System and method for controlling processes

• US 5,754,424 A
• Filed: 02/27/1996
• Issued: 05/19/1998
• Est. Priority Date: 02/27/1996
• Status: Expired due to Term

• Current Assignee: Katherine E. Melvin
• Original Assignee: Kenneth P. Melvin
• Inventors: Melvin, Kenneth P.
• Primary Examiner(s): RUGGIERO, JOSEPH F
• Application Number:
  US08/607,469
• Time in Patent Office: 812 Days
• Field of Search: 364/148, 364/152-156, 364/157, 364/158, 364/160-163, 364/176, 364/177, 364/174, 318/561, 318/609, 318/610, 318/600, 318/601
• US Class Current: 700/37
• CPC Class Codes: G05B 11/36 with provision for obtaining particular characteristics, e.g. proportional, integral, differential
• CPC Class Codes: G05B 15/02 electric

Abstract

A multi-purpose controller for analog variables of one dimension which closely controls such variables to set point by making sequential corrections to the level of input to the system of the variable based on: system response time, relationship of controller output to the variable, deviation of the variable from set point, and change in system load as determined by combining change in error with change in system input. The controller has a self tuning capability and can provide generation of the set point according to a set of variable parameters. It can be configured as a stand alone controller, as an intelligent Input/Output device for another intelligent device, or be resident in another intelligent device such as a programmable controller, computer or other microprocessor based device.

Claims

1. A controller to control the operation of a system, said system having at least one controllable input and at least one measurable variable, said system including a first signal conditioner to convert a control signal from said controller to an input form required by said system for said controllable input and, a second signal conditioner to detect and convert said measurable variable to a data signal having a format compatible with said controller, said controller comprising:

a processor having two input terminals and an output terminal, one of said input terminals disposed to be coupled to said second signal conditioner of said system to receive said data signal, the second of said input terminals disposed to receive a signal representative of a user selected set point for the operation of said system, and said output terminal disposed to be coupled to said first signal conditioner to provide said control signal for conversion to said at least one controllable input of said system, said processor to control the operation of said controller and to sequentially calculate and develop said control signal to control the operation of said system; and

a memory coupled to said processor and said second signal conditioner to store previous data signals, and intermediate results and values from said processor;

wherein said processor sequentially approximates a present error as the difference between a present set point and a present value of said measurable variable, and calculates a present slope in error of said system;

wherein said processor sequentially sets a present control signal from said controller proportional to a sum of a previous control signal, the present error, and the present slope in error.

2. The controller as in claim 1 further comprising a time generator coupled to said processor, wherein said processor approximates a response time for said system as the total time of the difference in time between when a difference in said control signal is applied to a first coupling unit and when a difference is noted in said data signal that is representative of a change in said measurable variable by said controller for an unloaded system.

3. The controller as in claim 1 wherein said processor further determines a system characteristic that is proportional to a change in said control signal from said controller to a resultant change in said data signal that is representative of a change in said measurable variable.

4. The controller as in claim 3 wherein the determination of said present control signal also includes said system characteristic multiplied with each of said present error and said present slope in error of said system to convert the units of said error and slope in error to units of said control signal.

5. The controller of claim 4 wherein the determination of said present control signal further includes the addition of a value that is proportional to a change in said present control signal.

6. The controller of claim 2 wherein said processor further determines a system characteristic that is proportional to a change in said control signal to a resultant change in said measurable variable.

7. The controller of claim 1 wherein said controller is disposed to be incorporated as an intelligent input/output device for another intelligent input/output device with at least one additional input/output device to control an aspect of the operation of said system.

8. The controller of claims 1 wherein said controller, in communication with an intelligent device forms part of a network of such devices wherein each of said intelligent devices controls a different aspect of operation of said system.

9. The controller of claim 1 further includes a bus coupled to said processor via which information from and to other intelligent devices can be transmitted and received.

10. The controller of claim 1 further includes an interface disposed to receive operational parameters from a local operator interface, said interface being coupled to said processor.

