能源消费量模拟分析和预测的神经网络方法  

Neural Network Method for Simulation and Prediction of Energy Consumption

作　　者：彭建良[1] 李新建[1] 王斌[1] 陈宝书[1] 

机构地区：[1]中国矿业大学工商管理学院

出　　处：《系统工程理论与实践》1998年第7期76-83,共8页Systems Engineering-Theory & Practice

摘　　要：基于我国１９７８～１９９４年间能源消费量的实际数据，采用人工神经网络方法对影响能源消费量的主要因素进行了模拟和分析。然后，采用相关因素预测法和组合预测法预测了能源消费量。经实际样本检验表明：组合预测法的预测精度较高。最后，应用组合预测法预测了我国１９９５～２０００年的能源消费量。Based on the energy consumption data of 1978-1994,the simulation and analysis for its main influencing factors is made by the use of neural network method.Then both the relevant factor procedure and multimethod combined procedure are used for prediction of energy consumption. The latter is tested to have better accuracy, proved by the sample of actual data.So finally the multi method combined procedure is applied to make prediction of energy consumption for the year 1995-2000.

关 键 词：能源消费量 人工神经网络 预测 

分 类 号：TK012[动力工程及工程热物理]