摘要:城市犯罪和公共安全是最重要的城市公共问题之一。作为产学研载体的大型工业园区大量的兴起，随之而来的园区内犯罪活动也日益增多。基于建筑信息模型(BIM)，结合闭路电视监控系统(CCTV)以及工业园区内的警卫巡逻，是一个有效预防犯罪的方法。在固有大型工业园区环境基础上，本文提出了一个大型工业园区内保安巡逻系统的数学模型。根据现有区内建筑物及车道、人行道等的分布，经过定量分析，获得空间的可视性、接近性和本征性，找出园区内防犯薄弱点，针对防范薄弱区域增添CCTV，确定保安的巡逻路径。最后通过基因遗传算法，对保安巡逻路径进行拓扑优化，得出最佳巡逻路径。通过文中的案例分析，本文所提出的模型可有效的预防大型工业园区内的犯罪发生，并大大提高了分析的效率。

Abstract: The crime and public safety in urban life is one of the most important public urban issues with the development of social market economic. As the large industrial parks which are the centers of industry-academic-research are significantly built, the crime prevention in these large industrial parks becomes more and more important. Closed-circuit television and guard patrol system are the major methods to prevent the crime through environment design. On the basis of environmental design and safe planning of large industrial park, this paper presents a mathematical model on the guard patrol system of large industrial park. Firstly, according to the distributions of buildings, roads, and sidewalks, the weak regions (relatively un-safe regions) can be obtained through the quantitative analysis of visibility, accessibility, and characteristic in the current site. Then, the track points of guard patrol system can be confirmed according to these weak regions. Finally, through the genetic algorithm, the tracks of guard patrol system are topologically optimized, and the best optimized guard patrol track is obtained. The case study illustrates the efficiency and accuracy of the proposed method.