Development of Traffic Network and Car Parks Simulation Model for Estimating Effects of Parking Guidance and Information Systems

世界上已经有50多个城市安装了停车诱导和信息系统(PGI)。PGI系统为驾驶员提供停车场的位置、方向和可用性信息，以重新分配停车需求，减少热门停车场的排队和拥堵。本研究旨在建立带有停车场的交通网络仿真模型，以评估PGI系统的效果。为了了解驾驶员对PGI系统的反应，松山市进行了三次停车场现场调查。尽管知晓率很高，但使用该系统的汽车司机的比例相对较低，司机并没有像他们可能的那样改变停车选择行为。根据上述偏好调查，不同类型的停车可用性信息对停车场选择具有足够的重要性，这意味着司机的停车选择可能存在一些实际约束，如购物时间和地点约束。PGI系统有望减少这些约束。网络模拟模型由三个子模型组成：需求子模型、性能子模型和信息提供子模型。需求模型描述了驾驶员的停车选择和路径选择行为。性能子模型计算网络和停车场中车辆沿需求模型给出的路线的移动。信息子模型提供三种不同类型的可用性信息。模型计算结果表明，提供可用性信息将总体上重新分配停车需求，减少司机在网络和停车场中的出行和排队时间。然而，这取决于拥堵程度和满位/空位信息在拥堵情况下可能并不总是有效。

More than 50 cities in the world have already installed Parking Guidance and Information (PGI) systems. PGI systems provides location, direction and availability information of car parks to car drivers in order to redistribute parking demand and to reduce queues and congestion in the popular car parks.This study aims to develop traffic network simulation model with car parks for estimating effects of PGI systems.Three times of car park site survery have been made in Matsuyama city for understanding drivers'response to the PGI system. In spite of the high awareness rate, the percentage of car drivers who used the system was relatively low and drivers did not change car park choice behavior as much as they might be. According to the Stated Preference survey, different types of parking availability information were sufficiently important for car park choice.This means that there might be some actual constraints for drivers'car park choice such as shopping time and place constraints.PGI systems are expected to reduce those constraints.The network simulation model consists of three sub models ; demand sub-model, performance sub-model and information providing sub-model. The demand model describes a driver's car park choice and route choice behavior. The performance submodel calculates the movement of a vehicle in the network and car parks along the route given through the demand model. The information sub-model provides three different types of availability information. The results of the model calculation showed that providing availability information would generally redistribute parking demand and reduce traveling and queuing time of car drivers in the network and car parks. However, it depends on the congestion level and FULL/SPACES information might not be always effective under the congested situation.