Abstract representation of an individual that makes one trip per day.
{
"id": 1,
"mode_choice": {
"type": "Logit",
"value": {
"mu": 2.0,
"u": 0.5
}
},
"modes": [
{
"type": "Constant",
"value": 1.0
},
{
"type": "Trip",
"value": {
"departure_time_model": {
"type": "ContinuousChoice",
"value": {
"choice_model": {
"type": "Logit",
"value": {
"mu": 1.0,
"u": 0.5
}
},
"period": [
0.0,
200.0
]
}
},
"destination_schedule_utility": {
"type": "None"
},
"legs": [
{
"class": {
"type": "Road",
"value": {
"destination": 1,
"origin": 0,
"vehicle": 0
}
},
"schedule_utility": {
"type": "AlphaBetaGamma",
"value": {
"beta": 0.01,
"gamma": 0.04,
"t_star_high": 100.0,
"t_star_low": 100.0
}
},
"stopping_time": 600.0,
"travel_utility": {
"type": "Polynomial",
"value": {
"b": -0.02
}
}
}
],
"origin_delay": 300.0,
"origin_schedule_utility": {
"type": "None"
},
"pre_compute_route": true,
"total_travel_utility": {
"type": "Polynomial",
"value": {
"c": 0.001
}
}
}
}
]
}
Id used when writing the results of the agents.
Value must be greater or equal to 0.0
Modes accessible to the agent.
Must contain a minimum of 1 items
Mode of transportation available to an agent.
An activity (e.g., staying home, traveling) that always provide the same utility level.
Representation of a utility (or monetary) amount.
A trip consisting in a sequence of legs (either on the road or virtual).
Representation of the mode of transportation for a trip with one or more legs, consisting in traveling on the road or virtually.
The trip is a sequence of legs, where each leg contains a travel part (either on the road, with a given origin, destination and vehicle; or virtually, using a given travel-time function) and a stopping part (with a fixed and given stopping time).
The destination of a leg does not have to be equal to the origin of the following leg, i.e., the agents are allowed to teleport from one node to another (and even change their vehicle).
The departure time from origin is the only choice variable (the departure time from any following leg is equal to the arrival time at the stopping point of the previous leg, plus the stopping time of the previous leg).
The route chosen for each (road) leg of the trip are the fastest route (in term of expected travel time), given the expected departure time from the origin of the leg.
The arrival time at destination is the arrival time at the stopping point of the last leg, plus the stopping time for this last leg.
The total trip utility is composed of:
origin_schedule_utility. - A function of total travel time of the trip (i.e., the sum of the travel time of each leg, excluding stopping time): total_travel_utility. - A function of arrival time at the stopping point for each leg: leg's schedule_utility. - A function of travel time for each leg (excluding stopping time): leg's travel_utility. - A function of arrival time at destination (which accounts for the stopping time of the last leg): destination_schedule_utility.When the utility for a given component is not specified, it is assumed to be null.
In practice, one of total_travel_utility or legs' travel_utility is usually null but this is not enforced by the model.
The legs of the trips.
The full trip consists realizing this legs one after the other.
Must contain a minimum of 1 items
A leg of a trip.
Type of the leg (road or virtual).
A leg with travel on the road.
A leg of a trip on the road network.
Origin node of the leg.
Value must be greater or equal to 0.0
Destination node of the leg.
Value must be greater or equal to 0.0
Vehicle used for the leg.
Value must be greater or equal to 0.0
A virtual leg, with a fixed TTF, independent from the road network.
Constant or piecewise-linear function.
A piecewise-linear function.
y values of the function.
Representation of time duration or timestamp, expressed in seconds.
Starting x value.
Interval between two x values.
A constant function.
Time spent at the stopping point of the leg, before starting the next leg (if any).
Travel utility for this specific leg (a function of the travel time for this leg).
{
"type": "Polynomial",
"value": {
"b": -10.0
}
}
{
"type": "Polynomial",
"value": {
"b": -5.0,
"c": -2.0
}
}
A polynomial function of degree 4.
Constant, linear, quadratic and cubic functions are special cases.
Coefficient of degree 0.
Coefficient of degree 1.
Coefficient of degree 2.
Coefficient of degree 3.
Coefficient of degree 4.
Schedule utility at the stopping point (a function of the arrival time at the stopping point).
The schedule utility is always null.
The schedule utility is computed using the alpha-beta-gamma model.
There is a penalty beta for leaving / arriving early and a penalty gamma for leaving / arriving late.
Compute the schedule-delay utility using Vickrey's alpha-beta-gamma model
The earliest desired arrival (or departure) time.
The latest desired arrival (or departure) time (must not be smaller than t_star_low).
The penalty for early arrivals (or departures), in utility per second.
The penalty for late arrivals (or departures), in utility per second.
