Advanced adaptive neuro-fuzzy inference system controller for optimizing pH neutralization process control

Abstract

The heavy reliance of modern industries on chemical processes to facilitate the mass production of cosmetics, beverages, food products, and pharmaceuticals has in turn contributed to the heightened significance of pH value regulation that supports product quality assurance. However, the process of pH control is difficult due to its highly sensitive, dynamic, and nonlinear nature. The conventional control approaches like proportional integral derivative (PID) and proportional integral (PI) controller are inept at handling the complex process of pH control. Thereby, in this work adaptive neuro-fuzzy inference system (ANFIS), which combines the accuracy of fuzzy inference system (FIS) and learning capability of adaptive neural network (ANN) is applied for pH process regulation. Moreover, the controller operation is improved further with the application of chicken swarm optimization (CSO) for tuning its input parameters. The primary goal is to accomplish effective load regulation and appropriate set-point tracking using smoother control signal. According to the derived simulation outcomes, it is observed that both the industrial and standard structure of the proposed chicken swarm (CS)-ANFIS controller outperforms other existing control techniques with better disturbance rejection, set-point tracking and excellent sensitivity to change in model parameters.