Bio-inspired Hybrid Algorithm for Air Temperature Process Control

D. Pamela and K. Gearad Joe Nigel

The essence of modernizing industrial and commercial systems requires blending of computing and control of process into various stages of machine operations and data process to reduce cost. The design of a controller with optimum control values is the major criteria that decides the controller performance. The controller design can vary from a wide range of conventional models to sophisticated adaptive control models. The thrust of controller design is the estimation of controller parameter. The optimum control parameter value gives the best controller performance. Application of soft computing techniques in estimating the controller parameter proves to be rewarding. Genetic Algorithms (GA‟s) are a novel search method that mimics the process of natural evolution. The genetic algorithm has no prior knowledge of the accurate solution and hence depends completely on the responses from its environment to arrive at the optimal solution. In this chapter, the application of advanced and proven methodology of parameter estimation using Genetic Algorithm has been discussed. The process taken for study is an air temperature control process used in withering tea leaves. The best tea manufacturing process is usually initiated by proper plucking of shoot, handling of leaf, Withering, Fermenting and Blending. Of all these steps, withering is the first and foremost step and it is the most essential process that decided the quality of the end product. The withering has to be done under controlled temperature so the leaves are withered properly. Hence it is vital to have a perfectly tuned controller to control the air temperature. The following sections will explain in detail how GA has been used to tune a adaptive controller called Adaptive Smith Predictor Controller, to get the best performance of the controller.

Volume 12 | 04-Special Issue

Pages: 1762-1768

DOI: 10.5373/JARDCS/V12SP4/20201659