论文部分内容阅读
对西北地区半干旱气候区小麦黄矮病1992—2009年发生、流行情况进行长期监测、分析,选择制约小麦黄矮病发生、流行的23个因素,利用三层人工神经网络可以逼近任意连续函数,对非线性预测系统进行模拟处理的特点,分析所选预测分子,提出一套完整的建立BP人工神经网络模型的方法,并建立陕西省BP神经网络长期预测模型。对1992—2006年数据进行网络训练,利用2007—2009年数据进行测试。结果表明,以发病率为指标,输出结果误差在0.001~0.034之间;以发病级别作为预测结果,模型计算得出的数值与实际病级完全吻合,准确率为100%。说明利用神经网络建立小麦黄矮病预测模型是可行的。
Long-term monitoring and analysis of the occurrence and prevalence of yellow dwarf in wheat in the semi-arid climate region in the semi-arid climate of Northwest China from 1992 to 2009 were conducted. The 23 factors restricting the occurrence and prevalence of wheat yellow-dwarf disease were selected. By using the three-layer artificial neural network, , The nonlinear prediction system is simulated and analyzed. The selected predictors are analyzed and a complete set of methods to establish BP artificial neural network model are proposed. The long-term BP neural network prediction model in Shaanxi Province is also established. Network training on 1992-2006 data, using 2007-2009 data for testing. The results showed that the incidence of error as an indicator, the output error between 0.001 ~ 0.034; the level of disease as a prediction result, the model calculated values exactly match the actual level, the accuracy rate of 100%. It is feasible to use neural network to establish wheat yellow dwarf disease prediction model.