IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0912956
(2013-06-07)
|
등록번호 |
US-9886008
(2018-02-06)
|
발명자
/ 주소 |
- Gahinet, Pascal
- Chen, Rong
- Eryilmaz, Bora
- Singh, Baljeet
|
출원인 / 주소 |
|
대리인 / 주소 |
|
인용정보 |
피인용 횟수 :
0 인용 특허 :
16 |
초록
▼
Embodiments provide techniques, computer-readable media, and devices for allowing users to perform interactive design of controllers, such as PID controllers, in a free-form modeling environment. Users can tune controllers using characteristics familiar to typical users rather than having to specify
Embodiments provide techniques, computer-readable media, and devices for allowing users to perform interactive design of controllers, such as PID controllers, in a free-form modeling environment. Users can tune controllers using characteristics familiar to typical users rather than having to specify gain values for the controller, which may be difficult for a user to relate to the performance of a controller.
대표청구항
▼
1. A method comprising: for a representation of a Proportional Integral Derivative (PID) controller, the representation including a first free parameter, a second free parameter, a first fixed parameter, and a second fixed parameter,selecting a plurality of values for the first free parameter;select
1. A method comprising: for a representation of a Proportional Integral Derivative (PID) controller, the representation including a first free parameter, a second free parameter, a first fixed parameter, and a second fixed parameter,selecting a plurality of values for the first free parameter;selecting a plurality of values for the second free parameter;evaluating, by processing logic, a merit function for pairs that include a first value for the first free parameter and a second value for the second free parameter;selecting, by the processing logic, a pair of the pairs for which the merit function is evaluated, the selected pair satisfying the merit function for a determined crossover frequency, where the determined crossover frequency is one of the first fixed parameter or the second fixed parameter, wherethe merit function includes a setpoint tracking measure term,is an input disturbance rejection measure term, anda robust stability measure term;determining, by the processing logic, one or more of a Proportional (P) gain, an Integral (I) gain, or a Derivative (D) gain of the PID controller based on the selected pair of values for the first free parameter and the second free parameter satisfying the merit function; andgenerating, by the processing logic, executable instructions to implement the PID controller. 2. The method of claim 1, where the merit function is represented as: F=w1FRT+w2FDR+w3FDMw1+w2+w3where,w1 is a first weight,w2 is a second weight,w3 is a third weight,FRT is the setpoint tracking measure term,FDR is the input disturbance rejection measure term, andFDM is the robust stability measure term. 3. The method of claim 2, where the setpoint tracking measure term is represented as: FRT=10e-π(T+T2-1)where T=maxω(T(jω)), and T is a closed-loop transfer function. 4. The method of claim 2, where the input disturbance rejection measure term is represented as: FDR=[cos2(φL2)CL(0)]0.3whereφL is a total phase lead, CL(0) is the gain of CL(jω) where jω=0. 5. The method of claim 2, where the robust stability measure term is represented as: FDM=δtargetδ=tan(PMtarget2)S-Twhereδ is a disc margin,δtarget is a target disc margin,PMtarget is a target phase margin,S is a loop sensitivity function, andT is a closed loop transfer function. 6. The method of claim 1, where the first fixed parameter is a gain cross over frequency, ωc, and the second fixed parameter is a phase margin, θm. 7. The method of claim 6, where the first free parameter is α and the second free parameter is β, and α and β vary between 0 and 90 degrees. 8. The method of claim 1, where the evaluating includes at least one of: a direct search technique;a gradient-descent technique; anda gridding technique. 9. The method of claim 1, where the evaluating includes a gridding technique performed at determined increments. 10. The method of claim 1, where the merit function is represented as: F=max(FRT,FDR,FDM)where,FRT is the setpoint tracking measure term,FDR is the input disturbance rejection measure term, andFDM is the robust stability measure term. 11. The method of claim 10, where the setpoint tracking measure term is represented as: FRT=10e-π(T+T2-1)where T=maxω(T(jω)), and T is a closed-loop transfer function. 12. The method of claim 10, where the input disturbance rejection measure term is represented as: FDR=[cos2(φL2)CL(0)]0.3whereφL is a total phase lead, CL(0) is the gain of CL(jω) where jω=0. 13. The method of claim 10, where the robust stability measure term is represented as: FDM=δtargetδ=tan(PMtarget2)S-Twhereδ is a disc margin,δtarget is a target disc margin,PMtarget is a target phase margin,S is a loop sensitivity function, andT is a closed loop transfer function. 14. The method of claim 1 further comprising: identifying a constraint for use with at least one of the first free parameter and the second free parameter; anddiscarding at least one of the first free parameter or the second free parameter for violating the constraint. 15. The method of claim 14, where the first free parameter is a and the second free parameter is β, andthe constraint is Δφ−90<β−α,where Δφ represents a total phase shift of the PID controller. 16. The method of claim 1 further comprising: discarding at least one pair of the first free parameter and the second free parameter that fail to satisfy a Nyquist stability test. 17. The method of claim 1, where the selected pair of values for the first free parameter and the second free parameter satisfies the merit function by producing a smallest value of the merit function. 