Transportation planning system and method thereof

Abstract

A transportation planning system and method are disclosed. The system includes an origin-to-destination demand prediction module, a vehicle selection module, and a semi-dynamic route assignment module. The origin-to-destination demand prediction module predicts the origin-to-destination demand for each origin-to-destination pair in a target area within a given time period. The vehicle selection module predicts the vehicle selection results for each origin-to-destination pair based on variables of vehicle and user attributes. The semi-dynamic route assignment module executes a semi-dynamic route assignment procedure for this time period based on the origin-to-destination demand and includes a route information unit and a route assignment unit. The route information unit establishes a route library for each origin-to-destination pair, including multiple routes, and extracts the route variables for each route to obtain route information. The route assignment unit predicts the traffic volume for each route based on the vehicle selection results and the route information for static assignment, and defers the origin-to-destination demand for any route to the next time period if a user cannot reach the destination of that route within the given time period.

Claims

1. A transportation planning system, comprising: a demand forecasting module for forecasting the demand for each origin-destination pair of a target area within a time period; a vehicle selection module for forecasting the vehicle selection result for each origin-destination pair based on variables of a vehicle and user attributes; and a semi-dynamic route assignment module for executing a semi-dynamic route assignment procedure for the time period based on the demand for origin-destination pairs, and comprising a route information unit and a route assignment unit, wherein the route information unit establishes a route library including a plurality of routes for each origin-destination pair, and extracts a route variable for each route to obtain route information; In this semi-dynamic route assignment procedure, the route assignment unit predicts the traffic volume of each route based on the vehicle selection result and the route information to perform a static assignment. If a user on any route cannot reach the destination of the route within the specified time period, the origin and destination demand of the route is deferred to the next time period. In the next time period, the static assignment is performed on the deferred origin and destination demand of the route until the user reaches the destination of the route.

2. The transportation planning system as described in claim 1, wherein the route assignment unit is used to determine whether the user of the route can reach the destination of the route within the time period by calculating the shortest travel time from the origin of the route to any node on the route.

3. The transportation planning system as described in claim 2, wherein if the road user is unable to reach the end of the route within the time period, the route assignment unit determines that the road user is at a stop point on the route at the end of the time period, replaces the traffic volume of all downstream road segments after the stop point with 0, and defers the start and end demand of the route to the next time period.

4. The transportation planning system as described in claim 1, wherein the route database includes all the road segments traversed by each route.

5. The transportation planning system as claimed in claim 1, wherein the origin-to-destination demand forecasting module has an origin-to-destination demand forecasting artificial intelligence model to forecast the origin-to-destination demand, the origin-to-destination demand forecasting artificial intelligence model being obtained by training a deep learning model and a plurality of variable categories, wherein the variable categories include one or more of the following: transportation environment and usage characteristics, socio-economic and urban development, meteorological and external environment, time characteristics and historical trends, external influences and events.

6. The transportation planning system as claimed in claim 1, wherein the route assignment unit has a route assignment artificial intelligence model to predict traffic volume for each route, the route assignment artificial intelligence model being trained according to a deep learning model and a plurality of variable categories, wherein the variable categories include one or more of route data variables, transfer station data variables, route length, road level, usage intensity, road type, capacity, speed and speed limit.

7. A transportation planning method, comprising: predicting the origin-destination demand of each origin-destination pair including a target area in a time period using an origin-destination demand prediction module; predicting a vehicle selection result for each origin-destination pair using a vehicle selection module based on variables of a vehicle and user attributes; executing a semi-dynamic route assignment procedure for the time period using a semi-dynamic route assignment module based on the origin-destination demand, thereby establishing a route library including a plurality of routes for each origin-destination pair using a route information unit of the semi-dynamic route assignment module, and extracting a route variable for each route to obtain route information; and predicting the traffic volume of each route in the semi-dynamic route assignment procedure using a route assignment unit of the semi-dynamic route assignment module based on the vehicle selection result and the route information to perform a static assignment. When a user on any of the routes in the semi-dynamic route assignment program fails to reach the destination of the route within a given time period, the route assignment unit of the semi-dynamic route assignment module postpones the origin-destination requirement of the route to the next time period; and in the next time period, the route assignment unit of the semi-dynamic route assignment module performs the static assignment on the postponed origin-destination requirement of the route until the user reaches the destination of the route.

8. The transportation planning method as described in claim 7, wherein the step of deferring the origin-end requirement of a route to the next time period when the route assignment unit via the semi-dynamic route assignment module cannot reach the destination of the route within a time period includes: The semi-dynamic route assignment module calculates the shortest travel time from the origin to any node on the route to determine whether the user can reach the destination within the specified time period. If the user cannot reach the destination within the specified time period, the semi-dynamic route assignment module determines that the user has a stop on the route at the end of the specified time period. The semi-dynamic route assignment module replaces the traffic volume of all downstream road segments after the stop with 0 and postpones the origin and destination demand of the route to the next specified time period.

9. The transportation planning method as described in claim 7, wherein the route library includes all the road segments traversed by each route.

10. The transportation planning method as described in claim 7 further includes: A training procedure is performed based on a deep learning model and a plurality of variable categories to obtain an artificial intelligence model for demand forecasting, wherein the variable categories include one or more of the following: transportation environment and usage characteristics, socio-economic and urban development, meteorological and external environment, time characteristics and historical trends, external influences and events.

11. The transportation planning method as described in claim 7 further includes: A route assignment artificial intelligence model is obtained by training a deep learning model and a plurality of variable categories, wherein the variable categories include one or more of the following: route data variables, transfer station data variables, route length, road level, usage intensity, road type, capacity, speed and speed limit.

Images