11. The controller of claim 1 further includes a display to display to a user selected operational parameters and operating values of said system prior to, during and after operation.

12. The controller as in claim 1 wherein said present slope of error is the change in set point of said measurable variable from a previous interval to a present interval minus the change in said measurable variable of said system from the same previous interval to the same present interval.

13. The controller as in claim 12:

further comprising a time generator coupled to said processor, wherein said processor approximates a response time for said system as the total time of the difference in time between when a difference in said control signal is applied to said first coupling unit and when a difference is noted in said data signal representative of a change in measurable variable by said controller.

14. The controller as in claim 12 wherein a calculated control signal is proportional to said present control signal plus a change in said control signal.

15. The controller as in claim 2 wherein the portion of said present error added as a portion of said present control signal is proportional to a number of time interval cycles occurring within said response time of said system.

16. A controller to control the operation of a system, said system having at least one controllable input and at least one measurable variable, said system including a first signal conditioner to convert a control signal from said controller to an input form required by said system for said controllable input and, a second signal conditioner to detect and convert said measurable variable to a data signal having a format compatible with said controller, said controller comprising:

a processor having two input terminals and an output terminal, one of said input terminals disposed to be coupled to said second signal conditioner of said system to receive said data signal, the second of said input terminals disposed to receive a signal representative of a user selected set point for the operation of said system, and said output terminal disposed to be coupled to said first signal conditioner to provide said control signal for conversion to said at least one controllable input of said system, said processor to control the operation of said controller and to sequentially calculate and develop said control signal to control the operation of said system;

a memory coupled to said processor and said second signal conditioner to store previous data signals, and intermediate results and values from said processor; and

a time generator coupled to said processor to generate a base cycle time of said processor;

wherein said processor approximates a system response time for said system as the total time of the difference in time between when a difference in said control signal is applied to a first coupling unit and when a difference is noted in said data signal representative of a change in measurable variable by said controller;

wherein said processor determines a system characteristic that is proportional to the ratio of a change in said control signal to a resultant change in said measurable variable;

a present error between actual measurable variable and the set point for the measurable variable; and

a present slope in error of said measurable variable;

wherein a cycle time of said processor is the time necessary to make measurements and perform calculations; and

wherein said processor sequentially sets a present control signal proportional to the sum of a previous control signal, said present error multiplied by said system characteristic and said cycle time divided by said system response time, and a present change in load that is arrived at by doubling said present slope in error multiplied by said system characteristic.

17. A controller to control the operation of a system, said system having at least one controllable input and at least one measurable variable, said system including a first signal conditioner to convert a control signal from said controller to an input form required by said system for said controllable input and, a second signal conditioner to detect and convert said measurable variable to a data signal having a format compatible with said controller, said controller comprising:

a processor having two input terminals and an output terminal, one of said input terminals disposed to be coupled to said second signal conditioner of said system to receive said data signal, the second of said input terminals disposed to receive a signal representative of a user selected set point for the operation of said system, and said output terminal disposed to be coupled to said first signal conditioner to provide said control signal for conversion to said at least one controllable input of said system, said processor to control the operation of said controller and to sequentially calculate and develop said control signal to control the operation of said system;

a memory coupled to said processor and said second signal conditioner to store previous data signals, and intermediate results and values from said processor; and

a time generator coupled to said processor to generate a base cycle time of said processor;

wherein said controller controls the rate at which the operation of said system is changed over a period of time to achieve operation at a selected set point by making adjustments to a present set point at selected intervals, initially several parameter and variable values are preset or initialized and stored in said memory including final set point, previous set point, previous rate of change of set point, maximum rate of change of set point and acceleration of set point, and the value of said output of said second signal conditioner corresponding to said measurable value is stored in said memory; and

wherein said processor during each cycle time compares said present set point with said final set point to determine if a change in said present set point is necessary,

if said previous set point equals said final set point, said present rate of change of set point and acceleration of set point are each set to zero, resulting in said set point remaining unchanged and the present set point being set equal said previous set point, and

if said previous set point is less than or greater than said final set point, the values of said present rate of change of set point and acceleration of set point may be adjusted to cause said present set point to change to achieve said final set point.

