针对地表水环境质量问题,运用人工神经网络理论和方法,建立地表水环境质量评价的BP人工神经网络模型。并将训练样本进行归一化处理,同时利用RAND函数对训练样本进行插值保证神经网络充分学习。通过实例进行评价分析,说明用BP人工神经网络方法评价地表水环境质量是可行的。该模型具有很强的学习、联想和容错功能,其分析结果和过程都接近人脑的思维过程和分析方法,使得地表水环境质量评价结果的精度大大提高。
According to artificial neural network theory and method,a BP artificial neural network model for surface water environment quality assessment was built.The training samples were normalized.The RAND function was used to construct enough training samples in order to keep the network full learning.The result shows that the application of BP neural network in surface water environment quality assessment is feasible.This model possesses strong functions of study,association and fault tolerance,moreover,both its analysis results and process approach the metal process and analysis method of human brain,which greatly improves the accuracy for surface water environment quality assessment.