Process for rapidly controlling a process variable without overshoot using a time domain polynomial feedback controller
View Patent ↗A method for controlling a process variable as it approaches a predetermined value (setpoint) so that the setpoint is not exceeded. The method employs a time domain polynomial equation in a feedback configuration and utilizes a controller that acts as an On/Off controller until the process variable approaches setpoint. As the process variable approaches setpoint, the controller acts as a fast responding analog controller thereby “tailoring” a control variable to precisely bring the process variable to the setpoint without exceeding or overshooting the setpoint.
1. A method, as applied in the field of thermal or chemical fluid process control, for controlling a process parameter to a setpoint using a continuous time domain polynomial feedback controller, the method comprising the steps of:
accepting an error signal from an error signal calculator and calculating a continuous type controller output using a time domain polynomial algorithm, which includes an n th order, odd-ordered exponent, in a feedback configuration;
integrating and averaging said error signal over time;
adding said integrated and averaged error signal to said output;
calculating a bias tuning parameter based on user defined criteria;
subtracting said bias tuning parameter from said output; and
using said output to control a final control element;
whereby said process parameter, indicating the state of said fluid process and under the control of said controller output, approaches but does not exceed said setpoint.
2. The method of claim 1 wherein said method further comprises ingredient addition, whereby said controller reduces product variability when used to control ingredient addition into a production batch.
3. The method of claim 1 wherein said algorithm includes a polynomial equation of the form Y=A(x) p −B, wherein Y is a variable expressing said controller output, A is a variable expressing gain, x is a variable expressing said error signal, P is a variable expressing said exponent, and B is a variable expressing said bias tuning parameter.
4. The method of claim 1 wherein said algorithm includes a polynomial equation of the form Y=A(x) p +B(x)−C, wherein Y is a variable expressing said controller output, A is a variable expressing gain, x is a variable expressing said error signal, P is a variable expressing said exponent, B is a variable expressing said bias tuning parameter, and C is a variable expressing an averaged-integrated error signal.
5. The method of claim 1 wherein the output data of said controller is calculated using the polynomial equation Output c =K a (Error) p −K Bias .
6. The method of claim 1 wherein the output data of said controller is calculated using the polynomial equation Output c =K a (Error) p +K b −K Bias .