18. The controller of claim 17 wherein an absolute value of said present rate of change of set point is limited at each cycle time to not exceed said preselected maximum rate of change of set point.

19. The controller of claim 17 wherein during each cycle time if said previous set point is not equal to said final set point, a first value of said set point is estimated at a next unit of time without a change in said previous set point, a second value of set point is estimated where a rate of change of said set point is to be changed for said system to reach said final set point, said first and second values are compared, and from the result of said comparison a next present set point is determined.

20. The controller of claim 19 wherein if said final set point is greater than said present set point, where said first value is compared to said second value,

if said second value is greater than said first value, said present rate of change of set point is compared to said maximum rate of change of said set point,

if said present rate of change of set point has the lesser value, said acceleration of set point remains unchanged and a new present set point is selected proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point, and

if said maximum rate of change of said set point has the lesser value, said rate of change of set point is set equal to said maximum rate of change of set point, said acceleration of set point is set to zero, and a new present set point is selected proportional to the sum of said maximum rate of change of set point multiplied by said cycle time, and said previous set point, and

if said first value is greater than said second value, said acceleration of set point is set proportional to the negative of said present rate of change of set point squared divided by the difference between said present set point and said final set point, and said next present set point is selected proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point.

21. The controller of claim 19 wherein if said final set point is less than said present set point, where said first value is compared to said second value,

if said first value is greater than said second value, said present rate of change of set point is compared to the negative of said maximum rate of change of said set point,

if said present rate of change of set point has the greater value, said acceleration of set point is set to the negative value of itself and a new present set point is selected proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point, and

if said negative of said maximum rate of change of said set point has the greater value, said rate of change of set point is set equal to the negative of said maximum rate of change of set point, said acceleration of set point is set to zero, and a new present set point is selected proportional to the sum of said maximum rate of change of set point multiplied by said cycle time, and said previous set point, and

if said first value is less than said second value, said acceleration of set point is set proportional to said present rate of change of set point squared divided by the difference between said present set point and said final set point, and said new present set point is selected proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point.

22. A method for controlling the operation of a system by developing a control signal, said system having at least one controllable input and at least one measurable variable, said system including a first signal conditioner to convert said control signal from a controller to an input form required by said system for said controllable input and, a second signal conditioner to detect and convert said measurable variable to a data signal having a format compatible with said controller, said controller disposed to receive a signal representative of a user selected set point for the operation of said system, said method of developing said control signal comprising the steps of:

a. sequentially approximating a present error as the difference between a present set point and a present value of said measurable variable; and

calculate a present slope in error of said system;

b. sequentially setting said present control signal proportional to a sum of a previous control signal, the present error, and the present slope in error; and

c. approximating a response time for said system as the total time of the difference in time between when a difference in said control signal is applied to said first coupling unit and when a difference is noted in said data signal that is representative of a chance in said measurable variable.

23. The method of claim 22 further including the step of:

d. approximating a system characteristic that is proportional to a change in said control signal to a resultant change in said data signal that is representative of a change in said measurable variable.

24. The method as in claim 22 further including the step of:

d. calculating said present slope of error as the change in set point of said measurable variable from a previous interval to a present interval minus the change in said measurable variable of said system from the same previous interval to the same present interval.

25. The method as in claim 24 further including the steps of:

e. approximating a response time for said system as the total time of the difference in time between when a difference in said control signal is applied to a first coupling unit and when a difference is noted in said data signal that is representative of a change in said measurable variable.

26. The method as in claim 24 further including the step of:

e. calculating a control signal that is proportional to said present control signal plus a change in said control signal.

27. The method as in claim 22 wherein the portion of said present error of step a. added as a portion of said present control signal in step b. is proportional to a number of time interval cycles occurring within said response time of said system.

