Recently, Neural Network (NN) has been an important reasoning method as aforeland of researcher in Artificial Intelligence (Al). However, there are many problems inits application, such as the long time of training ~. bad global-searching ability and lack ofsystematic means in designing NN's structure. Based on deeply studying GeneticA1gorithm(GA).~ Fuzzy Logic Rules(FLR)~ Production Rules~ Radial Basis FunctionNetwork(RBFN), this paper apply the combination of Genetic Algorithm and NeuralNetworl?and the combination of Fuzzy Logic Rules and Neural Network, and RadialBasis Function Network in the object recognition according to the need of "intelligencereconnaissance and data fusion Expert System ". Thus the problems appearing in the usingof NN in the object recognition, for example recognition-rate.~ training-speed, is resolved insome degree.
|