An Intelligent Control Method For Nonlinear System Using Fuzzy Neural Network Based On Genetic Algorithms

Research on neural network has been developing rapidly in recent years. Neural network's features display its great capacity in solving highly nonlinear and indefinite control system. It has important significance in all kinds of science areas. This paper applies Fuzzy Neural Network based on Genetic Algorithm to the nonlinear system of power electronic circuits. Because ANN can be used without the specific model of object, and it stores useful information as distributed manner, we usually make use of the topological structure and weights of neural network to realize nonlinear mapping, which make those weights full of meaning, also reserve the training algorithm of ANN. As it is known, we hope to get the universal best answer, so in the processing of weight training, we adopt GA to avoid the defect of BP algorithm, which is very easy to get into the local best answer. Based on a summary of domestic and abroad study, this paper tries to establish fuzzy optimizing neural network real time flood forecasting model, which combines Artificial Neural Network (ANN), Fuzzy Sets theory and Genetic Algorithms (GAs). The framework and research fruits are listed as follows:The historical development of ANN is reviewed from theory and practice. Great emphasis is place on the lacks of ANN, and the structure and algorithm of it. At the same time, the objective and scheme are proposed based on this.The methodologies of fuzzy set, ANN and GAs are studied, and the historical development of Fuzzy Neural Network (FNN) is reviewed from theory and practice as well. Then the thought of using fuzzy technology and GAs to construct FNN is proposed. Great emphasis is place on research methods of FNN based fuzzy set and GAs.Aiming at the fuzziness and complex relations of nonlinear system and their factors, combining excellent knowledge expressing ability of fuzzy mathematics with fine learning ability of ANN, real time control process is implemented using fuzzy optimizing neural network. In addition, the BP algorithm is a kind of algorithm which is in use widely in multiplayer neural network, because the BP algorithm is a kind of gradient descended searching algorithm in essence, it has weaknesses such as slow convergent speed, liable to fell into error function's local extreme value point, insurable to find global extreme value point for multi-modal and non-differential function in larger searching zone, which restrict neural network's application in all fields. So the modification to BP algorithm is necessary. A modified training algorithm based on GAs is proposed; it can improve the convergent speed and precision.No existing referential fornula are ready to calculate parameters, such as number of hidden layer nodes in fuzzy set, neural network and parameters in GAs, and commonly parameters are determined using simulating methods. After analyzing the influnce of parameters on computing results of model, optimal parameters is finally specified. The simulation results show that the topology of neural network has been simplied. Not only the error goal is met with, but also the generalization capability is improved.Finally, the dissertation concludes with a series of vistas of future FNN based on GAs.