Application of GA_BP to the Matte Grade Model Based on Neural Network

The advantages and disadvantages of Back Propagation Algorithm (BP) and Genetic Algorithm (GA) are analysed, combination of the two algorithms forms GA-BP.Compared the training and testing results of the matte grade neural network model with the GA-BP and BP respectively indicates that the GA-BP is superior to BP, and it can search answer quickly in overall range.The training results also explains that nerual network technique is feasible and effective to the modeling for copper flash smelting process.