Bi-level GA and GIS for Multi-objective TSP Route Planning

Abstract Route planning is usually carried out to achieve a single objective such as to minimize transport cost, distance traveled or travel time. This article explores an approach to multi-objective route planning using a genetic algorithm (GA) and geographical information system (GIS) approach. The method is applied to the case of a tourist sight-seeing itinerary, where a route is planned by a tour operator to cover a set of places of interest within a given area. The route planning takes into account four criteria including travel time, vehicle operating cost, safety and surrounding scenic view quality. The multi-objective route planning in this paper can be viewed as an extension of the traditional traveling salesman problem (TSP) since a tourist needs to pass through a number of sight points. The four criteria are quantified using the spatial analytic functions of GIS and a generalized cost for each link is calculated. As different criteria play different roles in the route selection process, and the best order of the multiple points needs to be determined, a bi-level GA has been devised. The upper level aims to determine the weights of each criterion, while the lower level attempts to determine the best order of the sights to be visited based on the new generalized cost that is derived from the weights at the upper level. Both levels collaborate during the iterations and the route with the minimal generalized cost is thus determined. The above sight-seeing route planning methodology has been examined in a region within the central area of Singapore covering 19 places of interest.

[1]  Fred Glover,et al.  Tabu Search - Part II , 1989, INFORMS J. Comput..

[2]  Teuvo Kohonen,et al.  Self-Organization and Associative Memory , 1988 .

[3]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[4]  Teuvo Kohonen,et al.  Self-organization and associative memory: 3rd edition , 1989 .

[5]  Gunar E. Liepins,et al.  Schema Analysis of the Traveling Salesman Problem Using Genetic Algorithms , 1992, Complex Syst..

[6]  Brian Logan,et al.  Route Planning with Ga , 1996 .

[7]  Fred W. Glover,et al.  Tabu Search - Part I , 1989, INFORMS J. Comput..

[8]  Chang Wook Ahn,et al.  A genetic algorithm for shortest path routing problem and the sizing of populations , 2002, IEEE Trans. Evol. Comput..

[9]  Dirk Van Gucht,et al.  Parallel Genetic Algorithms Applied to the Traveling Salesman Problem , 1991, SIAM J. Optim..

[10]  C. D. Gelatt,et al.  Optimization by Simulated Annealing , 1983, Science.

[11]  Bryant A. Julstrom,et al.  Very greedy crossover in a genetic algorithm for the traveling salesman problem , 1995, SAC '95.

[12]  D. J. Smith,et al.  A Study of Permutation Crossover Operators on the Traveling Salesman Problem , 1987, ICGA.

[13]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[14]  David J. Forkenbrock,et al.  GUIDEBOOK FOR ASSESSING THE SOCIAL AND ECONOMIC EFFECTS OF TRANSPORTATION PROJECTS , 2001 .