28. A method for controlling the operation of a system by developing a control signal, said system having at least one controllable input and at least one measurable variable, said system including a first signal conditioner to convert said control signal from a controller to an input form required by said system for said controllable input and, a second signal conditioner to detect and convert said measurable variable to a data signal having a format compatible with said controller, said controller disposed to receive a signal representative of a user selected set point for the operation of said system, wherein cycle time is the time necessary to make measurements and perform calculations, said method of developing said control signal comprising the steps of:

a. generating a base cycle time;

b. approximating a system response time for said system as the total time of the difference in time between when a difference in said control signal is applied to said first coupling unit and when a difference is noted in said data signal that is representative of a change in said measurable variable;

c. sequentially approximating a present error as the difference between a present set point and a present value of said measurable variable;

d. calculating a present slope in error of said system;

e. approximating a system characteristic that is proportional to a change in said control signal to a resultant change in said data signal that is representative of a change in said measurable variable;

f. approximating a present change in load by doubling said present slope in error multiplied by said system characteristic; and

g. sequentially setting a present control signal proportional to the sum of a previous control signal, said present error multiplied by said system characteristic and said cycle time divided by said system response time, and said present change in load.

29. A method for controlling the operation of a system by developing a control signal, said system having at least one controllable input and at least one measurable variable, said system including a first signal conditioner to convert said control signal from a controller to an input form required by said system for said controllable input and, a second signal conditioner to detect and convert said measurable variable to a data signal having a format compatible with said controller, said controller disposed to receive a signal representative of a user selected set point for the operation of said system, said method of developing said control signal comprising the steps of:

a. generating a base cycle time;

b. presetting or initializing final set point, previous set point, previous rate of change of set point, maximum rate of change of set point and acceleration of set point, and the value of said data signal corresponding to said measurable value;

c. comparing, during each cycle time, a present set point with said final set point to determine if a change in said present set point is necessary;

d. if as a result of step c. said previous set point equals said final set point, said present rate of change of set point and acceleration of set point are each set to zero, resulting in said set point remaining unchanged and the present set point being set equal said previous set point; and

e. if as a result of step d. said previous set point is less than or greater than said final set point, the values of said present rate of change of set point and acceleration of set point may be adjusted to cause said present set point to change to achieve said final set point.

30. The method of claim 29 further including the step of:

f. limiting the absolute value of said present rate of change of set point at each cycle time to not exceed said preselected maximum rate of change of set point.

31. The method of claim 29 wherein step e. further includes the steps of:

f. estimating a first value of said set point at a next unit of time without a change in said previous set point;

g. estimating a second value of said set point where said rate of change of said set point is to be changed for said system to reach final set point;

h. comparing said first and second values of steps f. and g.; and

i. determining a next present set point from the result of step h.

32. The method of claim 31 wherein step i. further includes the steps of:

j. determining if said final set point is greater than said present set point;

k. if step j. is true and if said second value is greater than said first value in step h., comparing said present rate of change of set point to said maximum rate of change of said set point;

l. from step k. if said present rate of change of set point is the lesser value, said acceleration of set point remains unchanged and selecting a new present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point;

m. from step k. if said present rate of change of set point is the greater value, setting said rate of change of set point equal to said maximum rate of change of set point, said acceleration of set point to zero, and selecting a new present set point proportional to the sum of said maximum rate of change of set point multiplied by said cycle time, and said previous set point; and

n. if step j. is true and if said first value is greater than said second value in step h., setting said acceleration of set point proportional to the negative of said present rate of change of set point squared divided by the difference between said present set point and said final set point, and said next present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point.

33. The method of claim 31 wherein step i. further includes the steps of:

j. determining if said final set point is less than said present set point;

k. if step j. is true and if said first value is greater than said second value in step h., comparing said present rate of change of set point to the negative of said maximum rate of change of said set point;

l. from step k. if said present rate of change of set point is the greater value, setting said acceleration of set point to the negative of said acceleration of set point and selecting a new present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point;

m. from step k. if said present rate of change of set point is the lesser value, setting said rate of change of set point equal to the negative value of said maximum rate of change of set point, said acceleration of set point to zero, and selecting a new present set point proportional to the sum of the negative value of said maximum rate of change of set point multiplied by said cycle time, and said previous set point; and

n. if step j. is true and said first value is greater than said second value in step h., setting said acceleration of set point proportional to said present rate of change of set point squared divided by the difference between said present set point and said final set point, and said next present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point.

