Adaptive neuro-fuzzy inference system identification model for smart control valves with static friction

The study of static friction in control engineering is the subject of many researches due to its impact on degradation of performance of the control loops. Mathematical model of systems with static friction is not straight forward. Precise and proper model of this phenomenon is a key factor in model-based control to mitigate its effect. By increasing number of smart valve in industry, demand for identification of such valves is rising. In these valves, identification of process is limited to control signal (OP) and valve position (MV). By taking advantage of Hammerstein approach, identification is divided in two parts, linear dynamic part and nonlinear static part. In this paper, adaptive neuro-fuzzy inference system (ANFIS) is used for identification of nonlinear static part of the plant. The linear dynamic part can be identified using linear identification methods. Results reveal that ANFIS which integrates both neural networks and fuzzy logic principles and has potential to capture the benefits of both in a single framework can capture well the key model of the systems with smart valves involved in static friction.