基于遗传算法的天然气消费量预测研究  

Prediction of Natural Gas Consumption Research Based on Genetic Algorithm

作　　者：帅训波[1] 杨遂发[1] 周相广[1] 马书南[2] 

机构地区：[1]中国石油勘探开发研究院廊坊分院地球物理与信息研究所,河北廊坊065007 [2]北京工业大学计算机学院,北京100022

出　　处：《德州学院学报》2009年第2期60-62,76,共4页Journal of Dezhou University

摘　　要：天然气消费量受众多因素影响,应用单一预测方法产生的预测值误差较大.提出了一种天然气消费量优化组合预测模型的构造方法,给出了优化组合预测模型提高预测精度的理论依据,并应用实数编码遗传算法对优化组合模型进行求解.实例计算结果分析表明,优化组合预测模型具有更高的预测精度.Consumption of natural gas is affected by many factors, so there is big error between the prediction and the real consumption for almost every single prediction method. To solve the problem, a construction method of optimization combination prediction model is proposed. Besides, a genetic algorithm based real code is designed to search for optimization parameters values of the combination model. Prediction result of the optimization combination model constructed by this method is better than that with single method, which is proved not only by theories but also by the experiments results.

关 键 词：天然气 消费预测 优化组合 遗传算法 

分 类 号：F224[经济管理—国民经济] TP301[自动化与计算机技术—计算机系统结构]