MITSIM accepts as input time dependent origin to destination (OD) tables. Each OD table contains a time tag - the time that this table becomes effective - and travel demands for individual OD pairs. The travel demand for each OD pair is characterized by the following data items: an average departure rate and its standard deviation, a distribution factor, and a list of paths that connect this OD pair. The last 3 data items are optional and default values are supplied.
The time intervals that individual OD tables remain effective may have different lengths. This allows the use of shorter time intervals during peak periods and longer time intervals during the off-peak. The OD tables are sorted chronologically in the input file according to their time tags. The simulator stores only the current OD tables and reads the new ones when the current one expires.
OD tables can be specified individually for each vehicle type or, alternatively, the simulator can randomly assign a type to each generated vehicle based on a fleet-mix specified in the parameter file ( See Table C.1). Vehicle type is a combination of vehicle class (e.g. high performance passenger cars, low performance passenger cars, buses, trucks, trailer trucks, etc.), and group based on lane-use privilege (e.g. HOV and ETC) and access to information (e.g. guided and unguided).
The standard deviation of the average departure rate represents the randomness of total travel demand for the corresponding OD pair. The departure rate used in a particular simulation run is randomly sampled based on the given average departure rate and its standard deviation (default is to sample from a normal distribution). The distribution factor, which is a value between 0 and 1, determines the percentage of vehicles departing deterministically. For example, a distribution factor of 0.4 indicates 40% of vehicles will depart according to constant headways and the remaining 60% of vehicles according to Poisson distribution (random). The choice of distribution factor should be made based on traffic conditions. A distribution factor of close to 0, for example, can be used for low congestion traffic conditions and a value close to 1 for high congestion [May(1990)].
The inter-departure time between two vehicles is the reciprocal of the departure rate if the vehicle's departure is deterministic. For Poisson departures, inter-departure time is randomly drawn from negative exponential distribution, i.e:
where:
An alternative method for specifying travel demand is to use a vehicle trip table file. The vehicle trip table contains a list of scheduled vehicle departures, sorted chronologically according to departure times. Each vehicle departure is described by its origin, destination, type, and a predetermined path. Type and path are optional and can be determined based on the fleet mix and the route choice model respectively.
After a vehicle is generated, a set of vehicle and driver characteristics and a pre-trip path are assigned. The details are described in the following two sections.
