【正文】 guidance and information systems includes reports of case studies that show that there are quantifiable benefits to be derived from installing variable message signs indicating car parking space availabilty [6]. Benefits are described as quantitative, in terms of time saved, and qualitative in terms of public image and driver safety. WebTAG [7] guidance touches on the subject briefly in discussion of travel costs by including ?parking “costs”(which notionally include time spent searching and queuing for a space and walking to the final destination)?. The authors? perception is that car park accessibility isnormally considered after the urban design is plete and car park policy has been determined. The missing link is in the transport planning policy contribution to the initial design of urban areas with 16 respect to car park provision and accessibility. This apparent deficiency has recently been addressed in three different locations reported here. Each has sought ways to investigate the effects on the road work of car park planning policy and accessibility. All have looked for methods to minimize urban congestion caused by drivers searching for a car park space. All have used an SParamics microsimulation model to test the design options. A study in Nieuwegein (The Netherlands) modelled a large expansion in travel demand and the provision of car park spaces for a major town centre redevelopment, where Saturday afternoon shopping was the critical period. It incorporated ITS within the microsimulation model to deliver information to drivers on availability of spaces and routes to car parks. Another study, in Rochdale (England), models the distribution of spaces in conjunction with major town centre development plans. The goal is to optimise the provision of car parks with respect to adjacent land use and to minimise town centre congestion by considering car park access early in the design process. The third study, in Takapuna (New Zealand), is also investigating the effect of city centre expansion. It uses bespoke software to model the car park demand and a microsimulation model to assign the demand to the work. Once again the goal is to understand the effect of car park policy and minimise city centre congestion. CAR PARK MODELLING IN MICROSIMULATION 17 Typical design option tests for a microsimulation model include changes to road layout, public transport priority schemes, optimisation of signals, or changes in demand. Each individual vehicle in the simulation will react to these changes, and the congestion they cause, as it moves to its destination. When testing the effect of car park policy decisions, the emphasis moves from examination of the effect of changes to the road work to examination of the effect of changes in the destination for that part of the trip undertaken in a car. The simulation model must now include the capability to distinguish between the driver?s destination and the vehicle?s parking location and make dynamic choices between these locations. 18 Figutr 1： Rochdale city centre carparks Arrivals Car parks are an entity within the microsimulation model, and are linked to zone destinations and car parks may serve more than one zone. Allocation of vehicles to car parks is undertaken by limiting car park access to specific trip purposes. The model includes car parking charges and the distances between car parks and associated zones as ponents of the generalised trip cost. As each vehicle type may have different cost coefficients, the modeler may differentiate between drivers who will accept a longer walk and those who will accept a higher charge. 19 If a car park is full then vehicle drivers within the simulation wait at the entrance for a predetermined time, after which they reassess their choice of car park and possibly proceed to another. Using an external software controller it is possible to monitor car park occupancy within the simulation and change a vehicle?s destination before it reaches the queue. As an example of how this methodology can be used to implement a car park policy model, consider a city centre zone with a mix of retail and mercial use with several car parks available within reasonable walking distance. Drivers will have a preferred location based on their proposed length of stay and the car park charging drivers may have a contract for permit parking. A car park may have multiple adjacent entrances, each coded with a restriction to force vehicles to accept the appropriate parking charge. The effect in the simulation is that short stay vehicles enter car parks closer to their destination or with a lower charge. The long stay vehicles enter via the entry links with the higher charges or accept a longer walk time. The modeller can test responses to car parking changes by adjusting entry charges for different car parks or by varying the level of permit parking. Land use changes may be modelled by adjusting the proportion of driver and vehicle types using a particular zone and related car parks. Departures The assignment of all vehicles to an SParamics road work is 20 controlled by a detailed (5 minute) time release profile. In its simplest form of use, the journey origin car park is determined by finding the minimum journey cost,which includes the walk time, or vehicles may simply be released in proportion to the size of the car park. If more control is required, such as the ability to match departures to arrivals at the same car park, the release may be triggered by an external software controller linked to the simulation model which uses an algorithm to determine when to release vehicles and where they originate on the work. This may be associated with a car park occupancy monitoring system and be used to match vehicle arrivals with a