Recursive Least Square and Genetic Algorithm Based Tool for PID Controllers Tuning

The objective of this paper is to present a recursive least square and genetic algorithm based tool for PID controller tuning. This tool was developed to be used in PID controllers embedded in PLC - Programmable Logic Controllers. In such software, first the plant is identified and then, based in the plant model, the PID controller parameters are found. The recursive least square is used to do the plant identification, while the genetic algorithm is used to the PID tuning. Initially, it establishes communication with the PLC, and after a disturbance, it acquires the data. In the sequence, the software searches for the new controller parameters and simulates it in order to evaluate its performance. Experimental and simulation results are provided, showing the effectiveness of the tool.