城市停车场路径规划系统设计与实现  

Design and implementation of urban parking path planning system

作　　者：孙剑萍 徐昀 李晓鹏 汤兆平 SUN Jianping;XU Yun;LI Xiaopeng;TANG Zhaoping(School of Transportation and Logistics,East China Jiaotong University,Nanchang 330013,China;Shenhua Zhunneng Group Co.,Ltd.,Ordos 010300,China)

机构地区：[1]华东交通大学交通运输与物流学院,江西南昌330013 [2]神华准能集团有限责任公司,内蒙古鄂尔多斯010300

出　　处：《现代电子技术》2021年第21期83-89,共7页Modern Electronics Technique

基　　金：国家自然科学基金项目(51965017)。

摘　　要：针对停车过程中有效泊位信息预测及行车路径选择问题,采用BP神经网络预测空闲泊位数,建立最低出行成本路径选择多目标规划模型,设计基于路径方向引导的双向搜索Dijkstra改进算法,基于ArcGIS Engine城市路网地理信息及Python爬取百度地图智慧交通信息,开发设计城市停车场路径规划系统,对城市停车场泊位进行预测及路径规划。研究实例显示:设计开发的系统能够有效地规避驾驶人无效寻泊,并可根据个人意愿,选择距离、出行时间、出行成本最低的三种模式,进行选择行车路线规划,对提高停车路径规划效率、优化停车场资源分配具有重要的参考意义。In order to solve the problem of effective berth information prediction and driving path selection in the parking process,BP neural network is used to predict the number of free parking spaces,establish a multi-objective planning model for the lowest travel cost path selection,and design an improved bidirectional search Dijkstra algorithm based on path direction guidance.On the basis of ArcGIS Engine′s urban road network geographic information and Python′s crawling Baidu map intelligent traffic information,an urban parking lot path planning system is developed and designed to carry out path planning and parking space prediction for urban parking lots.Research examples show that the system developed in this paper can effectively avoid inefficient parking by drivers,and can choose three modes of distance,travel time and lowest travel cost to plan the driving route according to personal wishes.It has important reference significance to improve the efficiency of parking path planning and optimize the allocation of parking resources.

关 键 词：泊位预测 路径选择 多目标规划 BP神经网络 双向Dijkstra算法 ArcGIS技术 系统设计 

分 类 号：TN99-34[电子电信—信号与信息处理] U495[电子电信—信息与通信工程]