18. One or more non-transitory computer-readable media comprising program instructions, the program instructions when executed by a processing element operable to: initiate an interactive tuning interface, the interactive tuning interface configured to: compute one or more loop responses for a model of a controller used with a model of a plant,graphically display the one or more computed loop responses,compute performance and robustness information for the model of the controller, andgraphically display the computed performance and robustness information;linearize at least a portion of the model of the plant to produce a linearized plant model;receive one or more inputs specifying a gain crossover frequency and an open-loop phase margin for the model of the controller;automatically tune the model of the controller by algebraically solving one or more first parameters of the model of the controller based on the specified gain crossover frequency and the open-loop phase margin,optimizing one or more second parameters of the model of the controller;determine one or more of a Proportional (P) gain, an Integral (I) gain, or a Derivative (D) gain for the model of the controller based on the one or more first parameters and the one or more second parameters;generate executable instructions to implement the model of the controller; anddisplay, at the interactive tuning interface, a response of the model of the controller during execution of the model of the controller with the model of the plant, where the model of the controller represents the controller using values for a first free parameter and a second free parameter that are determined by satisfying a merit function, andthe merit function includes a setpoint tracking measure term,an input disturbance rejection measure term, anda robust stability measure term. 19. The one or more non-transitory computer-readable media of claim 18 where the merit function is represented as: F=w1FRT+w2FDR+w3FDMw1+w2+w3where,w1 is a first weight,w2 is a second weight,w3 is a third weight,FRT is the setpoint tracking measure term,FDR is the input disturbance rejection measure term, andFDM is the robust stability measure term. 20. The one or more non-transitory computer-readable media of claim 18, where the merit function is represented as: F=max(FRT,FDR,FDM)where,FRT is the setpoint tracking measure term,FDR is the input disturbance rejection measure term, andFDM is the robust stability measure term. 21. The one or more non-transitory computer-readable media of claim 18, wherein the first fixed parameter is a gain cross over frequency, ωc, and the second fixed parameter is a phase margin, θm. 22. The one or more non-transitory computer-readable media of claim 21, where the first free parameter is α and the second free parameter is β, and α and β vary between 0 and 90 degrees. 23. A method for tuning a proportional integral derivative (PID) controller, the method comprising: for a design objective for the PID controller, the design objective including: a closed-loop stability; anda robustness measure,determining a first value that represents a gain crossover frequency for an open-loop response of the PID controller;determining a second value that represents a phase margin for the open-loop response;determining automatically, by processing logic, at least two free parameters of the PID controller to enable the PID controller to satisfy the design objective, the determining the at least two free parameters based on a first transfer function;determining automatically, by the processing logic, two setpoint weights of the PID controller, the two setpoint weights determining a setpoint overshoot of the PID controller, the determining the two setpoint weights based on a second transfer function, the second transfer function independently adjustable relative to the first transfer function to provide the PID controller with two degrees of freedom (2DOF), the determining the two setpoint weights including evaluating a merit function using selected pairs of values for the two setpoint weights to identify a pair of values for the two setpoint weights that satisfy the merit function, the merit function including the second transfer function;determining, by the processing logic, one or more of a Proportional (P) gain, an Integral (I) gain, or a Derivative (D) gain of the PID controller based on the at least two free parameters determined to enable the PID controller to satisfy the design objective and the pair of values for the two setpoint weights identified as satisfying the merit function; andgenerating, by the processing logic, executable instructions to implement the PID controller. 24. The method of claim 23 where the at least two free parameters are α and β, and α and β vary between 0 and 90 degrees, andthe two setpoint weights are b and c. 25. The method of claim 23 where the evaluating the merit function includes at least one of: a direct search technique;a gradient-descent technique; anda gridding technique. 26. The method of claim 23 wherein the merit function includes an upper bound on the second transfer function and a lower bound on the second transfer function. 27. The method of claim 23 wherein the PID controller is configured to control a model of a plant, the method further comprising: presenting a plot of a response of the plant to a change in a setpoint signal. 28. The method of claim 27 further comprising: receiving a change to a performance characteristic of the PID controller;generating a new PID controller that satisfies the received change to the performance characteristic. 29. The method of claim 28 wherein the performance characteristic of the PID controller is at least one of: a response time;a bandwidth frequency; anda phase margin. 