The Research And Application Of Neural Networks In Power Investment

Artificial Neural Network (ANN) is a new intelligent information processing theory developed during the process of imitating brain to solve problem. The processing ability is similar to the brain's learning, recognizing and remembering by imitating and realizing abstract thing and associative memory. Because it has features of high non-linearity, parallel processing, adaptability and self-organization, it has been applied in many fields.With the deep process of power reform the condition that the projects of power investment face is complex much more. In the new market condition with competition, the projects of power investment face up with many uncertain factors, such as price, lending rate, the macroscopically policy of the government and require of environment. The evaluation of project must be emphasized to ensure the security and profitability.The application of ANN in power investment is studied in the paper. Firstly, the basic knowledge about the ANN is introduced. The characteristic of ANN is introduced in detail. So do with the BP, the RN and the ART. Secondly, the paper analyses the characteristic of the Recurrent Net with Fuzzy Optimum. By adding additional and internally by-pass of feedback into network model, the ability to process dynamic information is improved.The paper simulates the model that is applied to evaluate a project by using Matlab software. The example shows that the calculating result of this model is as the same as the practice and the result of other models such as the BP neural network model of fuzzy optimization, the RNN model and the grey decision model.