Research On The Algorithms Of Fuzzy-neuro Model-free Control System Design For Plants With Uncertainties

The designs and applications of fuzzy-neuro model-free control systems for plants with big uncertainties are discussed in this thesis. Series fuzzy neuron control methods are proposed and the simulation experiments with the examples of some industrial plants are made. The main contents of this thesis are as follows:1. A survey of the development of neural networks , the neuro-control systems andfuzzy control system is summarized.2. The basic theories of neuron model-free control are introduced, which includes the neuron model for control, the learning strategy and the neuron control method.3. To multi-model plants, the fuzzy integral compound control method with a self-tuning parameter is proposed. The control simulation tests of the hydraulic turbine generator are made and the results show that the proposed controller has strong robustness and satisfied performances .4. To nonlinear plants with uncertainties,the model-free control method with two neurons is proposed .Applied to a pH neutralization process with heavy nonlinear character, the experiments are made to show the effectivenes and robustness of the proposed method5. To plants with big time delay, the neuron PID control method with fuzzy-tuninggain is proposed. Being used to control the water temperature process, the simulation tests results prove the validity of the proposed method.