Adaptive adjustment system for nonlinear dynamic system
By automatically adjusting the regulation parameters in the nonlinear dynamic system of the motor vehicle, the problem of complex and costly manual parameterization in the prior art is solved, and efficient regulation system adaptability and efficiency are achieved.
Patent Information
- Application Number
- CN202510339700.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-03-25
- Filing Date
- 2025-03-21
- Publication Date
- 2025-09-26
AI Technical Summary
In the prior art, nonlinear dynamic system adjustment devices of motor vehicles need to be manually parameterized for different operating states, which leads to complex system design and high costs. In addition, the number and range of adjustment parameters are limited, which may lead to insufficient adjustment behavior.
A regulation system is provided, which optimizes the parameter set by automatically adjusting the parameters of the proportional, differential and integral components within the operating range, storing the regulation parameters using a lookup table or a distribution table, and automatically adapting the parameters according to the regulation deviation and adaptation conditions.
It realizes automatic parameter adaptation within different operating ranges, reduces the need for manual parameterization, improves the efficiency and adaptability of the regulation system, and reduces system complexity and application costs.
Smart Images
Figure CN120704107A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a regulating device for a technical system, in particular a nonlinear dynamic system. Background Art
[0002] Technical devices in motor vehicles, such as steering systems, traction control mechanisms, and braking systems, often require PI or PID control to ensure functional safety. However, these devices exhibit highly nonlinear and dynamic behavior, leading to the design of control systems often being designed in an operating-state-dependent manner, with control parameters varying depending on the operating state. This often results in a complex overall system and significant operational complexity, as the control system must be individually designed for each operating state. Consequently, in the case of a traction control system, as an example of a technical device to be controlled, the operational complexity is substantial, with parameterization for various operating ranges typically lasting approximately 18 months. Because parameterization is performed manually, the number of control parameters and operating ranges is often limited, and system dependencies are therefore greatly simplified, approximated, or not considered. This can lead to inadequate control behavior in certain operating states.
[0003] It is therefore an object of the present invention to provide an improved regulating system which can independently carry out parameter adaptation for different operating ranges, wherein in particular the parameters can also be adjusted during normal operation. Summary of the Invention
[0004] This object is achieved by a regulating system for a technical device, in particular in a motor vehicle, according to claim 1 and by a method for operating a regulating system according to the complementary claim.
[0005] Further refinements are described in the dependent claims.
[0006] According to a first aspect, a method for operating a control system for controlling a technical device of the technical system using a proportional component and at least one of a differential component and an integral component as control components is provided, comprising the following steps:
[0007] - providing control parameters of a parameter set for calculating control components as a function of an operating range determined by an operating point and / or an operating state of the technical system;
[0008] - performing a control of the technical device based on the control deviation and the control parameter;
[0009] Adapting the control parameters for one or more operating ranges in response to the presence of an adaptation condition based on the control behavior within an optimization range assigned to the adaptation condition, wherein the optimization range determines the duration within which the course of the control deviation is used for adapting the control parameters.
[0010] To regulate the nonlinear dynamic behavior of mapping technology devices based on PI, PD, or PID, control parameters are often predefined in an operating range-dependent and / or operating state-dependent manner. Control parameters can correspond to control factors, which are applied by multiplying the corresponding proportional, differential, and / or integral components. The control device accordingly uses the control parameters predefined for this operating range for regulation and changes the control parameters when a new operating range is determined. Operating ranges are often defined in a rule-based manner, for example, by value ranges for individual operating variables or operating state variables, and a parameter set consisting of control parameters is assigned to each of the thus defined operating ranges.
[0011] For example, parameter sets associated with the operating range can be maintained in a lookup table or an assignment table, so that the regulating system always has access to the operating range associated with the current operating point or operating state.
[0012] The aforementioned control system now provides for adapting one or more parameter sets if it is determined that the control quality within a past time period is insufficient. The control quality is determined based on the temporal profile of the control deviation that occurs.
[0013] In this context, adaptation conditions are defined which specify when one or more of the parameter sets must be adapted.
[0014] Furthermore, the first adaptation condition may provide that adaptation is performed if a substantial undershoot or overshoot occurs. A substantial undershoot or overshoot is determined if the control deviation in terms of value is greater than a predefined first deviation amount within a predetermined first time period, and the control deviation subsequently changes sign and the control deviation in terms of value exceeds a further deviation amount.
[0015] Furthermore, the second adaptation condition can provide that the adaptation is carried out if the control deviation is above a predetermined second deviation amount for a predetermined second duration.
[0016] If the presence of one of the adaptation conditions is thus detected, an optimization range is first determined. For the first adaptation condition, the optimization range corresponds to the time period between the time at which the control deviation begins to exceed a predetermined first deviation amount and the time at which, after a change in sign of the control deviation, an excess of a further deviation amount is determined.
