Soft computing techniques and their applications in intelligent industrial control systems: A survey

Abstract

Soft computing involves a series of methods that are compatible with imprecise information and complex human cognition. In the face of industrial control problems, soft computing techniques show strong intelligence, robustness and cost-effectiveness. This study dedicates to providing a survey on soft computing techniques and their applications in industrial control systems. The methodologies of soft computing are mainly classified in terms of fuzzy logic, neural computing, and genetic algorithms. The challenges surrounding modern industrial control systems are summarized based on the difficulties in information acquisition, the difficulties in modeling control rules, the difficulties in control system optimization, and the requirements for robustness. Then, this study reviews soft-computing-related achievements that have been developed to tackle these challenges. Afterwards, we present a retrospect of practical industrial control applications in the fields including transportation, intelligent machines, process industry as well as energy engineering. Finally, future research directions are discussed from different perspectives. This study demonstrates that soft computing methods can endow industry control processes with many merits, thus having great application potential. It is hoped that this survey can serve as a reference and provide convenience for scholars and practitioners in the fields of industrial control and computer science.