Human Simulated Intelligent Fuzzy Control For Time-Varying Processes With Large Time Delay

Time-delay, which is frequently a source of instability and performance deterioration, occurs in many industrial processes. Therefore, controller synthesis for time-delay processes has been one of the most challenging issues. Also, in practice, these processes almost present some time-varying parameters, environmental noise and other disturbances that enhance the difficulties of controller design. In view of the forgoing factors, two kinds of control schemes are proposed for time-varying processes with large time delay: one is Single-Neuron-PID-based Smith predictor, which is proved to have the structure of internal model controller, and the other is Human-Simulated Intelligent Fuzzy Control. This dissertation is organized as follows.First, the characteristic of control issue of time-varying processes with large time delay is investigated. Although Smith predictor can effectively remove the time delay from the closed-loop characteristic equation, all the factors mentioned above may cause the performance deteriorate considerably. Hence, then, a novel single neuron internal model controller is proposed to improve the robustness of Smith predictor. The improved intelligent neuron PID is used as the main controller while Smith predictor is used as the compensator. The learning rule of single neuron adaptive PID is improved and the proportional coefficient can be self-tuning online according to human-simulated intelligent technique. Then the proposed scheme is proved to be equal to the internal model controller and some simulation examples are included to show that the proposed scheme has the advantages of small overshoot, good...