Skip to Main Content
Owing to the problem of control difficulty for the complex system, which has the characteristics of the large inertia, the pure time-delay and the model uncertainty in the industrial processes, a new PID-type fuzzy neural network controller (FNNC) based on Takagi-Sugeno-Kang (TSK) inference is proposed. Real-coded Chaotic Quantum-inspired Genetic Algorithm (RCQGA) is used to optimize the membership function parameters and TSK parameter sets simultaneously with faster convergence speed and more powerful optimizing ability. The pure time-delay effect of the complex object is compensated by a Smith predictor combined with Radial Basis Function (RBF) neural network identifier. The structure and control tactics of the controller are presented and tested by simulations and experiments in the heating furnace system. The proposed algorithm, as confirmed by the results of simulation and experiment compared with the Smith-Fuzzy-PID controller, exhibits good dynamic adjustment, high steady-state accuracy, strong resistant ability to interference and good robustness.