30. The method of claim 28 further comprising: presenting an updated plot of the response of the plant based on the new PID controller. 31. The method of claim 30 wherein the updated plot is presented in real-time. 32. The method of claim 23 wherein the merit function is given by: F2DOF=maxω≤ωcmax[k1log10(T2DOF(ω)Tmax),k2log10(TminT2DOF(ω))]whereωc is a closed-loop bandwidth of a one degree of freedom loop of the PID controller,k1 and k2 are constants, andTmin and Tmax are lower and upper bounds. 33. The method of claim 32 wherein Tmin and Tmax, are given by: Tmin(ω)=min(1max(1,ω/(ωc/1.5)),T2DOF(ω)), Tmax(ω)=1. 34. One or more non-transitory computer-readable medium comprising program instructions, the program instructions when executed by a processing element operable to: for a first design objective and a second design objective for the PID controller, the first design objective including: a closed-loop stability, anda robustness measure, and the second design objective including reference tracking, select a first value for a first fixed parameter (α);select a second value for a second fixed parameter (β); andadjust a plurality of free parameters of the PID controller when the first value and the second value are specified by tuning the PID controller so that the tuned PID controller satisfies the second design objective, where:the plurality of free parameters of the PID controller include γ and φ, andthe first and second fixed parameters and the plurality of free parameters are used to express the PID controller in a continuous time expression represented as: C2(s)=Kccosφzcosβωccosγcosφs(sinγs+ωccosγωc)(sinφs+ωccosφsinαs+ωccosα)whereC2(s)=C(s)+Cf(s),C(s) represents a serial compensator, andCf(s) represents a feedforward compensator; anddetermine one or more of a Proportional (P) gain, an Integral (I) gain, or a Derivative (D) gain of the PID controller based on values for the plurality of free parameters, the first value for the first fixed parameter, and the second value for the second fixed parameter, where the values for the plurality of free parameters are determined by evaluating selected values of the plurality of free parameters using a merit function; andgenerate executable instructions to implement the PID controller. 35. One or more non-transitory computer-readable media comprising program instructions, the program instructions when executed by a processing element operable to: for a representation of a proportional integral derivative (PID) controller that includes a first free parameter, a second free parameter, a first fixed parameter, and a second fixed parameter,select a plurality of values for the first free parameter;select a plurality of values for the second free parameter;search a first range that includes one or more of the plurality of values for the first free parameter and the second free parameter, the search performed using an optimization technique;identify a constraint for use with the one or more of the plurality of values for the first free parameter and the second free parameter;discard one or more values for the first free parameter and the second free parameter that violate the constraint or that fail to satisfy a Nyquist stability test;evaluate a first merit function for pairs that include a value of the first free parameter and a value of the second free parameter that are not discarded and that satisfy the Nyquist stability test, the merit function operating in the optimization technique;select a pair of the first merit function evaluated pairs of values for the first free parameter and the second free parameter that satisfies the first merit function for a determined crossover frequency, where the crossover frequency is one of the first fixed parameter or the second fixed parameter, where the first free parameter is α and the second free parameter is β, andthe first merit function selects a value for α and β;identify a plurality of values for a third free parameter;identify a plurality of values for a fourth free parameter;search a second range that includes one or more of the plurality of values for the third free parameter and the fourth free parameter;evaluate a second merit function for pairs that include a value of the third free parameter and a value of the fourth free parameter;select a pair of the second merit function evaluated pairs that satisfies the second merit function, where the first free parameter is γ and the second free parameter is φ, and γ and φ are used in the expression: C2(s)=Kccosφzcosβωccosγcosφs(sinγs+ωccosγωc)(sinφs+ωccosφsinαs+ωccosα)whereC2(s)=C(s)+Cf(s),C(s) represents a serial compensator, andCf(s) represents a feedforward compensator; anddetermine one or more of a Proportional (P) gain, an Integral (I) gain, or a Derivative (D) gain of the PID controller based on the selected pair of values for the first free parameter and the second free parameter that satisfies the first merit function; andgenerate executable instructions to implement the PID controller. 36. The one or more non-transitory computer-readable medium of claim 33 wherein the second merit function is configured to tune reference tracking performance of the PID controller. 37. The one or more non-transitory computer-readable medium of claim 35 wherein the pair of values for the first free parameter and the second free parameter satisfy the first merit function by producing a smallest value of the first merit function, and the selected pair of values for the third free parameter and the fourth free parameter first free parameter satisfy the second merit function by producing a smallest value of the second merit function.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.