【摘要】 对误差反向传播(error back propagation,简称BP)人工神经网络在水质评价中应用的原理进行了分析,并将其应用于地下水质量评价。首先,利用MATLAB7的神经网络工具箱,根据GB/T 14848-93《地下水质量标准》,构建出10-11-5三层结构的BP人工神经网络模型,并对某油田区的地下水水质进行了评价,进而将评价结果与内梅罗指数法的评价结果进行了对比分析。结果表明:BP人工神经网络模型能更真实地反映研究区地下水质量状况,评价结果客观、准确、合理,将其应用于地下水质量评价时适应性更强,计算简便,在该领域有一定的实用价值和应用前景。更多还原

【Abstract】 In this paper,the error backpropagation(referred to as BP) artificial neural network principle was analyzed and applied to groundwater quality evaluation.Based on the neural network toolbox of MATLAB7,a 10-11-5 three-tier structure of BP neural network model was created to evaluate oilfield groundwater quality.The evaluation results were compared with that of Nemerow index method.The results reflect the reality of the groundwater situation of the studied oilfield and indicate that the BP neural network for evaluation of groundwater quality is more adaptable and convenient.This method can be able to provide objective,accurate and reasonable results and true quality of groundwater,which can be widely applied in the field of groundwater quality evaluation.更多还原