摘要
城市需水量预测对合理分配水资源起着重要作用,但城市需水量本身的影响因素众多,其预测是典型的不确定性问题。集对分析是处理不确定问题的新方法,它通过联系度展示了研究对象间关系的详细结构,对处理不确定性问题十分有效。采用集对分析相似预测法,结合集对分析同异反模式中的"择近原则"和相似分析法中相似的历史原因产生相似结果的原则,在预测过程中将需水量时间序列作为需水量的影响因子,利用一定时长的历史数据来建立基于集对分析原理的相似预测法模型。通过对克拉玛依市三坪地区需水量的预测发现,集对分析相似预测法不但计算简单,而且精度优于传统的ARIMA模型。
The forecast of urban water demanded was very important when water resources needed to be distributed reasonably. There were many factors which could affect the results when we predicted the urban water demanded. It was a classical uncertainty problem, and at the same time, Set Pair Analysis was a new method to solve the uncertainty problems. It could express the detailed structure between studied variables with the connection degree So Set Pair Analysis was always effective when we tried to work out the uncertainty problems. Using the Similar Prediction Method based on Set Pair Analysis to forecast the urban water demanded,with the principle of ＂picking up the nearest one＂in Set Pair Analysis and ＂similar reasons leading to similar results＂in the Similar Prediction Method. This paper treated the time series as the factors and buih the model based on the Similar Prediction Method and Set Pair Analysis with a number of data which was recorded for some time. At the same time, this paper used the data of water demanded in Sanping, Karamay of Xinjiang to test the model based on the Similar Prediction Method and Set Pair Analysis, which showed the method was simple and had a higher accuracy compared with the traditional ARIMA model.
出处
《资源开发与市场》
CAS
CSSCI
2017年第4期408-410,441,共4页
Resource Development & Market
关键词
集对分析原理
相似预测法
不确定性问题
预测
Set Pair Analysis
similar prediction method
uncertainty problems
forecast
