Adaptive feed forward method for temperature control

• US 9,582,754 B1
• Filed: 05/17/2016
• Issued: 02/28/2017
• Est. Priority Date: 05/17/2016
• Status: Active Grant

• Current Assignee: Collins Research, Inc.
• Original Assignee: Collins Research, Inc.
• Inventors: Collins, Roger
• Primary Examiner(s): Holmes, Michael B
• Application Number:
  US15/156,383
• Time in Patent Office: 287 Days
• Field of Search: 706/25
• US Class Current: 1/1
• CPC Class Codes: G05B 13/0265 the criterion being a learning criterion
• CPC Class Codes: G05D 23/00 Control of temperature
• CPC Class Codes: G05D 23/1917 using digital means
• CPC Class Codes: G05D 23/1919 characterised by the type of controller

Abstract

A method of temperature control for maintaining actual temperature at a set point uses a learning period from which to acquire parameters that are used to control temperature. The method defines a learning period as one complete oscillation of the actual temperature about the set temperature. During such an oscillation period, the actual temperature will have been exclusively above and below the set temperature for one segment each. The invention provides a method for adjusting a variable in order to maintain a process measurement at a predetermined constant value, the method defining a sample period, wherein the sample period is a period of time represented by a complete oscillation of the process measurement about the predetermined constant value and wherein the complete oscillation includes a first time period during which the adaptable measurement is continuously above the predetermined constant value and a second time period during which the adaptable measurement is continuously below the predetermined value and wherein the first time period and second time period occur substantially sequentially.

Claims

1. A method of temperature control for maintaining actual temperature at a set point comprising the steps of:

setting a temperature controller at the set point;

taking temperature readings of the actual temperature at discrete time periods over the course of a learning period;

determining the sum of the actual temperatures taken within each of the discrete time periods over the course of the learning period;

determining output within each discrete time period over the course of the learning period;

determining the sum of the output within each discrete time period over the course of the learning period;

determining a learned feed forward parameter, wherein the learned feed forward parameter is a quotient determined by a proportion of a first part, wherein the first part is the product of the set point and the sum of the output over the course of the learning period, and a second part, wherein the second part is the sum of the actual temperature values over the course of the learning period; and

applying the learned feed forward parameter to modify the output in order to maintain the actual temperature at the set point.

2. The method of claim 1, wherein each of the discrete time period is substantially a one second period of time.

3. The method of claim 1, wherein the learning period includes a first time period in which the actual temperature is continuously below the set point and a second time period in which the actual temperature is continuously above the set point and wherein the first time period and the second time period occur sequentially.

4. The method of claim 1, wherein the learning period includes a first time period in which the actual temperature is continuously above the set point and a second time period in which actual temperature is continuously below the set point and wherein the first time period and the second time period occur sequentially.

5. The method of claim 1 wherein the learning period includes at least one first time period wherein the actual temperature is continuously above the set point and at least one second time period wherein the actual temperature is continuously below the set point.

6. The method of claim 1, further comprising a plurality of learning periods.

7. The method of claim 1, wherein a feed forward parameter is initially set to a factory default setting.

8. The method of claim 1, wherein the temperature controller is a PID temperature controller.

9. A system for calculating a learned feed forward parameter to apply to an output value in order to regulate temperature by maintaining an actual temperature at a set point, the system comprising:

a temperature controller configured at a set point;

a temperature reader configured to measure the actual temperature at fixed intervals;

a blower configured to operate during each of the fixed intervals to provide an output at each of the fixed intervals, wherein the output is defined as a blower duty cycle; and

a central processing unit configured to calculate the learned feed forward parameter for regulating the blower in order to control the output for approximating the actual temperature to the set temperature, wherein the learned feed forward parameter is the output that the temperature controller would produce if there were no error and no history of error, and wherein the learned feed forward parameter is a quotient determined by a proportion of a first part, wherein the first part is the product of the set point and the sum of the output over the course of a learning period, and a second part, wherein the second part is the sum of the actual temperature over the course of the learning period.

10. The system of claim 9, wherein the central processing unit applies the learned feed forward parameter to the output value to control the blower duty cycle in order to maintain the actual temperature substantially at the set point.

11. The system of claim 9, wherein the error value is the difference between the actual temperature and the set point.

12. The system of claim 9, wherein the central processing unit is included in the temperature controller.

13. The system of claim 9, wherein the temperature controller is a proportional integral derivative temperature controller for further calculating a proportional integral derivative feed forward parameter.

14. The system of claim 9, wherein the learning period includes a first time period in which the actual temperature is continuously below the set point and a second time period in which actual temperature is continuously above the set point and wherein the first time period and the second time period occur substantially sequentially.

15. The system of claim 9, wherein the learning period includes a first time period in which the actual temperature is continuously above the set point and a second time period in which the actual temperature is continuously below the set point and wherein the first time period and the second time period occur substantially sequentially.

16. The system of claim 9, wherein the actual temperature and blower output are each calculated once per second over the course of the learning period.

17. The system of claim 13, wherein the learned feed forward parameter is capped at a maximum of two hundred percent of the proportional integral derivative feed forward parameter.

18. The system of claim 13, wherein the learned feed forward parameter will not fall below a minimum of fifty percent of the proportional integral derivative feed forward parameter.

19. A temperature controller for maintaining an actual temperature at a set point comprising a central processing unit configured to calculate a learned feed forward parameter, wherein the learned feed forward parameter is a quotient determined by a proportion of a first part, wherein the first part is the product of the set point and the sum of the output over the course of a learning period, and a second part, wherein the second part is the sum of the actual temperature over the course of a learning period.

20. The temperature controller of claim 19, wherein the temperature controller is a proportional integral derivative temperature controller.

21. The temperature controller of claim 19, wherein the learning period is a time period within which the actual temperature is continuously above the set point and a second time period in which actual temperature is continuously below the set point and wherein the first time period and the second time period occur substantially sequentially.

22. The temperature controller of claim 20, wherein the one complete oscillation is a period within which the actual temperature is continuously below the set point and a second time period in which process value is continuously above the set point and wherein the first time period and the second time period occur substantially sequentially.

23. A method for adjusting a variable in order to maintain an adaptable measurement at a predetermined constant value, the method comprising:

defining a learning period, wherein the learning period is a period of time represented by a complete oscillation of the adaptable measurement about the predetermined constant value and wherein the complete oscillation includes a first time period during which the adaptable measurement is continuously above the predetermined constant value and a second time period during which the adaptable measurement is continuously below the predetermined value and wherein the first time period and second time period occur substantially sequentially.