초록 ▼

A predictive and self-tuning PI controller includes an apparatus for variably assigning gains thereto in accordance with specified time functions following a change in set-point. The control gains are computed in accordance with a plurality of input parameters and subsequently continuously adjusted

A predictive and self-tuning PI controller includes an apparatus for variably assigning gains thereto in accordance with specified time functions following a change in set-point. The control gains are computed in accordance with a plurality of input parameters and subsequently continuously adjusted to set-point errors based on a GPC approach thereby to optimize performance indices derived from simple and classical user specifications. Variation and tuning of control gains in accordance with set-point errors permits the PI controller to be useful in the general area of process control and applicable to a wider range of processes compared to traditional PI control, including time-delay processes, unstable processes and processes with time-varying dynamics.

대표청구항 ▼

1. A process control system for controlling a process including: input means for inputting a first signal representing a set-point for a process and a second signal representing a process variable; a summer generating an error signal representing a difference between said first and second signa

1. A process control system for controlling a process including: input means for inputting a first signal representing a set-point for a process and a second signal representing a process variable; a summer generating an error signal representing a difference between said first and second signals; a controller connected to provide a control signal for the process as a function of said error signal and of control gains thereof; and an associated control structure providing signals to said controller, said associated control structure providing predictive gain values for said controller and self tuning gains of said controller. 2. The process control system of claim 1, wherein said associated control structure comprises gain design means responsive to an input parameter set from a user and generating a time function signal for controlling gains in accordance with said input parameter set.3. The process control system of claim 2, wherein said input parameter set includes parameters representing at least one of: a first-order mathematical model of the process, a parameter representing an estimate of a time-delay of the process, parameters representing a desired closed-loop damping factor and natural frequency, and a sampling interval and wherein said gain design means generates said time function signal in response to each parameter of said parameter set input thereto.4. The process control system of claim 3, wherein said input parameter set further includes a prediction horizon.5. The process control system of claim 4, wherein said gain design means generates a default prediction horizon and replaces said default prediction horizon by said prediction horizon when included in said input parameter set.6. The process control system of claim 3, further comprising gains scheduling means, said gains scheduling means responding to said set-point and to a schedule depending on elapsed time following a set-point change for assigning gains to said controller in accordance with said time function signal generated by said gain design means, whereby time scheduled gains are assigned to said controller.7. The process control system of claim 6, further comprising gains self-tuning means, said gains self-tuning means responding to a difference between said set-point signal and a controlled process variable for repeatedly adjusting weighting parameters of said time function signal to optimize control performance and for providing a weighting parameter adjusted time function signal to said time scheduled gains scheduling means for self-tuning gains assigned to said controller.8. The process control system of claim 2, further comprising gains self-tuning means, said self-tuning means responding to a difference between said set-point signal and a controlled process variable for repeatedly adjusting weighting parameters of said time function signal to optimize control performance and for providing a weighting parameter adjusted time function signal to said time scheduled gains scheduling means for self-tuning gains assigned to said controller.9. The process control system of claim 8, further comprising gains scheduling means, said gains scheduling means responding to said set-point signal and to a schedule depending on elapsed time following a set-point change for assigning gains to said controller in accordance with said time function signal generated by said gain design means, whereby time scheduled gains are assigned to said controller.10. The process control system of claim 9, further comprising gains self-tuning means, said self-tuning means responding to a difference between said set-point signal and a controlled process variable for repeatedly adjusting weighting parameters of said time function signal to optimize control performance and for providing a weighting parameter adjusted time function signal to said time scheduled gains scheduling means for self-tuning gains assigned to said controller.11. The process control system of c laim 10, wherein said input parameter set includes parameters representing each of: a first-order mathematical model of the process, a parameter representing an estimate of a time-delay of the process, parameters representing a desired closed-loop damping factor and natural frequency, and a sampling interval and wherein said gain design means generates said time function signal in response to each parameter of said parameter set input thereto.12. A PI control apparatus comprising: input means for inputting a first signal representing a set-point for a process and a second signal representing a process variable; means for generating an error signal representing a difference between said first and second signals; control means for providing a signal to the process as a function of said error signal and of control gains thereof; gain design means receiving a user input parameter set and generating time functions relating the control gains to at least one input parameter of said input parameter set and to elapsed time following a set-point change; gains scheduling means for updating the PI control gains based on said elapsed time according to a time schedule depicted by said time functions; and gains self-tuning means for continuously updating and fine tuning parameters of said time functions based on said error signal. 13. The PI control apparatus as recited in claim 12, wherein said user input parameter set comprises: two parameters (a, b) of a first-order mathematical model of the process; one parameter (L) representing a time delay of the process; two parameters (&xgr;, &ohgr; n) representing a desired closed-loop damping factor and natural frequency; one parameter (T s) representing a sampling interval of the data acquisition system; and one optional parameter (P) representing a prediction horizon. 14. The PI control apparatus as recited in claim 12, wherein the said time functions are given in the equations (24)-(27) of Table 1.15. The PI control apparatus as recited in claim 13, wherein said gains design means assigns said optional parameter representing the prediction horizon a default value of 10 when said optional parameter is unspecified.16. The PI control apparatus as recited in claim 12, wherein a first pair of time functions are used in generating the PI control gains if said elapsed time is less than said time delay, and a second pair of time functions different from said first pair of time functions are used in generating the PI control gains if said elapsed time is more than said time delay.17. The PI control apparatus as recited in claim 12, wherein the parameters of said time functions are updated according to equations (30)-(31) in Table 2, and default adaptation gains of ß1=ß2=0.00001 are used therein.