Research On Optimization Design Of Electric Apparatus Using Chaotic Neural Networks

With the increase of voltage level of high voltage transmission system, miniaturization and complexity of the electric apparatus have brought forward higher requirement for the insulation performance of the electrical equipment. Accordingly, the structural optimization plays a very important role for investigating and developing the electric apparatus. Actually, there are many optimized variables for the electric apparatus. However, for the traditional optimizing methods, it is very difficult to solve the intricate computation for the optimum mathematical model with higher efficiency and precision. Aiming at these problems, the intelligent optimization approaches have widely applied for designing the electric apparatus.Because the Chaotic Neural Network possess stronger nonlinear searching ability and higher searching efficiency. The Neural Network and chaotic principle has been combined and introduced to optimize the electric apparatus. Moreover, the combination of the Chaotic Neural Network with modern optimization methods have been applied for designing the high voltage electric apparatus for realizing the optimization with higher approaching capability and efficiency. For the reason that the traditional Artificial Neural Network can not satisfy the requirements of the optimization design of electric apparatus with multi-variables and complex energy function. A novel Chaotic Neural Network based on the inverse problem solution of the electric field has been introduced. The whole optimization procedure of Chaotic Neural Network takes the combination of temporary chaotic dramatic and gradient descent dramatic in the optimum searching process into considerations. In optimizing, the searching direction is determined by the optimization programming. And gradually reducing the searching space, thus the efficiency of the optimization of the electric apparatus can be improved.The generation of chaotic variables is an important part in the Chaotic NeuralNetwork. In this thesis, the generating principle and process have been discussed, and the simulation results clearly illustrate the influence of the main network parameters on the chaotic neural network.Finally, the Chaotic Neural Network is employed for solving the complexity nonlinear oscillating mathematical function and searching the optimum. Moreover, the optimization design of the contact for the 220kV SF6 circuit breaker and the 40.5kV vacuum interrupter has been successfully performed. From the optimization results, it can be found that the feasibility and validity of the proposed Chaotic Neural Network have been verified. And the basis of numerical computation and optimization for electric apparatus has been provided.