基于改进粒子群优化-反向传播神经网络算法的小麦储藏品质预测模型

在使用反向传播神经网络(back propagation neural network,BPNN)预测小麦的储藏品质时,由于其易陷入局部极值且收敛速度慢,导致预测误差较大且稳定性较差,由此提出一种改进粒子群(improved particle swarm optimization,IPSO)算法优化的BPNN预测模型.采用非线性函数动态调整粒子群算法中的惯性权重和学习因子,优化BPNN中的权值参数,进而构建IPSO-BPNN预测模型.为验证该模型的准确性和稳定性,将其与BPNN模型、PSO-BPNN模型进行对比,结果表明:IPSO-BPNN模型预测的均方误差显著降低,有助于提高小麦储藏品质预测的准确性和可靠性.

Research on Prediction Model of Wheat Storage Quality Based onIPSO-BP Neural Network

It is prone to fall into local extremes and has slow convergence speed when using BP neural network to predict the storage quality of wheat, which leads to large prediction errors and poor stability. For this reason, an IPSO-BP neural network prediction model is proposed. A non-linear function is used to dynamically adjust the inertia weights and learning factors of particle swarm optimization algorithm in this paper, so as to optimize the weight parameters in the BP neural network, and then build an improved BP neural network prediction model optimized by the particle swarm optimization algorithm. To verify the accuracy and stability of the model, the BP neural network model and PSO-BP neural network model were taken as reference. The results show that the mean square error of IPSO-BP neural network model prediction is significantly reduced, which helps to improve the accuracy and reliability of wheat storage quality prediction.