[0017] For the second adaptation condition, the optimization range corresponds to the duration for which the control deviation exceeds the second deviation amount or the second duration or a predefined duration that is greater than the second duration.
[0018] The optimization range determines the duration within which the course of the control deviation is used for adapting the control parameters.
[0019] The control device can be operated with a predefined control cycle time, for example, between 5 ms and 100 ms. The optimization range can be predefined as a duration in ms or as a number of control cycles. For example, the control cycle time can be 20 ms, and the optimization range can include two or a few up to several dozen control cycles.
[0020] If the adaptation condition is met, then for each of the control parameters, ie the control factors Kp, Ki and Kd, the control model is now in the form
[0021] u=Ki∫e dt+Kd de / dt+e*Kp
[0022] The respective contribution to the manipulated variable u is determined based on the control deviation e. In particular, it is determined to what extent the proportional component, the differential component, and / or the integral component contribute to the control deviation.
[0023] For this purpose, the contribution variable B is determined individually for each control parameter Kp, Ki and Kd. Kp 、B Ki 、B Kd The contribution variable resulting from the numerical sum of the corresponding terms for the controlled component is:
[0024]
[0025] The integral component is determined at the time of the start of the optimization range, and the time step of the control cycle is j=1 . . . N for the optimization range.
[0026] Thus, for example, the contribution of the proportional component to the control deviation B Kp It can be the sum of the quotient of the value of the proportional component for the considered time step and the sum of the values of the proportional component, the differential component and / or the integral component for the considered time step over all time steps of the control cycle of the optimization range. Similarly, the contribution of the differential component to the control deviation B Kd It can be the sum of the quotient of the value of the differential component for the considered time step and the sum of the values of the proportional component, the differential component and / or the integral component for the considered time step over all time steps of the control cycle of the optimization range. Similarly, the contribution of the integral component to the control deviation B KiThis can be the sum of the quotient of the value of the integral component for the considered time step and the sum of the values of the proportional, differential and integral components for the considered time step, formed over all time steps of the control loop of the optimization range.
[0027] If a transition occurs between operating ranges within the optimization process, the contribution of the manipulated variable for each operating range is determined individually for each control cycle. This results in contribution variables for the proportional, differential, and integral components for a plurality of operating ranges, each of which serves as a basis for adapting the manipulated variable in the relevant operating range.
[0028] The parameter set of the control parameters for the operating range occurring within the optimization range is thus automatically adapted according to the contribution variables of the control components. The control parameters for the respective operating range are now adjusted according to their assigned contribution variables with the corresponding learning rates LRp, LRd, LRi.
[0029]
[0030] The learning rates LRp, LRd, and LRi can be predefined depending on the respective operating range. In particular, the learning rate can be selected based on the control deviation. For example, the larger the square error of the control deviation, the higher the learning rate. For very small square errors, the learning rate will also be very low. This allows the learning behavior to converge (the learning rate converges to 0).
[0031] Furthermore, the adaptation of the control variables for the different operating ranges can be weighted in addition depending on the time portion of the optimization range that the different operating ranges occupy.
[0032] The control parameter is always adjusted in such a way that it is changed in the direction of the manipulated variable that results in a compensation of the control deviation. In other words, in the case of a positive control deviation, the control parameter is reduced, and vice versa. This makes it possible to adapt the corresponding operating range in the event of a misadjustment of the control parameter without having to perform manual parameterization. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] The following is a more detailed explanation of the embodiments based on the accompanying drawings, wherein:
[0034] Figure 1 A schematic diagram showing a control system with automatic parameter adjustment for the respective operating range;
[0035] Figure 2 shows a flow chart illustrating a method for performing parameter adjustment; and
[0036] Figure 3a and 3b The course of the actual value of the control is shown to illustrate the occurrence of the first adaptation condition or the second adaptation condition. DETAILED DESCRIPTION
[0037] Figure 1 A block diagram is shown for illustrating a control system 1 for a technical device 2 in a technical system 3, such as a vehicle. The invention is described below with reference to a traction control system of a vehicle.
[0038] Control system 1 includes operating range or operating state identification based on operating variables B and operating state variables Z. In the case of a traction control mechanism, operating variables B may be speed, acceleration, tire temperature, etc. Operating state variables Z may include weather conditions, road surface type, tire tread, ambient temperature, etc. Operating variables B and operating state variables Z may be detected by sensory means or determined using a model-based approach. By dividing the operating variables and operating state variables into ranges, different operating ranges can be identified and defined in operating range block 4 for which parameter sets should be determined or adapted.
