城市派出所空间位置优化及警力分配

提高公安派出所的治安防控能力是解决我国城市治安问题的有效途径之一。本文从公安派出所空间位置优化和警力分配两个角度出发,针对当前公安派出所空间位置布局较为随意及警力资源紧缺的现状,以及多目标优化的需求,结合新兴空间大数据和犯罪数据,进行派出所多目标空间优化研究,包括派出所空间布局的数学期望、量化指标以及多目标空间优化模型,并在相关理论模型的基础上利用犯罪热点和房屋建筑分布的空间异质性开展派出所警力（治安和户籍分类）的空间分配研究。以兰州市中心城区为分析区域开展公安派出所空间位置优化和警力分配实验,结果表明：① 在不增加派出所数量、最大维持原有空间布局的前提下,本文提出的方法能有效降低派出所服务区重叠度（17.2%）和平均响应时间（6.67 s）,提高面积覆盖度（12.01%）和需求点覆盖度（7.25%）,并有效提高研究区内各区域应急响应时间的公平性（基尼系数由0.382降低到0.268）;② 数据分析发现,城市犯罪热点以及房屋建筑空间分布均存在空间分异现象。理解并量化分析空间分异特征,有助于优化派出所治安和户籍警力的配置。

Spatial Location Optimization of Urban Police Stations and Police Allocation

Public security is a significant issue in China currently. A prevention and control system was implemented in an attempt to effectively increase public security. Improving public security through controlling the placement of police stations is one of the ways to solve issues of urban crime in China. A method was proposed to optimize police station locations and the allocation of police members. This method uses qualitative principles, quantitative indexes, and the multi-objective models of the location-allocation of police stations, taking into account police resource constraints. The method was divided into three steps: (1) The existing police stations were streamlined using the minimum facility model; (2) The maximum coverage model was used to optimize the police stations in urban key areas; (3) The improved minimum impedance model was introduced for global optimization with consideration to social equity. Method parameters were defined and discussed, and police station location optimizaion was analyzed based on crime and building data. Results showed spatial heterogeneity in crime hotspots and building distributions. These results form the foundation of this paper's theory. Using Lanzhou to carry out empirical research, the results show that the proposed method can effectively reduce the degree of overlap and the average response time of police. The results were as follows: (1) The method can effectively improve the coverage of demand points and service areas. Additionally, it can effectively improve the workload of emergency police services, as the Gini coefficient of emergency response time decreased significantly. The degree of overlap in police service areas decreased by 17.2%, and the average response time decreased by 6.67 seconds. The coverage of high-demand points, coverage of key areas, and general area coverage increased by 7.25%, 3.00%, and 12.01%, respectively, and the Gini coefficient of response time decreased from 0.382 to 0.268; (2) Through the analysis of spatial heterogeneity of crime hotspots and building distribution, it was found that this method can support the allocation of security police members and the police members in charge of household registration. Using this approach, more security police members were allocated to the blocks with higher crime, and more household registration police members to the blocks with larger numbers of residents according to the housing area and the number of crimes. As a result, the number of police members vary between police stations, and the ratio of security police members to household registration police members in the same police station is also different.