Adaptive control based on genetic algorithm and fuzzy tuning for unknown systems with time-delay

Considering unknown systems with time-delay, a neural network approach for on-line parameter estimation is presented. The unknown steady-state gain and time-delay of the systems are estimated on-line by Adaline network and then used to modify parameters of Smith predictor in real-time. An adaptive neuron controller based on fuzzy tuning and genetic algorithm is designed in this paper. Fuzzy rules are used to adjust the proportional learning rate, integral learning rate and derivative learning rate of the controller. An improved genetic algorithm is proposed in this paper. This algorithm is an on-line technique, and employed to optimize the gain of controller. This method can be applied for slow time-varying and uncertainty systems with time-delay. Simulation results show that the method is efficient and practical.