由于剥叶性能直接影响到甘蔗收获机械的收割性能,为了客观有效地对剥叶性能进行预测,提出了BP神经网络预测方法.针对剥叶元件性能的特点,采用正交试验法构造训练样本,以保证网络具有较高的泛化能力,同时对该训练样本建立了回归分析模型,以检验BP网络模型的输出精度.在此基础上,利用已建立的神经网络预测模型对影响剥叶性能的各因素取值的不同组合进行仿真分析,以确定各因素取值的最优组合.结果表明,BP神经网络的预测模型比回归模型具有更高的输出精度,进行剥叶元件的性能预测与优化是可行且有效的. In order to forecast the stripping performance objectively and effectively, a BP neural network prediction method is proposed.Because of the characteristics of stripping device, the orthogonal test is used to construct the training samples, In order to ensure the network has a high generalization ability, and established a regression analysis model of the training samples to test the output accuracy of the BP network model.On the basis of this, using the established neural network prediction model of the impact of stripping performance The results show that the prediction model of BP neural network has higher output precision than the regression model, and the performance prediction and optimization of the stripping elements are carried out Is feasible and effective.