基于神经网络参数优化的岩体边坡稳定性评价

高陡岩体边坡的稳定性是一个内部存在多种相互联系、相互影响因素的复杂系统，而BP人工神经网络属于非线性动态系统，较适合用于评价高陡岩体边坡稳定性。分析了BP网络模型参数对高陡岩体边坡稳定性评价精度的影响，并提出了对模型参数进行优化，以提高预测精度的若干办法。用一工程实例对参数优化后的BP神经网络在高陡岩体边坡稳定性评价中的应用效果进行了检验。研究表明，用经参数优化的BP人工神经网络模型预测高陡岩体边坡稳定性是可行的，预测结果虽然与实际状态存在一定的误差，但仍可以相对准确地反映边坡稳定状况。

Prediction research on water quality improvement through water diversion project for a city in plain river network area

Along with quick social and economic development,the water quality of a city in the plain river network area is influenced seriously and the water quality can not meet the management goal of water function requirement,so a water diversion project is planned for improving the water quality.In order to analyze the improvement effect by project water diversion,in the light of different project layouts and water diversion scale,a 1D hydraulic and water quality model is applied to analyze the water quality variation condition with 40-year hydrological data.According to the analysis result,the optimized water diversion flow that makes the water quality meet the management goal of water function area is determined and the optimized project layout and water diversion scheme is proposed.