34. A computer-implemented method for controlling the operation of a system by developing a control value, said system having at least one controllable input and at least one measurable variable, said computer-implemented method comprising the steps of:

a. entering a user selected set point for the operation of said system;

b. sequentially approximating a present error as the difference between a present set point and a present value of said measurable variable; and

calculating a present slope in error of said system;

c. sequentially setting a present control value proportional to a sum of a previous control signal, the present error, and the present slope in error; and

d. approximating a response time for said system as the total time of the difference in time between when a difference in said control value is applied to said system and when a difference is noted in said measurable variable.

35. The computer-implemented method of claim 34 further including the step of:

d. approximating a system characteristic that is proportional to a change in said control value to a resultant change in said measurable variable.

36. The computer-implemented method as in claim 34 further including the step of:

e. calculating said present slope of error as the change in set point of said measurable variable from a previous interval to a present interval minus the change in said measurable variable of said system from the same previous interval to the same present interval.

37. The computer-implemented method as in claim 36 further including the steps of:

f. approximating a response time for said system as the total time of the difference in time between when a difference in said control value is applied to said system and when a difference is noted in said measurable variable; and

g. setting said time intervals for control value changes, measurements and calculation proportional to said response time for said system.

38. The computer-implemented method as in claim 36 further including the step of:

g. calculating a control value that is proportional to said present control value plus a change in said control value.

39. The computer-implemented method as in claim 34 wherein the portion of said present error of step b. added as a portion of said present control value in step c. is proportional to a number of time interval cycles occurring within said response time of said system.

40. A computer-implemented method for controlling the operation of a system by developing a control value, said system having at least one controllable input and at least one measurable variable, wherein cycle time is the time necessary to make measurements and preform calculations, said method of developing said control value comprising the steps of:

a. entering a user selected set point for the operation of said system;

b. generating a base cycle time;

c. approximating a system response time for said system as the total time of the difference in time between when a difference in said control value is applied to said controllable input of said system and when a difference is noted in said measurable variable;

d. sequentially approximating a present error as the difference between a present set point and a present value of said measurable variable;

e. calculating a present slope in error of said system;

f. approximating a system characteristic that is proportional to a change in said control signal to a resultant change in said data signal that is representative of a change in said measurable variable;

g. approximating a present change in load by doubling said present slope in error multiplied by said system characteristic; and

h. sequentially setting a present control value proportional to the sum of a previous control value, said present error multiplied by said system characteristic and said cycle time divided by said system response time, and said present change in load.

41. A computer-implemented method for controlling the operation of a system by developing a control value, said system having at least one controllable input and at least one measurable variable, said method of developing said control value comprising the steps of:

a. entering a user selected set point for the operation of said system;

b. generating a base cycle time;

c. presetting or initializing final set point, previous set point;

previous rate of change of set point, maximum rate of change of set point and acceleration of set point, and said measurable value;

d. comparing, during each cycle time, a present set point with said final set point to determine if a change in said present set point is necessary;

e. if as a result of step d. said previous set point equals said final set point, said present rate of change of set point and acceleration of set point are each set to zero, resulting in said set point remaining unchanged and the present set point being set equal said previous set point; and

f. if as a result of step e. said previous set point is less than or greater than said final set point, the values of said present rate of change of set point and acceleration of set point may be adjusted to cause said present set point to change to achieve said final set point.

42. The computer-implemented method of claim 41 further including the step of:

g. limiting the absolute value of said present rate of change of set point at each cycle time to not exceed said preselected maximum rate of change of set point.

43. The computer-implemented method of claim 41 wherein step f. further includes the steps of:

g. estimating a first value of said set point at a next unit of time without a change in said previous set point;

h. estimating a second value of set point where a rate of change of said set point is to be changed for said system to reach said final set point;

i. comparing said first and second values of steps g. and h.; and

j. determining a next present set point from the result of step i.

