利用Minkowski距离逼近道路网络距离算法研究

发布时间：2018-01-11 04:03

  本文关键词：利用Minkowski距离逼近道路网络距离算法研究　出处：《武汉大学学报(信息科学版)》2017年10期 　论文类型：期刊论文

  更多相关文章： Minkowski距离 道路网络 距离逼近 弯曲度 地理加权回归分析

【摘要】：道路网络背景下的距离度量(如道路网络距离、旅行时间)是在空间分析或空间统计过程中常用的距离度量,但在科研过程中由于道路数据的可获得性和精度等方面的限制,该类距离的计算可能较为困难。Minkowski距离函数是欧氏空间中的广义距离函数,其参数p值的不同代表着对空间不同的度量。利用Minkowski的通用性和灵活性(参数p不同的取值),研究如何更好地逼近道路网络距离。同时,探索不同道路网络的部分计量特征(如密度、弯曲度等)与最优p值之间的关系。实验证明,相对于最常用的欧氏距离度量,优选p值后的Minkowski距离函数能够更大程度上逼近道路距离。而通过对道路网络计量特征与最优p值之间的关系的分析,指出了弯曲度与最优p值之间的对应关系,它对于p值的选择具有重要的指导意义。此外,为了验证Minkowski距离逼近算法的可行性,以地理加权回归分析为例,通过对比传统的欧氏距离度量、最优Minkowski距离度量和道路网络距离(旅行时间)对模型解算结果的影响,指出优选后Minkowski距离一定程度上更接近于采用旅行时间对模型解算的结果。
[Abstract]:Distance measurement (such as distance of road network, travel time) in the background of road network is commonly used in the process of spatial analysis or spatial statistics. However, due to the limitation of road data availability and precision in scientific research, it may be difficult to calculate this kind of distance. Minkowski distance function is a generalized distance function in Euclidean space. The difference of parameter p represents different measurements of space. By using the generality and flexibility of Minkowski, the paper studies how to approach the distance of road network better. At the same time. The relationship between the partial metrological characteristics (such as density, curvature, etc.) of different road networks and the optimal p value is explored. The experiments show that this method is relative to the most commonly used Euclidean distance measure. The Minkowski distance function after selecting p value can approach the road distance to a greater extent, and through the analysis of the relationship between the metrological characteristics of road network and the optimal p value. The corresponding relationship between the curvature and the optimal p value is pointed out, which has important guiding significance for the selection of p value. In addition, in order to verify the feasibility of the Minkowski distance approximation algorithm. Taking geographical weighted regression analysis as an example, the effects of traditional Euclidean distance metric, optimal Minkowski distance measurement and road network distance (travel time) on the results of the model are compared. It is pointed out that the Minkowski distance after optimal selection is closer to the result of using travel time to solve the model.
【作者单位】： 武汉大学遥感信息工程学院;中国民航科学技术研究院航空安全技术实验室;72515部队;
【基金】：国家自然科学基金(41401455,U1533102) 地理国情监测国家测绘地理信息局重点实验室开放基金(2015NGSM10)~~
【分类号】：P208;U491
【正文快照】： 项目资助:国家自然科学基金(41401455,U1533102);地理国情监测国家测绘地理信息局重点实验室开放基金(2015NGSM10)。在空间分析和统计技术应用的过程中,空间距离度量的使用非常普遍,如插值技术、空间相关性分析和空间聚类分析等。而欧氏距离(Euclide-an distance)度量往往是默，

本文编号：1408011