Research On Neuron Control With Tuning Gain

The neuro-model-free control is one of important development branches in intelligent control. The design of the neuron control systems and tuning the neuron controller parameters are studied in this thesis. The simulation experiments of applying the proposed methods are made. The main content of this thesis includes the following:1. The principle of neuron control, fuzzy control and grey predictive control is introduced.2. Based on the idea of tuning neuron controller parameter on-line by fuzzy logic, two neuron control methods with fuzzy tuning gain are proposed by using the table method or the formulation method, respectively. To the steel casting processes and the pH processes with nonlinearities and uncertainties, the neuron control systems are designed, the simulation results show that good performances are reached by using the proposed control methods.3. From the point of promoting response speed and restraining overshoot, the grey predictor is used to design the neuron control systems. The improved neuron control method with fuzzy-tuning gain is proposed. Apply this neuron control method, the experiments of controlling the continuous steel casting processes and the plants with time-delay are made. The simulation results show that satisfactory performances are obtained.