基于需求密度感知与GNG的移动设施动态选址 方法

摘要

针对社会设施的动态选址问题，本文提出一种基于需求密度感知和Growing Neural Gas Networks (GNG)的设施动态选址方案。该方法主要有以下三方面，第一、对区域进行划分；第二、获取区域的资源需求密度，并对低密度区域进行过滤；第三：基于区域的需求密度对有限的信息资源进行合理分配。本文方法与Kmeans进行了对比，实验结果表明所提方法可有效对区域需求进行拓扑感知，并可对有限的可移动设施进行合理性规划。 For the problem of facilities dynamic location, a method based on demand density perception and growing neural gas networks (GNG) was proposed. This method can be organized into three parts: firstly, divide the region into many unit areas; secondly, obtain the demand density of each unit area and filter out the areas with low-density; thirdly, allocate the limited information resources reasonably based on the demand density. An experiment comparison with Kmeans was done. The results show that the method proposed can effectively realize the topological perception of regional demand, and can reasonably plan the limited mobile facility dynamic.