[0039] In the parameter set selection block 5, a specific parameter set of control parameters Kp, Kd, Ki (Kp is the proportional factor, Kd is the differential factor, and Ki is the integrator factor, i.e., the integral factor) is assigned, using a lookup table or assignment table, to the identified operating range in which the technical system 3 is located. These control parameters can be defined individually for each operating range. The parameter set is fed to the control device 6, which can be implemented as a PD, PI, or PID control device.
[0040] The regulating device obtains a regulating deviation e between a setpoint preset RSoll and an actual value Rist, which can also correspond to an operating variable. The regulating deviation can correspond, for example, in the form of a speed difference between the vehicle speed and the speed of the drive wheels.
[0041] The control device 6 generates the manipulated variable u from the control deviation e:
[0042] u=Ki∫e dt+Kd de / dt+e*Kp
[0043] The regulating device can also be designed as a differential regulating device, wherein the control deviation e then corresponds to the differential component.
[0044] The manipulated variable u output by the control device 6 can be adapted to the requested motor torque so that, according to the setpoint specification R, an optimal slip (Schlupf), i.e., an optimal deviation between the wheel speed and the vehicle speed, can be achieved. The parameter set selection block 5 enables the selection of a suitable parameter set based on the identified operating state. The parameter set can then be adapted based on the quality of the control.
[0045] In the adaptation block 7, the regulation parameters are adapted according to the operating range and updated in the parameter set selection block 5. In the adaptation block 7, the following is performed according to Figure 2 The flow chart describes the method in more detail.
[0046] In step S1 , the control deviation e is determined, and in step S2 , it is checked whether an adaptation condition exists. The control deviation e is therefore temporarily stored, and the time course of the control deviation is analyzed.
[0047] If Figure 3a As shown by way of example in , the control deviation e is greater than a predetermined first deviation amount AB1 for a predetermined first duration Z1 and then the sign of the control deviation e changes and exceeds a further deviation amount ABD for a predetermined further shorter duration, then a first adaptation condition may exist.
[0048] Alternatively or additionally, if as in Figure 3b If, as shown by way of example in FIG. 1 , the control deviation e lies above a second predetermined deviation amount AB2 within a second predetermined time period Z2 , a second adaptation condition can be identified.
[0049] Alternatively or additionally, other criteria for the adaptation conditions may also be defined.
[0050] If the adaptation condition is met (alternative expression: yes), the method continues with step S3 , otherwise (alternative expression: no), the method jumps to step S10 .
[0051] In step S3, an optimization range H is determined. For the first adaptation condition, the optimization range H begins at the time when the control deviation e exceeds a predetermined first deviation amount AB1 by value and ends at the time when the control deviation e, with an opposite sign, exceeds a predetermined further deviation amount ABD by value. For the second adaptation condition, the optimization range H begins at the time when the control deviation e exceeds a predetermined second deviation amount AB2 by value and ends at the time when the control deviation e no longer exceeds the second deviation amount AB2 or after a second predetermined time period Z2.
[0052] In the subsequent step S4, the control components Ki∫e dt, e*Kp, wherein the control parameters Kp, Kd, Ki are used for the operating range determined for each time step.
[0053] In the following step S5, the contribution of the individual control components, i.e., the contribution of the differential component, the integral component, and the proportional component to the entire manipulated variable u determined for the relevant control cycle j, is now checked for each control cycle j=1 ... N within the optimization range. To this end, the contribution of the individual components for this control cycle is calculated as follows:
[0054]
[0055] In each control cycle, the control variables resulting from the corresponding operating range or operating state range are used.
[0056]
[0057] By summing the individual contributions of the controller components of a control loop, it is possible to determine which controller component, i.e., the proportional component, the differential component, and / or the integral component, makes the greatest contribution to the determination of the manipulated variable u for a specific operating range within the optimization range H. The summation is always performed only for control loops assigned to the same operating range. The contributions of the controller components are calculated and evaluated individually for multiple operating ranges within the optimization range H.
[0058] The manipulated variable u results in an actual value Rist, which is compared with a setpoint preset Rsoll. In the case of a traction control system, the setpoint preset and the actual preset can correspond to a setpoint drive torque or a setpoint speed or an actual drive torque or an actual speed. To adjust the parameter set for the operating range assignment, in step S6, for each time segment within the optimization range H assigned to a specific operating range, the ratio of the contribution of the proportional component, the differential component, or the integral component to the total contribution can be used as the contribution share.
[0059]
[0060] The adaptation of the control parameters of the parameter set for the relevant operating range occurring during this time period within the optimization range H is thus automatically carried out as a function of the contribution of the relevant control components.
