基于BP神经网络的重要水库水源地风险预测

Risk Prediction of Important Reservoir Water Source Based on BP Neural Network

水库是饮用水源的重要载体。近年来,社会经济急剧发展,大量污染物的排放严重威胁了水库饮用水源的健康。为保障人民群众饮水安全,准确识别水库型饮用水源地健康风险程度,基于模糊综合评价和BP神经网络建立了水库型饮用水源健康风险预测模型,构建了一个具有多层次多指标特性的饮用水源健康评价指标体系,并结合珊溪(赵山渡)库区的实例,对该库区饮用水源进行了健康风险预测。结果表明:该库区的水源健康等级为健康,与实际调研情况相符。神经网络模型可以用于对饮用水源健康风险预警,对水源地相关管理部门的水源健康评价工作具有一定的参考意义。

Reservoir is an important carrier of drinking water source.In recent years,with the rapid development of social economy,the discharge of a large number of pollutants has seriously threatened the health of drinking water sources in reservoirs.In order to ensure the security of the people drinking water and evaluate drinking water health accurately,based on fuzzy comprehensive evaluation and BP neural network,a health risk prediction model for drinking water source in reservoir is established in this paper,and a multi-level and multi-index system for health evaluation of drinking water source was established.Combing with the example of Shanxi(Zhaoshandu)reservoir area,this paper predicts the health risk of the drinking water source.The results show that the health level of the water source in the reservoir area is healthy,which is consistent with the actual investigation.BP neural network model can accurately predict the risk of water source safety and has certain reference significance for the assessment of water source health of relevant management departments of water source.