Model the time-dependent choices on activity, parking location, departure time and travel route,Hai-Jun Huang (黄海军)School of Economics and Management, Beijing University of Aeronautics and Astronautics, China e-mail: haijunhuangbuaa.edu.cn,Routing:,O,D,O,D,可证明:,再也找不到比,更好的流量分布，能够令网络总时间进一步减小。 这个流量分布是系统最优状态下的流量分布，如果不依靠收费调节，就需要靠行政命令来实现。现在我们不是用行政手段，只设置了一个收费站和恰当的收费水平，仍然靠自由市场的准则，居然得到了系统最优的状态（也就是用市场准则实现了计划经济的目标）！,Parking:,Parking has become an increasingly serious problem in most densely populated cities. Search time for parking has substantially increased in central business districts.,Parking, as an integral component of urban transportation systems, influences not only the parking system itself, but also the whole transportation system. Parking can significantly affect the level of traffic congestion.,Purposes of this study,Understand the comprehensive behaviour in peoples making decisions associated with parking and travelling, and analysing the consequences of various policies. Space-time complex (quasi-dynamics+network+equilibrium),THE PROBLEM,People depart from origin r , and do the following decision-makings:,(1) Activity choice (type, location, duration) (2) Parking location choice (3) Travel path choice,Main assumptions:,All commuters are homogenous. Parking facilities are characterized by type (on-street, off-street) and time-varying parking charges. The decision-making process on activity, travel and parking choices follows a hierarchical structure.What to do (activity type choice), where (activity location choice), how long for it (activity duration choice), where to park their cars (parking location choice), which path to reach the parking lot (travel route choice) Each activity is associated with a specific time window, outside which a traveller will suffer a large penalty.,Off-street parking in BJ,On-street parking in BJ,Travel and activity utilities (1),= the comprehensive utility departing from origin r to destination s in period t and performing activity j with duration,= the travel disutility departing from origin r to destination s in period t and performing activity j with duration,= the total utility of performing activity j,= the systematic components of utility associated with destination s,= the systematic components of utility associated with activity j at destination s,(1),Travel and activity utilities (2),= the in-vehicle travel time from origin r to location i,= the parking searching time,= the walking time from parking i to destination s,= the time-varying parking charge,= schedule delay penalty of early or late arrival,= the starting time of an activity for travellers departing from origin r in period t,(4),(5),(2),Parking search time,=parking accumulation at location i,=parking capacity of location i,The Parking accumulation can be determined in terms of path inflow rates from origins (See the paper for details),(10),Parking charge function,= parking charge discount,= hourly parking fee,= parking duration,Walking time,is the walking distance from location i to destination s,and is the walking speed.,(6),(7),(8),The schedule delay costs of early or late arrival at destination s,= the desired time window for starting activity j at destination s without any schedule delay penalty,(9),Behaviour Modeling: 1. Activitys destination/type/duration choices,Nested logit model,Marginal demand,+, and are the scaling parameters that measure the variability in choice behaviour among the travellers.,(17),(18),(19),2. Parking location choice equilibrium,Deterministic parking location choice equilibrium conditions,(20)-(22),(23),3. Route choice equilibrium,Deterministic route choice equilibrium conditions,(24)-(26),(27),MODEL FORMULATION: Equivalent variational inequality,(41),VIs constraints:,flow conservation at origin r,inflow entering link a in period k,the sum of all link travel times along the path,the number of potential travellers within the origin zone r at the end of period t-1,(32),(33),(34),(35),(38),(40),(39),Two classes of equilibria included:,Logit-based equilibrium: Choices of activity type, activity duration, activity destinationDeterministic equilibrium: Choices of parking sites, travel routeFlow conservation between successive periodsSolution Methods:,NUMERICAL EXAMPLE,Work zone,# of activity = 4 H: staying at home W: working E: having lunch/dinner S: joining social activity Five on-street parking locations (c, d, e, g, i) Four off-street parking locations (a, b, f, h),Input data for link travel time functions,Input data associated with activities,00:00 to 24:00 is discretized into 96 time periods each with 15 minutes,Input data for temporal utility profiles for the activities,Step parking charges,The off-street parking (garage) in home zone and the on-street parking around restaurant are free.,Convergence of solution algorithm,