Continuously Scheduled Model Parameter Based Adaptive Controller
First Claim
1. An adaptive controller for use in controlling a process, comprising:
- a controller input to receive an process variable input signal from the process;
a controller output to provide a process control signal for use in controlling the process;
a control block coupled between the controller input and the controller output that determines the control signal by implementing process control calculations using the input signal and a set of controller tuning parameters; and
a tuning block that determines new values for the set of controller tuning parameters during operation of the process using a set of stored process model parameter values and a process state variable, wherein the set of stored process model parameter values includes a process model parameter value for a particular process model parameter for each of a plurality of different process operating points defined by the process state variable;
the tuning block including;
a model adaptation routine that determines the stored process model parameter value for the process model parameter at each of the plurality of different process operating points;
a model parameter determination routine that determines a particular process model parameter value for a particular process operating point based on the process state variable value and the set of stored process model parameter values; and
a controller tuning parameter routine that determines the set of controller tuning parameters for the particular process operating point from the determined process model parameter value and a stored tuning rule.
1 Assignment
0 Petitions
Accused Products
Abstract
An adaptive process controller performs continuously scheduled process model parameter interpolation to determine a particular set of process model parameters which are used to develop controller tuning parameters for controller tuning. More particularly, a state-based, adaptive PID controller described herein uses a new technique to determine an appropriate process model to be used to perform adaptive tuning over the various operating regions of the plant, and in particular, uses a process model parameter determination technique that enables continuously scheduled process model parameter update over the various plant operating regions or points. The use of this continuously scheduled process model parameter update method provides for smoother transitions between tuning parameters used in the PID controller during adaptive tuning procedures which are implemented based on changes in the operating region or the operating point of the process, thereby providing for better overall control.
-
Citations
34 Claims
-
1. An adaptive controller for use in controlling a process, comprising:
-
a controller input to receive an process variable input signal from the process; a controller output to provide a process control signal for use in controlling the process; a control block coupled between the controller input and the controller output that determines the control signal by implementing process control calculations using the input signal and a set of controller tuning parameters; and a tuning block that determines new values for the set of controller tuning parameters during operation of the process using a set of stored process model parameter values and a process state variable, wherein the set of stored process model parameter values includes a process model parameter value for a particular process model parameter for each of a plurality of different process operating points defined by the process state variable;
the tuning block including;a model adaptation routine that determines the stored process model parameter value for the process model parameter at each of the plurality of different process operating points; a model parameter determination routine that determines a particular process model parameter value for a particular process operating point based on the process state variable value and the set of stored process model parameter values; and a controller tuning parameter routine that determines the set of controller tuning parameters for the particular process operating point from the determined process model parameter value and a stored tuning rule. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A method of adaptively tuning a process controller used to control a process, the method comprising:
-
storing a set of model parameter values characterizing the operation of the process during operation of the process, each of the model parameter values characterizing the operation of the process at a different process operating point with each of the different process operating points being associated with a different process state variable value; collecting process data during the operation of the process and performing a model characterization procedure based on the collected process data to determine a new model parameter value for one or more process operating points; updating the stored set of model parameter values with the new model parameter value; and tuning the process controller using the stored set of model parameter values, including; determining a current model parameter value for a current process operating point using two or more of the set of stored model parameter values and the process state variable; using the determined current model parameter value and a process model based tuning rule to define a set of controller tuning parameter values; and updating the process controller with the determined set of controller tuning parameter values. - View Dependent Claims (19, 20, 21, 22, 23, 24, 25, 26, 27)
-
-
28. An adaptive process controller system, for implementation on a computer processor to control a process, comprising:
-
a computer memory; a process controller routine stored on the computer memory and executable on the computer processor to implement a control algorithm that determines a process control signal for use in controlling the process based on a process variable input from the process and a set of controller tuning parameters; and a tuning routine stored on the computer memory and executable on the computer processor to determine new values for the set of controller tuning parameters during operation of the process using a set of stored process model parameter values and a process state variable, wherein the set of stored process model parameter values includes a process model parameter value for a particular process model parameter for each of a plurality of different process operating points defined by the process state variable;
the tuning routine including;a model adaptation routine that determines the stored process model parameter value for the process model parameter at each of the plurality of different process operating points, a model parameter determination routine that determines a particular process model parameter value for a particular process operating point based on the process state variable value and the set of stored process model parameter values; and a controller tuning parameter routine that determines the set of controller tuning parameters for the particular process operating point from the determined process model parameter value and a stored tuning rule. - View Dependent Claims (29, 30, 31, 32, 33, 34)
-
Specification