Fuzzy Neural Network Model Algorithm And Application,

Fuzzy neural network is one most promising field during artificial intelligence's three important fields at present. It exerts great action at fuzzy control, fuzzy decision-making, expert system, model identification and so on at last ten years, its theory, model, algorithm and application technology is important research problem in computer field and it receives high regard because of its advantage.We summarize fuzzy neural network's research actuality, development tide and application field in the article, expound basic conception of fuzzy logic system, artificial neural network, and fuzzy neural network; set up a normal fuzzy neural network model and study algorithm aim at actual problem, the node number of fuzzy layer, normal layer and rule layer is computable if the model has assured input and output pattern and fuzzy layer's subject function, and the model has good applicability to expert express and pattern identification. The normal fuzzy neural net study algorithm has shortcomings of slow convergence speed and easy local particle when study stylebooks are very big and different stylebooks' character difference is very small. Aim at above shortcomings, we use rough set theory and genetic algorithm to optimize study algorithm, and realize pattern character and stylebooks' automatic filter. Optimized network has quick convergence speed, steady performance and good extend capability.According to the normal fuzzy neural network's research production, and connect the scientific research project of the second factory in DaQing oil-field, we develop the project of water drive oil development effect analysis system, realize water drive oil development effect's automatic evaluation and automatic production of oil-well fracture scheme.