多模型自适应综合预报方法及其应用  

Multi-model adaptive comprehensive prediction method and its application

作　　者：刘劲松[1] 

机构地区：[1]黑龙江省信息中心,哈尔滨150030

出　　处：《东北农业大学学报》2009年第8期113-116,共4页Journal of Northeast Agricultural University

基　　金：黑龙江省农业厅合作项目

摘　　要：文章依据粮食产量演变规律,提出一种多模型自适应综合预报方法。预报系统由几个并行的自适应预报器组成,最终结果由准确度判别器择优决定。特殊情况,利用专家经验来修正。该法适用于系统具有阶跃性结构变化和时变参数的动态过程。在某省粮食产量预报中,应用此法获得了良好的效果。Based on the evolutionlaw of grain production yield to set up a multi-model adaptive comprehensive prediction methods. The prediction system is consisted of some parallel adaptive pre- dictors, and the final results are decided by checker. In special circumstances, the results need to fix by experts. This method applies to the system with a step structure change and dynamic process of time- varying parameters. The apphcation of this method obtained good results forecasting grain production yield of Heilongjiang Province.

关 键 词：粮食产量 时变参数 模型结构 自适应预报 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]