Analysis of Adaptive Neuro-Fuzzy Inference System in Different Environmental Conditions

Regardless of vast advances in field of control systems engineering, PID scheme still remains the most common algorithm in industrial use today. A Fuzzy Logic controllers (FLC) with an efficient understanding and small rule base that can be simply implemented in existing industrial controllers. In an industrial setting FLC is much more capable than the existing temperature controller for flowing water type applications. This includes compensating for thermo mass changes in system, dealing with unknown and variable delays, operating at different temperature setpoint without retuning etc. Also predicting the results by simulation would hide actual disturbances coming into picture during implementation. Hence this paper makes a differential study on various weather conditions and ANFIS designed controller analysis is presented with the help of identical industrial lab model. Once designed ANFIS controller is found to be suitable for concern weather condition (season) as compared to other. The graphical representation and mean error values depicts the uncertainties of ANFIS controller over wide span of time.