 306. For example, for a selected range of values of the pressure signal, the controller A 304A may be used, and for another selected range of values of the pressure signal, the controller B 304 B may be used.
In yet another embodiment, the switching logic processor 602 determines whether the controller A 304A or the controller B 304B is used to provide output to the PWM driver 306 based on a current flowing in the coil 106 (
According to another embodiment, a property of the load 314 may be measured by the sensor C 608. In one example, if the load 314 comprises a piston (not shown), sensor C 608 may provide a signal that is representative of a position of the piston, and this signal may be used by the switching logic processor 602 to determine whether the controller A 304A or the controller B 305B is used to provide output to the PWM driver 306. Logic for this example may be set forth as follows: if the piston displacement exceeds a selected threshold, the controller A 304A is selected; otherwise the controller B 304B is selected.
As illustrated by the foregoing examples provided in reference to
The input processor 402 provides input to a phase-in filter 702 and a phase-out filter 704. The phase-in filter 702 and the phase-out filter 704 optionally receive input from the external inputs element 502. The phase-in filter 702 and the phase-out filter 704 may also optionally receive inputs from other sensors (not shown), as described above in reference to
According to one example, the magnitude of the signal provided by the phase-in filter 702 to the controller A 304A is increased as the magnitude of the signal provided by the phase-in filter 702 to the controller A 304A is decreased, and vice versa. In one example, phase-in filter 702 may comprise a high-pass frequency filter (not shown), and the phase-out filter 704 may comprise a low-pass frequency filter (not shown). In this case, the high-pass filter and the low-pass filter may be designed so that the summation of their outputs is a unity gain function across the frequencies comprising the inputs from the input processor 402 to the phase-in filter 702 and the phase-out filter 704. The output signals from the controllers A 304A and B 304B will be responsive to the signals received from the phase-in filter 702 and the phase-out filter 704, respectively. Therefore, the resulting summation by the PWM driver 306 of the signals received from the controllers A 304A and B 304B will be responsive to the controller A 304A for high frequencies, and to the controller B 304B for low frequencies, with a smooth transition in the range where the low-pass filter and the high-pass filter cut-off frequencies overlap. According to one example, the low-pass filter may pass frequencies less than ˜10 Hz and the high-pass filter may pass frequencies greater than ˜10 Hz, with an overlap frequency range of approximately ˜(8 to 12) Hz.
According to other embodiments, the following optional inputs may be provided to the phase-in filter 702 and the phase-out filter 704: the external inputs element 502 (
For selected operating conditions (e.g., pulse rate, PWM driver, Ps and Pr) the PWM valve 308 (
A pre-calibration of the PWM valve 308 can be performed prior to installation in a feedback control system (such as the feedback control systems 300, 400, 500, 600 and 700 described above in reference to
Based on the re-calibration of the Pc vs. DC plot described above, the controllers, compensators, gain-scheduling tables, look-up tables, switching logic processors, phase-in and phase-out filters, etc., disclosed hereinabove, may be adapted by the CPU 902 via operative connections (not shown) to these elements. In one example, the limit logic processor 408 (
In other embodiments, the controllers 304 (
In a further example, the switching logic processor 602 (
In yet another example, the phase-in filter 702 and the phase-out filter 704 (
Beginning at a STEP 1102, a pressure command signal and a pressure feedback signal are received, as for example, by an input processor or a controller. One or more of the following processes may be performed to provide a processed signal: a scaling process, an analog-to-digital conversion process, and an error signal computation process. The processed signal may comprise a single signal, or a plurality of signals (e.g., a single processed error signal, or a processed pressure feedback signal and a processed pressure command signal).
At a STEP 1104, one or more compensation processes may be performed on the processed signal to provide a compensated signal. Exemplary compensation processes may include the following: 1) a linear compensation process, and a 2) nonlinear compensation process. In some embodiments, the compensation process may comprise implementing a lookup table based, at least in part, on data measurements of the control pressure Pc as a function of the duty cycle control signal, wherein the data measurements are performed for an open-loop condition of the feedback control system. In some embodiments, the compensation process may comprise a control law α(t) based, at least in part, on the EQUATION 2, above. In some embodiments, the compensation process may be responsive to one or more other input signals. For example, the other input signals may comprise: 1) signals representing a temperature of the fluid, 2) signals representing a bulk modulus of the fluid, and 3) signals representing a current gear number for a gear in an automatic transmission system.
At a STEP 1106, a duty cycle control signal is provided, responsive, at least in part, to the compensated signal.
At a STEP 1108, the duty cycle control signal may be modified according to a limit logic process, responsive to the duty cycle control signal and to one or more selected duty cycle limit values. The limit logic process may comprise one or more of the following: (1) modifying a controller integration process to prevent integral windup; (2) reducing the quantitative effect of the integration process on the duty cycle control signal; (3) preventing the duty cycle control signal from overshooting and undershooting the selected duty cycle limit values. If the selected duty cycle limit values are not exceeded, the duty cycle control signal may be provided for a next STEP 1110 without modification. Otherwise, the modified duty cycle control signal is provided for the STEP 1110.
