Special Performance Forecast for Shot Put Based on GA-BP

Abstract

by using the characteristics of global optimization of GA and local optimization of BP neural network, the calculation accuracy and convergence rate of the traditional BP neural network are improved. Then a special performance forecasting model for shot put based on GA-BP is established. The numerical results show that the mean square errors, the mean absolute value of calculation and prediction are diminished. Therefore GA-BP is suitable to solve nonlinear problems such as prediction of special performance, and has high accurateness and application value.