基于灰色关联分析的辽宁省粮食产量预测模型

Grain Yield Prediction Model of Liaoning Province Based on Grey Relational Analysis

利用辽宁省1985-2008年粮食生产相关影响因子指标统计数据,采用灰色关联分析与人工神经网络相结合的方法进行粮食产量预测。通过灰色关联度分析对8个指标进行定量分析,依据关联度的大小确定了机耕面积、有效灌溉面积、农业机械总动力、化肥施用量、粮食作物占总播种面积比重和农林牧渔业从业人员为影响辽宁省粮食生产的主要影响因素。以这些主要影响因素作为模型输入,建立了基于灰色关联人工神经网络的辽宁省粮食产量预测模型。预测结果表明,最大预测误差为2.09%,平均误差仅为0.89,表明该模型具有较高的预测精度和稳定性,为粮食产量预测提供了一条新的途径。

Statistic data on the impact factors related to food production of Liaoning Province from 1985 to 2008 were used to forecast the grain yield by the combined method of gray relational analysis and artificial neural network.8 indicators were quantified by gray correlation analysis.The following 6 indicators,machinery farming area,the effective irrigation area,the total power of agricultural machinery,chemical fertilizer,the proportion of food crops on the total sown area,the agriculture,forestry,animal husbandry and fishery employees,were determined as the main factors affecting food production of Liaoning Province,based on their correlation degree.And these 6 indicators were used as the model input.A model was established based on grey neural network model for grain production forecast in Liaoning Province.The predicted results showed that the maximum prediction error was 2.09%,the average error was only 0.89,this indicated that the model has high prediction accuracy and stability,and it could be used as a new way for grain yield prediction.