Neural Network Based PI Controller Parameter Calculation on a Boiler Drum Level System

摘要

The controller parameters influence the performance of the closed loop system. So we have to develop a tuning method for obtaining the optimum values of the controller parameters with respect to a particular process. Controller tuning is very much process dependent and any improper selection of the controller settings may lead to instability and affect performance of the closed loop system. Closed loop tuning methods like Ziegler-Nichols method depends on estimation of ultimate gain and ultimate time period. When trying different gains on an unknown process the amplitude of undampened oscillations can become unsafe or on the conversely for low initial gain settings the test can take a long time to reach sustained oscillation condition. This paper proposes a neural network based scheme to estimate ultimate gain and optimum proportional and integral value of PI controller within affordable time limit and safe input range when the parameters change.