Abstract: Urban parking problem has plagued people’s daily travel and puts a huge pressure on road traffic. In the case of free parking spaces in short supply, improving the utilization of parking spaces is one important way to solve the puzzle. Firstly, this paper develops a system prototype based on radio frequency identification technology and Baidu map API, obtaining real-time information on vacant parking spaces, capturing the driver request and providing parking spaces booking; then it quantifies the weight of four factors which influence the driver’s selection on parking spaces: the driving time from the current position to the parking lot, parking fee, the walking distance from parking lot to the right destination and the number of vacant parking lots. Then, in consideration of the vacant spaces are limited and real-time traffic conditions, the paper proposes a parking assignment model with 4 condition matrix, uniting the goal of user and system to a final goal which is to increase parking lots utilization. Finally, by using the Beijing road network data , traffic data , parking spaces data and other data in ArcGIS , a test is done, the test results show that , compared with ArcGIS location allocation model, the proposed allocation model is better, whose total system costs increased by 5% and the average utilization rate of parking spaces increased by 9%, the model is feasible.

Key words: real-time parking lot, location allocation, ArcGIS model

摘要： 以实时获取空余车位和智慧分配车位为目标，基于射频识别技术和地理信息技术开发智慧停车系统原型。采用模糊决策矩阵，定量化驾驶员从当前位置到停车场的驾驶时间、停车费用、停车场至目的地的步行距离和停车场的空余车位数这4个车位分配影响因素的权重；在考虑空余车位有限和实时交通状况的前提下，提出带有4个条件矩阵的车位分配模型，并利用某市道路网、交通状况、停车场等数据在ArcGIS中进行模型实验，实验结果表明，该智慧停车系统具有一定的可行性和实际意义。

关键词: 实时车位, 位置分配, ArcGIS模型