基于误差修正模型的河川年径流量预测研究

Study on prediction of annual runoff in river based on ECM model

对河川径流量预测进行了研究.考虑到年平均水位与年径流量关系密切,误差序列一般不是白噪声序列等因素,用传统的预测模型预测效果不佳.误差修正模型不需要假定误差序列是白噪声序列,只要变量间存在协整关系.利用年平均水位与年径流量的协整关系建立误差修正模型,并加以分析,得到了更好的预测模型.

This paper studied on the prediction of river runoff.The annual runoff was closely related with average water level,and error sequence isn＇t generally white noise sequence.The predictor of traditional forecasting model are not well.Error correction model doesn＇t assume that the error sequence is the white noise sequence,as long as they are co-integrated between the variables.This paper established error correction model by using co-integration of the annual runoff and average water level,and analyzed the results,obtained a better prediction model.