Optimization Strategy of BP Neural Network for Software Reliability Prediction

To improve the accuracy of software reliability prediction, a software reliability prediction model of BP neural network optimized by genetic algorithm is proposed. In the model, the software reliability factor is regarded as the input of BP neural network, and the software reliability accuracy rate is regarded as the output of neural network. Then the parameters of BP neural network are initially regarded as the particles in particle swarm, and the software reliability accuracy rate is regarded as the objective function of particle optimization. The optimal parameters of neural network are obtained through the cooperation between particle swarm and RBF neural network Predict by sex. The simulation results show that compared with the traditional software reliability prediction method, the BP neural network with genetic optimization has higher accuracy and faster convergence speed for software reliability prediction, and solves the problem of parameter optimization of traditional RBF neural network, which is more suitable for software reliability prediction.