Notes

Outline

NETWORK 2001 Transportation Planning under Multiple Objectives Woodam Chung John Sessions

Contents Old NETWORK algorithms New NETWORK algorithm An example of transportation planning problem with multiple goals Demonstration of NETWORK2001 software

Old Algorithms TRANSHIP Model (Kirby et al. 1981) Mixed Integer Programming (MIP)

Old Algorithms TIMBER TRAN. MODEL (Sullivan 1974) K-shortest path + MIP

Slide 5

New Challenges Very large network problems Transportation planning for a large area GIS applications Introduction of side constraints Volume upper bound Mill capacity Environmental considerations or requirements Introduction of multiple goals Transportation and road costs Open road mileage Earth being moved Environmental hazard probability (index)

New NETWORK Algorithm Solve a large network problem efficiently Flexible to consider additional constraints and multiple goals

New NETWORK Algorithm Flowchart

New NETWORK Algorithm Step1: k-shortest paths

New NETWORK Algorithm Step2: Simulated Annealing Assigning a route to each sale

New NETWORK Algorithm Step2: Simulated Annealing Assigning a route to each sale Evaluating a feasible solution Multiple goals, side constraints

New NETWORK Algorithm Step2: Simulated Annealing Assigning a route to each sale Evaluating a feasible solution Identifying the best solution

New NETWORK Algorithm Step3: Recalculate the equivalent variable costs Step4: Return to Step 1

Example Management Goals Minimizing total costs Minimizing total open road length

Example Input data

Example Results