基于案例推理的露天矿山排土场滑坡事故预警方法研究

摘要

通过整理100组排土场滑坡的案例库，提出了基于案例推理（CBR ）的排土场滑坡中长期预警方法，采用框架法表示滑坡案例，RBF神经网络和欧式距离相结合实现案例的检索，由用户和专家来完成案例的修正与调整。在高村排土场进行了实际应用，通过案例检索获得了相似的案例，确定了预警等级及合理的处置方案，为排土场滑坡事故预防提供了科学指导。%This paper has collected 100 sets of waste dump landslide accident cases and given the mid -and-long term early-warning method of waste dump landslide based on case -based reasoning (CBR) ,in which the frames are used to express cases ,the RBF neural network and the Euclidean distance are combined to realize the case retrieval and the case re-vision and adjustment is completed by users and experts .In the application of engineering example of Komura dump ,a simi-lar case is obtained by case retrieval ,the early warning level is determined and the following reasonable treatments are put forward ,which can provide a scientific guidance for proper preventive measures .