Development of Distributed System for Practical Vehicle Routing Problems Ivana Ćavar, Ante Galić, Juraj Fosin, Tonči Carić, Hrvoje Gold Faculty of Transport and Traffic Engineering, Vukelićeva 4, Zagreb, Croatia {ivana.cavar, ante.galic, juraj.fosin , tonci.caric, hrvoje.gold@fpz.hr} Abstract: In order to solve the practical Vehicle Routing Problems (VRP) by heuristic methods, the programming language Mars and the interactive program environment Venera have been developed and implemented. The implementation and verification of the heuristic algorithms will be improved by the application of new network (cluster/grid) technologies and by adapting Mars and Venera for the distributed and parallel method of operation. 1. INTRODUCTION The Vehicle Routing Problem (VRP) represents a general term for an entire class of problems which involve selection of the routes for the given group of vehicles that move from a pre-defined number of depots and need to visit the spatially distributed users. The VRP is a problem of combinatory optimisation and belongs to NP problems [3]. As assistance to the traffic technologist in solving the VRP problems, the programming language Mars, extended by the transport logistics related data types and specially formed instructions, as well as the interactive programming environment Venera have been developed [4]. 2. EVOLUTION OF MARS AND VENERA The programming language Mars was used to implement and test the algorithms for solving the VRP problems, i.e. constructive algorithms that generate starting solution and optimisation algorithms that iteratively improve existing solution. The capacitated vehicle routing problems (CVRP) have been solved separately. For these, not only the test problems were solved but also the problems of routing groups of vehicles in real traffic environment. One of such problems is also the problem of managing the fleet of refuse collecting company Čistoća d.o.o. Zagreb[2]. By using the heuristic algorithms the routes have been created for the given group of vehicles of fixed capacity and significant improvement has been achieved in comparison to the previously used routes obtained by experience. 3. APPLICATION OF GRID TECHNOLOGY With the aim of achieving even greater success in solving this type of NP problems, reconstruction of the Venera environment has been planned, so that the new graphic interface will allows the problems to be set and the results to be visualised on the local PC, whereas the problem will be solved on a group of available computers included in the (cluster/grid) network. The group of computers enables distributed / parallel method in solving the complex problem, thus reducing the computation time. The programming language Mars will be extended by new data structures (multiple depots, garages), by instructions for distributed / parallel performance and the possibility of simpler expansion of the language itself. The verification of the model of improving the transport organisation in actual logistic transportation environment will continue by solving the VRPTW (Vehicle Routing Problem with Time Windows) using the data about the collection of postal items of the Hrvatska pošta Company (Croatian Post). 4. CONCLUSION Further development of the programming language Mars and the interactive programming environment Venera is directed to the simplification of the implementation and the parallelisation of the heuristic algorithms, and their performance in the cluster/grid environment, resulting in faster solving of the practical VRP problems. REFERENCES [1] Carić, T., Ivaković, Č., Protega, V.: “Hybrid Chains of Heuristic Methods for Vehicle Routing Problem with Time Windows”, The 14th International DAAAM Symposium, Sarajevo, October 2003. [2] Carić, T., Galić, A., Gold, H.: “Interactive Programming Environment for Solving Vehicle Routing Problem”, ITS in Europe, Budapest, May 2004. [3] Carić, T.: "Improvement in Transport Organization by Usage of Heuristic Methods" Ph.D. thesis (in Croatian), Faculty of Traffic Sciences, Zagreb, June 2004. [4] Galić, A.: “The Model of Transportation Processes Based on VRP”, Bs.C. thesis (in Croatian), Faculty of Traffic Sciences, Zagreb, April 2004. [5] Toth, P., Vigo, D.: “The Vehicle Routing Problem”, SIAM, Philadelphia, 2002. [6] http://venera.fpz.hr This work is a part of the technological poly-project “CRO-GRID Applications Optimization of transport organization” and it is