44. The computer-implemented method of claim 43 wherein step j. further includes the steps of:

k. determining if said final set point is greater than said present set point;

l. if step k. is true and if said second value is greater than said first value in step i., comparing said present rate of change of set point to said maximum rate of change of said set point;

m. from step l. if said present rate of change of set point is the lesser value, said acceleration of set point remains unchanged and selecting a new present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point;

n. from step l. if said present rate of change of set point is the greater value, setting said rate of change of set point equal to said maximum rate of change of set point, said acceleration of set point to zero, and selecting a new present set point proportional to the sum of said maximum rate of change of set point multiplied by said cycle time, and said previous set point; and

o. if step k. is true and if said first value is greater than said second value in step i., setting said acceleration of set point proportional to the negative of said present rate of change of set point squared divided by the difference between said present set point and said final set point, and said next present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point.

45. The computer-implemented method of claim 43 wherein step j. further includes the steps of:

k. determining if said final set point is less than said present set point;

l. if step k. is true and if said first value is greater than said second value in step i., comparing said present rate of change of set point to the negative of said maximum rate of change of said set point;

m. from step l. if said present rate of change of set point is the greater value, setting said acceleration of set point to the negative of said acceleration of set point and selecting a new present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point;

n. from step l. if said present rate of change of set point is the lesser value, setting said rate of change of set point equal to the negative value of said maximum rate of change of set point, said acceleration of set point to zero, and selecting a new present set point proportional to the sum of the negative value of said maximum rate of change of set point multiplied by said cycle time, and said previous set point; and

o. if step k. is true and said first value is greater than said second value in step i., setting said acceleration of set point proportional to said present rate of change of set point squared divided by the difference between said present set point and said final set point, and said next present set point proportional to the sum of said acceleration of set point multiplied by said cycle time squared, said previous rate of change of set point multiplied by said cycle time, and said previous set point.

46. A method for controlling the operation of a system by developing a control signal, said system having at least one controllable input and at least one measurable variable, said system including a first signal conditioner to convert said control signal from a controller to an input form required by said system for said controllable input and, a second signal conditioner to detect and convert said measurable variable to a data signal having a format compatible with said controller, said controller disposed to receive a signal representative of a user selected set point for the operation of said system, said method of developing said control signal comprising the steps of:

a. sequentially approximating a present error as the difference between a present set point and a present value of said measurable variable; and

calculating a present slope in error of said system;

b. sequentially setting said present control signal proportional to a sum of a previous control signal, the present error, and the present slope in error; and

c. approximating a system characteristic that is proportional to a change in said control signal to a resultant change in said data signal that is representative of a change in said measurable variable.

47. The method as in claim 46 further including the step of:

d. multiplying said system characteristic with each of said present error and said present slope in error of said system to convert the units of said error and slope in error to units of said control signal.

48. The method of claim 47 further including the step of:

e. approximating a change in load on said system that is proportional to a present change in error converted to control signal units and a previous change in control signal; and

f. approximating a control signal that is proportional to a sum of said change in load on said system from step e.

49. A computer-implemented method for controlling the operation of a system by developing a control value, said system having at least one controllable input and at least one measurable variable, said computer-implemented method comprising the steps of:

a. entering a user selected set point for the operation of said system;

b. sequentially approximating a present error as the difference between a present set point and a present value of said measurable variable; and

calculating a present slope in error of said system;

c. sequentially setting said present control value proportional to a sum of a previous control signal, the present error, and the present slope in error; and

d. approximating a system characteristic that is proportional to a change in-said control value to a resultant change in said measurable variable.

50. The computer-implemented method as in claim 49 further including the step of:

e. multiplying said system characteristic with each of said present error and said present slope in error of said system to convert the units of said error and slope in error to units of said control value.

51. The computer-implemented method of claim 50 further including the steps of:

f. approximating a change in load on said system that is proportional to a present change in error converted to control signal units and a previous change in control signal; and

g. approximating a control value that is proportional to a sum of said change in load on said system from step f.