Optimized parameterization of active disturbance rejection control
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G05B-013/02
G05B-019/416
출원번호
US-0445345
(2014-07-29)
등록번호
US-10061275
(2018-08-28)
발명자
/ 주소
El-Shaer, Ahmed
Tian, Gang
Yajurvedi, Girish
출원인 / 주소
LINESTREAM TECHNOLOGIES
대리인 / 주소
Amin, Turocy & Watson LLP
인용정보
피인용 횟수 :
0인용 특허 :
37
초록▼
A system for tuning a control system uses a simplified tuning procedure to generate robustly stabilizing tuning parameters that reduce or eliminate undesired system oscillations in the presence of long system dead times or phase lag. A control method is used to establish a relationship between the p
A system for tuning a control system uses a simplified tuning procedure to generate robustly stabilizing tuning parameters that reduce or eliminate undesired system oscillations in the presence of long system dead times or phase lag. A control method is used to establish a relationship between the plant parameters of a controlled system and the tuning parameters of a parameterized active disturbance rejection controller determined to be optimal or substantially optimal for the control system. The plant parameters include the system gain, time constant, and dead time. Corresponding tuning parameters include the controller bandwidth and a system gain estimate. Using the system gain estimate as a tuning parameter can alleviate the influence of large dead times or phase lags on system response. Once established, these fixed relationships can be used to determine suitable tuning parameters for specific motion or process control applications based on the system gain and dominant constraints of the system.
대표청구항▼
1. A method for determining tuning parameters, comprising: determining, by a system comprising a processor using an iterative search algorithm that maintains a constant ratio between an observer bandwidth and a controller bandwidth, sets of values of the controller bandwidth and a system gain estima
1. A method for determining tuning parameters, comprising: determining, by a system comprising a processor using an iterative search algorithm that maintains a constant ratio between an observer bandwidth and a controller bandwidth, sets of values of the controller bandwidth and a system gain estimate that satisfy a robust stability constraint for respective sets of values of a system gain, a time constant, and a dead time;applying, by the system, a curve-fitting method to the sets of values of the controller bandwidth and the system gain estimate to yield a tuning parameter model;receiving, by the system, system parameter values for a control system that controls a controlled system, the system parameter values comprising at least received values of the system gain, the time constant, and the dead time of the controlled system, wherein the control system comprises a controller having an extended state observer that is a function of the system gain estimate and the observer bandwidth, and having a control law that is a function of the system gain estimate and the controller bandwidth;referencing, by the system, the tuning parameter model;determining, by the system based on the referencing, robustly stabilizing values of the controller bandwidth and the system gain estimate for the control system based on the received values of the system gain, the time constant, and the dead time; andtuning, by the system, the control system using the robustly stabilizing values of the controller bandwidth and the system gain estimate as tuning parameters to facilitate stable performance of the control system. 2. The method of claim 1, wherein the tuning comprises tuning, using the robustly stabilizing values of the controller bandwidth and the system gain estimate determined by the system, the controller having the extended state observer represented by ż=Az+Bu+L(y−Cz)and the control law represented by at least one of u=1b0(P(r-z1)-z2)oru=1b0(P(r-y)-z2)where u is a system input, y is a system output, z=[z1z2] is an extended state observer state vector, z1 and z2 are components of the extended state observer state vector, matrix L=[2ωoωo2], ωo is the observer bandwidth of the extended state observer, matrix A=[0100], matrix B=[bo0], matrix C=[1 0], P=ωc, ωc is the controller bandwidth, and b0 is the system gain estimate. 3. The method of claim 1, wherein the robust stability constraint is given by a structured singular value. 4. The method of claim 1, further comprising performing the receiving, the referencing, the determining, and the tuning periodically or substantially periodically during runtime of the control system. 5. The method of claim 1, wherein the receiving comprises at least one of receiving the system parameter values via manual input or receiving the system parameter values from an estimation system. 6. The method of claim 1, wherein the controlled system comprises at least one of an industrial robot, a positioning system, a pump, a spin dryer, a washing machine, a centrifuge, a conveyor, a palletizer, or a web tension control system. 7. The method of claim 1, wherein the tuning parameter model comprises at least one of an expression or a look-up table data structure that defines, as the sets of values of the controller bandwidth and the system gain estimate, values of the controller bandwidth and the system gain estimate as a function of the system gain, the time constant, and the dead time. 8. A system for determining robustly stabilizing tuning parameters, comprising: a memory; anda processor configured to execute executable components stored on the memory, the executable components comprising: an interface component configured to receive system parameter values for a control system that controls a controlled system, wherein the system parameter values comprise at least values of a system gain, a time constant, and a dead time of the controlled system, and the control system comprises a controller having an extended state observer that is a function of a system gain estimate of the system gain and an observer bandwidth, and a control law that is a function of the system gain estimate and a controller bandwidth;a tuning parameter determination component configured to generate a tuning