关键词: 神经网络 遗传算法 小波分析 智能建模 组件技术

Abstract:

A combined intelligent information processor is developed based on recombining and improving artificial neural networks（ANN） ,wavelet transformation （WT）, and genetic algorithm（GA）. Firstly, mass historical data and field data gathered by multi-sensors on spot are preprocessed using wavelet analyysis, which takes the preprocessed data as the input sample of the neural network model, and the synchronouslygenetic algorithm which has the ability of global optimization is adopted to dynamically modify the network structure and parameters and eliminate the rate tardiness of neural network training a nd relapse into local extremum. This processor can be used for accomplishing complex nonlinear modeling and data mining. Finally, an intelligent modelin g system with good expansibility is established by integrating the combined intelligent information processor with GIS using the component technology, a  nd the integrated scheme is described clearly in this paper. In order to verify the feasibility and validity of the modeling methods, a simulation examp   le is given for the runoff forecasting of an irrigation catchment.

Key words: neural network, genetic algorithm, wavelet transformation, intelligent modeling, component technology