At the STEP 1110, a drive current is provided, responsive to the duty cycle control signal.
At a STEP 1112, a fluid is provided at a control pressure Pc to a load, responsive to the drive current.
At a STEP 1114, a pressure feedback signal is provided, responsive to a measurement of the control pressure Pc.
At a STEP 1116, the pressure feedback signal is averaged, and the averaged value of the pressure feedback signal is provided for the STEP 1102.
Beginning at a STEP 1202, a pressure command signal and a pressure feedback signal are received, as for example, by an input processor or a controller. One or more of the following processes may be performed to provide a processed signal: a scaling process, an analog-to-digital conversion process, and an error signal computation process. The processed signal may comprise a single signal, or a plurality of signals (e.g., a single processed error signal, or a processed pressure feedback signal and a processed pressure command signal).
At a STEP 1204, a plurality of duty cycle control signals are provided, responsive to the processed signal.
At a STEP 1206, a switching logic processor receives the plurality of duty cycle control signals and one or more other input signals. A selected duty cycle control signal is selected from the plurality of duty cycle control signals, responsive, at least in part, to at least one other input signal. The other input signals may comprise one or more of the following signals: (1) the pressure feedback signal; (2) a signal representing a property of the fluid; (3) a signal representing a property of the PWM valve; (4) a signal representing a property of the load; and (5) a signal representing a current gear number for a gear in an automatic transmission system. In one example, the signal representing a property of the fluid may comprise a signal representing a temperature of the fluid.
At a STEP 1208, a drive current is provided, responsive to the selected duty cycle control signal of the STEP 1206.
At a STEP 1210, a fluid at a control pressure Pc is provided to a load, responsive to the drive current.
At a STEP 1212, the control pressure Pc is measured to provide the pressure feedback signal, responsive to the measured control pressure Pc. The pressure feedback signal is provided to the STEP 1202.
Beginning at a STEP 1302, a pressure command signal and a pressure feedback signal are received by an input processor. The input processor may perform one or more of the following: a scaling process, an analog-to-digital conversion process, and an error signal computation process. The input processor provides a processed signal that may comprise one, on a plurality of signals (e.g., an error signal, or a processed pressure feedback signal and a processed pressure command signal).
At a STEP 1304, a plurality of phase-in and phase-out filters receive the processed signal and provide a plurality of duty cycle control signals. In one example, the phase-in and phase-out filters may comprise bandpass filters. The phase-in and phase-out filters may provide filtered signals that are responsive, at least in part, to one or more other input signals. The other input signals may comprise one or more of the following signals: (1) the pressure feedback signal; (2) a signal representing a property of the fluid; (3) a signal representing a property of the PWM valve; (4) a signal representing a property of the load; and (5) a signal representing a current gear number for a gear in an automatic transmission system. In one example, the signal representing a property of the fluid may comprise a signal representing a temperature of the fluid.
At a STEP 1306, the filtered signals are provided to a plurality of controllers. Each controller generates a duty cycle control signal that is provided to a PWM driver.
At a STEP 1308, the duty cycle control signals are received by a PWM driver. The PWM driver combines the duty cycle control signals and generates a drive current.
At a STEP 1310, a PWM valve receives the drive current and provides a fluid at a control pressure Pc to a load, responsive to the drive current.
At a STEP 1312, a pressure sensor measures the control pressure Pc and provides the pressure feedback signal responsive to the measured control pressure Pc. The pressure feedback signal is provided to the input processor in accordance with the STEP 1302.
A number of embodiments of the present inventive concept have been described. Nevertheless, it will be understood that various modifications may be made without departing from the scope of the present teachings. For example, it should be understood that the functions described as being part of one module may in general be performed equivalently in another module.
Accordingly, it is to be understood that the inventive concept is not to be limited by the specific illustrated embodiments, but only by the scope of the appended claims. The description may provide examples of similar features as are recited in the claims, but it should not be assumed that such similar features are identical to those in the claims unless such identity is essential to comprehend the scope of the claim. In some instances the intended distinction between claim features and description features is underscored by using slightly different terminology.
This application is related to the pending U.S. patent application Ser. No. 10/874,133, filed Jun. 22, 2004, titled “CLOSED-LOOP VALVE-BASED TRANSMISSION CONTROL ALGORITHM”; and to U.S. Pat. No. 6,807,472, issued Oct. 19, 2004, titled “CLOSED LOOP CONTROL OF SHIFTING CLUTCH ACTUATORS IN AN AUTOMATIC SPEED CHANGE TRANSMISSION.” This pending patent application and issued patent are commonly owned by the assignee hereof, and are hereby fully incorporated by reference herein as though set forth in full, for teachings on closed-loop hydraulic control systems.