[0061] For the correcting parameters for the respective operating ranges, the correcting parameters Kp_neu, Kd_neu, Ki_neu are now adjusted in step S7 as a function of their assigned contribution variables with the respective learning rates LRp, LRd, LRi.
[0062]
[0063] The learning rates LRp, LRd, LRi can be predefined depending on the respective operating range. Alternatively, the learning rate can be selected depending on the control deviation.
[0064] Based on the contribution, the parameter set for the time segment of the operating range within the optimization range H is adjusted, with the adjustment being performed so that the manipulated variable u is increased or decreased so that it counteracts the control deviation e. In other words, if the control deviation is negative for the traction control, the torque manipulated variable is increased by increasing the corresponding control factor. Conversely, if the control deviation is positive, the control factor is reduced.
[0065] In step S8 , the adjusted control parameters are stored back in the parameter set selection block 5 depending on the operating range.
[0066] In step S9 , the manipulated variable u is determined by the regulating device 6 by summing the regulating components.
[0067] Then jump back to step S1.
Claims
1. A method, in particular an at least partially computer-implemented method, for operating a control system (1) of a technical device (2) for controlling a technical system (3) using a proportional component and at least one of a differential component and an integral component as control components, comprising the following steps: - providing control parameters (Kp, Kd, Ki) of a parameter set for calculating the control component as a function of an operating range determined by an operating point (B) and / or an operating state (Z) of the technical system (3); - performing a control of the technical device based on the control deviation (e) and the control parameters (Kp, Kd, Ki); - Adapting (S7) the control parameters (Kp, Kd, Ki) for one or more operating ranges based on the control behavior within an optimization range (H) assigned to the adaptation condition, depending on the presence of an adaptation condition, wherein the optimization range (H) determines the duration within which the change process of the control deviation (e) is used for adapting the control parameters (Kp, Kd, Ki).
2. A method according to claim 1, wherein a first adaptation condition stipulates that if a substantial undershoot or overshoot occurs, adaptation is performed, wherein the substantial undershoot or overshoot is determined if the numerical control deviation (e) is greater than a predetermined first deviation amount (AB1) within a predetermined first time duration (Z1) and the control deviation (e) subsequently changes sign and the numerical control deviation (e) exceeds another deviation amount (ABD), wherein the assigned optimization range (H) corresponds to the time duration between the time point at which the control deviation (e) begins to be above the predetermined first deviation amount (AB1) and the time point at which the exceeding of the another deviation amount (ABD) is determined after the change of sign of the control deviation (e).
3. A method according to claim 1 or 2, wherein a second adaptation condition stipulates that adaptation is performed if the adjustment deviation (e) is above a predetermined second deviation amount (AB2) for more than a predetermined second duration (Z2), wherein the optimization range (H) corresponds to the duration for which the adjustment deviation (e) exceeds the second deviation amount (AB2) or the second duration (Z2) for the second adaptation condition.
4. A method according to any one of claims 1 to 3, wherein a contribution share of each of the regulating components is determined within one or more time segments within an optimization range (H) respectively assigned to the operating range, wherein the contribution share determines the degree of adaptation of the regulating parameters (Kp, Kd, Ki) to be adapted.
5. The method according to claim 4, wherein for each control parameter (Kp, Kd, Ki), a contribution to the control component (B Kp , B Ki , B Kd ), the contribution being obtained by the sum of the values of the quotient of the value of the corresponding control component for the considered time step and the sum of the values of all control components over all time steps of the control loop to be assigned to the corresponding operating range in the optimization range (H).
6. A method according to claim 5, wherein, in order to adapt the parameter set of the control parameters (Kp, Kd, Ki) for the operating range occurring within the optimization range (H), an adaptation is performed for each control parameter (Kp, Kd, Ki) according to the share of the relevant contribution to the total contribution.
7. The method according to claim 6, wherein corresponding learning rates are taken into account for the adaptation and wherein the adjustment of the control parameters (Kp, Kd, Ki) for different operating ranges is weighted according to the time portion of the associated time segment in the optimization range (H). 8 . The method according to claim 7 , wherein the adjustment of the control variables (Kp, Kd, Ki) is performed in such a way that the adjustment is changed in the direction of a manipulated variable which counteracts the control deviation (e).
9. The method according to claim 1, wherein the method is used for operating a traction control system, wherein the control deviation indicates a difference between a setpoint speed and an actual speed or a difference between a predefined setpoint torque and an actual torque of the drive motor.
10. A device for performing one of the methods according to any one of claims 1 to 9.
11. A computer program product comprising instructions which, when executed by at least one data processing device, cause the data processing device to perform the steps of the method according to any one of claims 1 to 9.
12. A machine-readable storage medium comprising instructions which, when executed by at least one data processing device, cause the data processing device to perform the steps of the method according to any one of claims 1 to 9.