The study concerns on building a network used to provide a scientific and comprehensive alternative parking lot choice scheme when the driver's preferred parking lot is full. A parking supply-demand model is created combining with improved parking generation rate method and multiplication-weighted Voronoi diagram. Then the alternative relationship between two parking lots is determined according to the model, which can be materialized as the edge of the network. In the case, 49 parking lots in Nangang district, Harbin are used to set the parking lot network using the traditional and improved parking generation method to analyze their complex network basic parameters. The result is compared with the globally coupled network and nearest-neighbor coupled network. Data analysis shows that the parking lot determined by the connection algorithm has strong robustness, which means it is stable when facing a random attack. It also has a balanced performance in mean node degree, clustering coefficient, and other parameters comparing with the other two kinds of networks. It suggests that this method is a feasible method to select alternative parking lots when drivers cannot take service from their first parking choice.

研究涉及建立一个网络，当驾驶员的首选停车场已满时，该网络可提供科学而全面的替代停车场选择方案。结合改进的停车产生率方法和乘加权Voronoi图，建立了停车供需模型。然后根据模型确定两个停车场之间的替代关系，可以将其体现为网络的边缘。以哈尔滨市南岗区49个停车场为例，采用传统的改进停车场生成方法对停车场网络进行设置，分​​析其复杂的网络基本参数。将结果与全局耦合网络和最近邻居耦合网络进行比较。数据分析表明，由连接算法确定的停车场具有很强的鲁棒性，这意味着它在面对随机攻击时是稳定的。与其他两种网络相比，它在平均节点度，聚类系数和其他参数方面也具有平衡的性能。这表明当驾驶员无法从其第一个停车位选择服务时，该方法是选择其他停车场的可行方法。