洪水逃生最优路线决策方法研究与仿真

Research and Simulation of Flood Escape Routes Decision-Making Method

研究洪灾中避难路线的准确决策问题。洪水的行动路线带有很强的随机性,行动方向极易发生随机性改变,造成建立的路径选择决策模型很难根据洪水的突发变化进行随机调整,导致逃生路线选择模型的残差过大而发生崩溃。传统的逃生模型都没有考虑随机性突发状况对模型干扰因素,过大的决策模型残差导致洪水中避难路径的准确性不高。为解决上述问题,提出了一种采用特征筛选数据挖掘路径优化模型。选取洪水区域在撤退过程中的路线特征加入到模型中拥有残差决策,利用聚类方法对路线特征进行筛选,建立优化数据挖掘决策模型,从而选取洪水逃生路线。实验结果表明,提出的算法能够提高洪灾中避难路线选择的准确性。

Research accurate decision of asylum route in the flood. The movement structure of the flood is com- plex and changes greatly. This paper put forward a data mining path optimization model based on feature selection. The route characteristics of flooded evacuation was selected to join the model of residual decisions. Using clustering method classification feature, the optimal decision model of data mining was established. The experiment results show that the improved algorithm can improve the accuracy of decision of asylum route in the flood.