parameter model by identifying, based on an iterative search algorithm while maintaining a constant ratio between the observer bandwidth and the controller bandwidth, sets of values of the controller bandwidth and the system gain estimate that satisfy a robust stability constraint for respective sets of values of the system gain, the time constant, and the dead time, and applying a curve-fitting method to the sets of values of the controller bandwidth and the system gain estimate to yield the tuning parameter model anddetermine robustly stabilizing values of the controller bandwidth and the system gain estimate as a function of the system parameter values based on a referencing of the tuning parameter model; anda tuning component configured to tune the control system using the robustly stabilizing values of the controller bandwidth and the system gain estimate as tuning parameters to facilitate disturbance rejection for the control system. 9. The system of claim 8, wherein the tuning parameter model comprises at least one of an expression or a look-up table data structure that defines, as the sets of values of the controller bandwidth and the system gain estimate, values of the controller bandwidth and the system gain estimate as a function of the system gain, the time constant, and the dead time. 10. The system of claim 8, wherein the robust stability constraint is given by a structured singular value. 11. The system of claim 8, wherein the control system conforms to a transfer function represented by: G(s)=kτs+1e-Tdswhere s is a Laplace transform operator, k is the system gain, τ is the time constant, and Td is the dead time. 12. The system of claim 8, wherein the extended state observer is represented by ż=Az+Bu+L(y−Cz)and the control law is represented by at least one of u=1b0(P(r-z1)-z2)oru=1b0(P(r-y)-z2)where u is a system input, y is a system output, z=[z1z2] is an extended state observer state vector, z1 and z2 are components of the extended state observer state vector, matrix A=[0100], matrix B=[b00], matrix L=[2ωoωo2], matrix C=[1 0], ωo is the observer bandwidth of the extended state observer, ωc is the controller bandwidth, b0 is the system gain estimate, and P=ωc. 13. The system of claim 8, wherein the controlled system is at least one of an industrial robot, a positioning system, a pump, a spin dryer, a washing machine, a centrifuge, a conveyor, a palletizer, or a web tension control system. 14. The system of claim 8, wherein the tuning parameter determination component is configured to determine, as the robustly stabilizing values, values of the controller bandwidth and the system gain estimate defined in the tuning parameter model as corresponding to the system parameter values received by the interface component. 15. A non-transitory computer-readable medium having stored thereon executable instructions that, in response to execution, cause a computer system to perform operations, the operations comprising: generating a tuning parameter model, wherein the generating comprises: determining, using an iterative search algorithm that maintains a constant ratio between an observer bandwidth and a controller bandwidth, sets of values of the controller bandwidth and a system gain estimate that satisfy a robust stability constraint for corresponding sets of values of a system gain, a time constant, and a dead time, andapplying a curve-fitting method to the sets of values of the controller bandwidth and the system gain estimate to yield the tuning parameter model;receiving system parameter values for a control system that controls a mechanical system, wherein the system parameter values comprise at least values of the system gain, the time constant, and the dead time of the mechanical system, and the control system comprises a controller having an the extended state observer that is a function of the system gain estimate and the observer bandwidth, and a control law that is a function of the system gain estimate and the controller bandwidth;selecting, based on a referencing of the tuning parameter model, robustly stabilizing values of the controller bandwidth and the system gain estimate for the control system based on the system parameter values; andsetting at least one controller gain coefficient for the control system using the robustly stabilizing values of the controller bandwidth and the system gain estimate as tuning parameters to facilitate stable performance of the control system. 16. The non-transitory computer-readable medium of claim 15, wherein the generating comprises generating the tuning parameter model based on a determination of robustly stabilizing values of the controller bandwidth and the system gain estimate corresponding to the respective sets of values of the system gain, the time constant, and the dead time. 17. The non-transitory computer-readable medium of claim 15, wherein the robust stability constraint is given by a structured singular value. 18. The non-transitory computer-readable medium of claim 15, wherein the tuning parameter model comprises at least one of an expression or a look-up table data structure that defines the robustly stabilizing values of the controller bandwidth and the system gain estimate as a function of the system gain, the time constant, and the dead time. 19. The non-transitory computer-readable medium of claim 15, wherein the operations further comprise performing the receiving, the selecting, and the tuning periodically or substantially periodically during runtime of the control system. 20. The non-transitory computer-readable medium of claim 15, wherein the mechanical system is at least one of an industrial robot, a positioning system, a pump, a spin dryer, a washing machine, a centrifuge, a conveyor, a palletizer, or a web tension control system.
Bradley L. Hunter ; Christopher P. Cullen, Method and subsystem for generating a trajectory to be followed by a motor-driven stage when processing microstructures at a laser-processing site.
Matsubara Shunsuke (Oshino-mura JPX) Iwashita Yasusuke (Oshino-mura JPX) Okita Tadashi (Oshino-mura JPX), Tool life control method based on disturbance load torque of motor.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.