基于因子分析法优选河流水体中氮浓度预测方法的研究

定量的河流水体中氮浓度预测方法有很多种,如何优选出预测精度较高的方法一直是学术界多年来致力于研究的重点。本研究采用因子分析法对预测方法的精度评价指标进行分析,并建立了预测方法精度的评价模型,对回归分析法、神经网络法、灰色系统法和增长率统计法4种水体氮浓度预测方法进行综合评估,优选出精度较高的河流水体氮浓度预测模型——BP神经网络预测模型。结果表明,此评估模型对类似研究具有一定的参考价值,能为选择出合适的河流水体氮浓度预测方法提供依据。

Study on selecting forecast model of river water nitrogen concentration using factor analysis

There are many quantitative forecast models to predict river water nitrogen concentration, it is a key issue for research in academia field at present how to select a forecast model with higher estimate accuracy. The paper analyzes evaluation indexes of model accuracy by using factor analysis method, sets up an evaluation model of forecast accuracy. Then it predicts river water nitrogen concentration by using regression analysis, neural network method, grey system and growth rate statistic method, and carries on a comprehensive evaluation to the four kinds of forecast models by using evaluation model based on factor analysis method. It shows that BP neural network is a good forecast method to accurately predict river water nitrogen content than other three kinds. The results indicate that the evaluation model based on factor analysis method has a reference value in the similar studies, and it can provide evidence for selecting the suitable forecast models.