Feedback controller

• US 5,867,384 A
• Filed: 07/08/1997
• Issued: 02/02/1999
• Est. Priority Date: 07/08/1997
• Status: Expired due to Fees

• Current Assignee: Johnson Controls Technology Company (Johnson Controls International plc)
• Original Assignee: Johnson Service Company
• Inventors: Drees, Kirk H., Seem, John E., Ben-Aissa, Nebil
• Primary Examiner(s): Gordon, Paul P.
• Assistant Examiner(s): Garland, Steven R.
• Application Number:
  US08/889,682
• Time in Patent Office: 574 Days
• Field of Search: 364/148.01, 364/152-157, 364/160-63, 364/166, 364/176-179, 318/135, 318/636, 318/609, 318/610, 318/561
• US Class Current: 700/42
• CPC Class Codes: G05B 11/42 for obtaining a characteristic which is both proportional and time-dependent, e.g. P. I., P. I. D.
• CPC Class Codes: G05B 13/024 in which a parameter or coefficient is automatically adjusted to optimise the performance
• CPC Class Codes: G05B 5/01 electric
• CPC Class Codes: G05B 7/02 electric

Abstract

A closed loop feedback control system that operates to maintain a process variable at or near a process variable setpoint. The system includes a setpoint regulator that adjusts the value of a process variable in response to a setpoint associated therewith. A sensor associated with the setpoint regulator senses the process variable and outputs a process variable signal corresponding thereto. A controller connected to both the setpoint regulator and the sensor in a closed loop is programmed with PID logic to control operation of the setpoint regulator in response to the process variable signal received from the sensor. The controller generates a control signal that controls the setpoint regulator based on a non-uniform two-point sampling of the process variable signal. The controller delivers smooth, bump-free, control in spite of changes in the setpoint controller gains or system parameters. A digital filter is also implemented with the controller of the present invention to filter the process variable signal input into the controller. The filter includes both an amplitude to clip signal spikes and glitches and a low pass frequency filter to remove higher order noise-related frequencies from the signal, thereby reducing system memory requirements and increasing system computational speed.

Claims

1. A control system comprising:

a process regulator that adjusts the value of a process variable in response to a setpoint associated therewith;

a sensor associated with the process regulator that senses the process variable and that outputs a process variable signal corresponding thereto; and

a controller connected to both the process regulator and the sensor in a closed loop, the controller being programmed with PID logic to control operation of the process regulator in response to the process variable signal received from the sensor, and to generate a control signal, including proportional, integral and derivative function contributions, that controls the process regulator based on a non-uniform two-point sampling of the process variable signal.

2. The system of claim 1, wherein the controller generates a bumpless control signal after being released from an override condition, after a controller proportional gain is changed and after the controller is switched from a disabled state to an enabled state.

3. The system of claim 2, wherein the controller resets the integral function contribution to provide the bumpless control signal after being released from an override condition, after a change in the proportional gain, and after the controller is switched from a disabled state to an enabled state.

4. The system of claim 1, wherein the PID logic integral function co ntribution is calculated with second order accuracy.

5. The system of claim 4, wherein the controller calculates the integral function contribution through a trapezoidal integration technique.

6. The system of claim 1, wherein the controller integral function contribution minimizes control delay upon the controller returning to a control mode after reaching saturation at either a high or low control limit.

7. The system of claim 6, wherein the controller resets the integral function contribution upon reaching saturation to minimize delay in generating the control signal.

8. The system of claim 1, wherein the process variable comprises a temperature of a climate-controlled zone, and the setpoint comprises a zone temperature setpoint.

9. The system of claim 1, wherein the derivative contribution is generated by the controller through use of the process variable in a finite difference approximation.

10. The system of claim 9, wherein the finite difference approximation utilizes a non-uniform sampling period.

11. The system of claim 1, wherein the controller derivative contribution is limited to minimize the effects of noise through recursive implementation of the following equation:

##EQU5## where;

D is the derivative contribution, T_d is the derivative time, N is a constant which limits possible noise amplification, K is the controller gain, t is time, and y is the value of the process variable.

12. The system of claim 1, wherein the controller includes a built-in saturation detector that allows the controller to be utilized for control flow or logic coordination with additional control devices.

13. The system of claim 1, further comprising a digital filter coupled between the sensor and the controller to filter noise from the process variable signal.

14. The system of claim 13, wherein the digital filter comprises an adaptive amplitude filter that clips signal spikes and glitches in the process variable signal;

and

a low pass frequency filter coupled to the amplitude filter that filters high frequency noise from the process variable signal, thereby increasing computational speed and reducing memory requirements for the control system.

15. The system of claim 14, wherein the amplitude filter includes a window that permits small changes in the process variable signal, but that clips large changes in the process variable signal, the window width expanding exponentially according to the transient nature of the process variable signal.

16. The system of claim 14, wherein the frequency filter comprises a first order exponential filter.

17. The system of claim 14, wherein the frequency filter comprises a second order Butterworth filter that introduces minimal phase lag to the process variable signal, and rejects noise from the process variable signal above a predetermined cutoff frequency.

18. The system of claim 17, wherein the second order Butterworth filter includes a characteristic cutoff frequency limited by the sampling rate of the process variable signal and determined by the input rate of the process variable signal.

19. A feedback control method, comprising the steps of:

setting a process setpoint at a desired value;

determining deviation of a process variable from the process setpoint via a non-uniform sampling of the process variable; and

controlling the process variable in response to said step of determining deviation of a process variable to reduce the deviation in a manner that minimizes setpoint overshoot and undershoot wherein said controlling is accomplished via digital Proportional integral derivative (PID) logic that produces a control signal having a proportional function contribution, an integral function contribution and a derivative function contribution.

20. The method of claim 19, further comprising the step of calculating the integral contribution via a trapezoidal integration technique.

21. The method of claim 19, where in the step of determining deviation from the process setpoint via a non-uniform sampling of the process variable comprises calculating a finite difference approximation for the derivative contribution via the non-uniform sampling of the process variable.

22. The method of claim 19, further comprising the step of determining that saturation has been reached during the step of controlling the process variable when the control output has not changed from a high or low output limit for a predetermined period of time.

23. The method of claim 19, further comprising the step of resetting the integral function contribution as needed to provide a bumpless control output subsequent to state or parameter changes.

24. The method of claim 19, further comprising the step of filtering an input signal containing the process variable to minimize signal noise.

25. The method of claim 24, where in the step of filtering an input signal comprises the step of digitally filtering the signal to both clip signal spikes and signal glitches, and to eliminate high frequency signal noise.

26. The method of claim 25, wherein the step of filtering the input signal to clip signal spikes and glitches and further includes the step of exponentially expanding a filter window to accommodate rapid changes in signal value.