{
"type": "AlphaBetaGamma",
"value": {
"beta": 5.0,
"gamma": 20.0,
"t_star_high": 31350.0,
"t_star_low": 27900.0
}
}
Delay between the departure time of the trip and the start of the first leg.
Model used for the departure-time choice.
The departure time is always equal to the given value.
Representation of time duration or timestamp, expressed in seconds.
The departure time is chosen among a finite number of values.
Values among which the departure time is chosen.
Representation of time duration or timestamp, expressed in seconds.
Discrete choice model.
Choose the alternative with the largest utility.
Choose the alternative with the largest value.
Uniform random number between 0.0 and 1.0 to choose the alternative in case of tie.
Value must be greater or equal to 0.0 and lesser or equal to 1.0
Constants added to the value of each alternative.
The number of constants does not have to match the number of alternatives. If there are less constants than alternatives, then the constants are cycled over.
If None, no constant is added to the alternatives' value.
Choose the alternative using Logit probabilities.
A discrete or continuous Logit model
Uniform random number between 0.0 and 1.0 for inversion sampling.
Value must be greater or equal to 0.0 and lesser or equal to 1.0
Variance of the error terms, must be positive.
Value must be greater or equal to 0.0001
{
"type": "Deterministic",
"value": {
"constants": null,
"u": 0.5
}
}
Offset time added to the chosen departure-time value (can be negative).
The departure time is chosen according to a continuous choice model.
Interval in which the departure time is chosen.
Must contain a minimum of 2 items
Must contain a maximum of 2 items
Representation of time duration or timestamp, expressed in seconds.
Continuous choice model.
A discrete or continuous Logit model
Uniform random number between 0.0 and 1.0 for inversion sampling.
Value must be greater or equal to 0.0 and lesser or equal to 1.0
Variance of the error terms, must be positive.
Value must be greater or equal to 0.0001
Total travel utility of the trip (a function of the total travel time of the trip).
{
"type": "Polynomial",
"value": {
"b": -10.0
}
}
{
"type": "Polynomial",
"value": {
"b": -5.0,
"c": -2.0
}
}
A polynomial function of degree 4.
Constant, linear, quadratic and cubic functions are special cases.
Coefficient of degree 0.
Coefficient of degree 1.
Coefficient of degree 2.
Coefficient of degree 3.
Coefficient of degree 4.
Schedule utility at origin of the trip (a function of the departure time from origin).
The schedule utility is always null.
The schedule utility is computed using the alpha-beta-gamma model.
There is a penalty beta for leaving / arriving early and a penalty gamma for leaving / arriving late.
Compute the schedule-delay utility using Vickrey's alpha-beta-gamma model
The earliest desired arrival (or departure) time.
The latest desired arrival (or departure) time (must not be smaller than t_star_low).
The penalty for early arrivals (or departures), in utility per second.
The penalty for late arrivals (or departures), in utility per second.
{
"type": "AlphaBetaGamma",
"value": {
"beta": 5.0,
"gamma": 20.0,
"t_star_high": 31350.0,
"t_star_low": 27900.0
}
}
Schedule utility at destination of the trip (a function of the arrival time at destination).
The schedule utility is always null.
The schedule utility is computed using the alpha-beta-gamma model.
There is a penalty beta for leaving / arriving early and a penalty gamma for leaving / arriving late.
Compute the schedule-delay utility using Vickrey's alpha-beta-gamma model
The earliest desired arrival (or departure) time.
The latest desired arrival (or departure) time (must not be smaller than t_star_low).
The penalty for early arrivals (or departures), in utility per second.
The penalty for late arrivals (or departures), in utility per second.
{
"type": "AlphaBetaGamma",
"value": {
"beta": 5.0,
"gamma": 20.0,
"t_star_high": 31350.0,
"t_star_low": 27900.0
}
}
If true, the routes of the trip are computed during the pre-day model (faster). If false, they are computed during the within-day model (which means that the route for second and after legs is computed using the actual departure time, not the predicted one)..
Choice model used for mode choice.
When not specified, the first mode is always chosen.
Choose the alternative with the largest utility.
Choose the alternative with the largest value.
Uniform random number between 0.0 and 1.0 to choose the alternative in case of tie.
Value must be greater or equal to 0.0 and lesser or equal to 1.0
Constants added to the value of each alternative.
The number of constants does not have to match the number of alternatives. If there are less constants than alternatives, then the constants are cycled over.
If None, no constant is added to the alternatives' value.
Choose the alternative using Logit probabilities.
A discrete or continuous Logit model
Uniform random number between 0.0 and 1.0 for inversion sampling.
Value must be greater or equal to 0.0 and lesser or equal to 1.0
Variance of the error terms, must be positive.
Value must be greater or equal to 0.0001
{
"type": "Deterministic",
"value": {
"constants": null,
"u": 0.5
}
}