Electric heating – Heating devices – With power supply and voltage or current regulation or...
Reexamination Certificate
2002-08-09
2004-08-24
Hoang, Tu (Department: 3742)
Electric heating
Heating devices
With power supply and voltage or current regulation or...
C219S490000
Reexamination Certificate
active
06781096
ABSTRACT:
TECHNICAL FIELD
The invention relates generally to a method and apparatus for implementing a discrete-time control system, and more specifically to an apparatus and method for implementing a pattern-based control system implemented via matrix chains.
BACKGROUND
Almost all control systems require that they be tuned in light of the system in which they are to be deployed. To illustrate, consider the most frequently used control system, the proportional-integral-derivative controller (“PID controller”). A PID controller generates a control variable based upon the difference between a user-defined set point and a measured process variable. This difference is referred to as “the error.” The control variable is typically converted into a signal that is used to drive a control device that influences the process variable.
A PID controller calculates its control variable by summing together three quantities: (1) the error multiplied by a first gain factor, K
p
; (2) the integral of the error multiplied by a second gain factor, K
i
; and (3) the derivative of the error multiplied by a third gain factor, K
d
. The process of selecting the gain factors, K
p
, K
i
, and K
d
, (referred to as “tuning” a PID controller) is a matter that influences how well the control system behaves. If a PID controller is not tuned properly, the process variable may be slow to arrive at the user-defined set point. Additionally, an untuned PID controller may produce undesirable results such as causing the process variable to oscillate excessively about the set point, causing the process variable to exhibit a steady-state error, or causing the process variable to become unstable.
The process of optimally tuning a PID controller is mathematically complex, and requires a great deal of field testing to ensure that the chosen gain factors cause the control system to function properly. In short, the process is expensive and time consuming.
The above-discussed problem regarding the need to tune a control system to function properly with the system in which it is deployed is common to almost all forms of control systems. Thus, as is made evident by the foregoing discussion, a need exists for a generic, self-tuning control system that can be deployed in wide variety of systems. A desirable embodiment of such a control system will be capable of learning to tune itself based upon its interaction with the system it is controlling.
SUMMARY OF THE INVENTION
Against this backdrop, the present invention has been developed. A method of controlling a system having a plurality of state variables, so as to drive a process variable to a set point, may be arrived at by execution of the following steps. Initially, a matrix chain comprised of a plurality of matrices, including a first matrix, at least one intermediate matrix, and a last matrix is established. Each matrix generates an output based upon a plurality of inputs. The inputs for the first matrix are state variables. For the at least one intermediate matrix, one input is the output from another matrix and another input is either a state variable or the output from another matrix. For the last matrix, at least one input is the output from another matrix, and the output of the last matrix is used to drive a control device that influences the process variable. The output is further used as an input to either the first matrix or one of the at least one intermediate matrices. The matrix chain initially exhibits a systemic degree of reflectivity that is a multiple of 360°. Throughout the duration of controlling the system, the state variables are input to the set of matrices. Finally, throughout the duration of controlling the system, the control device is driven with the output of the last matrix.
According to another embodiment of the invention, a processor is programmed to execute the above-described steps.
According to yet another embodiment of the invention, a method of controlling a system having a plurality of state variables, so as to drive a process variable to a set point, may be achieved by detecting the state variables of the system. Thereafter, the set of state variables are input to a matrix chain. The matrix chain is comprised of a plurality of matrices in which an output from one matrix in the chain is an input to a second matrix in the chain. At least one of the matrices initially exhibits symmetry. At least one matrix feeds its output to a rearward matrix of the matrix chain. A control device that influences the process variable is driven with the output of a matrix.
According to yet another embodiment of the invention, a processor is programmed to execute the above-described steps.
According to yet another embodiment of the invention, an application specific integrated circuit is arranged and configured to effectuate the above-described steps.
REFERENCES:
patent: 4740664 (1988-04-01), Payne et al.
patent: 5079407 (1992-01-01), Baker
patent: 5870001 (1999-02-01), Osterling et al.
patent: 6384384 (2002-05-01), Connolly et al.
patent: 6622101 (2003-09-01), Oechsner et al.
Stephen Wolfram, “A New Kind of Science,” Wolfram Media, Inc., 2002, Chapter 9, pp. 473, 476, 526, Notes for Chapter 2, p. 869.
Hoang Tu
Matrix Innovations, LLC
Merchant & Gould P.C.
LandOfFree
Apparatus and method for pattern-based control of a system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for pattern-based control of a system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for pattern-based control of a system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3357285