Traveling salesman approach for solving petrol distribution using simulated annealing

Abstract

This research presents an attempt to solve a logistic company’s problem of delivering petrolto petrol station in the state of Johor. This delivery system is formulated as a travelling salesmanproblem (TSP). TSP involves finding an optimal route for visiting stations and returning to point oforigin, where the inter-station distance is symmetric and known. This real world application is adeceptive simple combinatorial problem and our approach is to develop solutions based on the idea oflocal search and meta-heuristics. As a standard problem, we have chosen a solution is a deceptivelysimple combinatorial problem and we defined it simply as the time spends or distance travelled bysalesman visiting n cities (or nodes) cyclically. In one tour the vehicle visits each station just once andfinishes up where he started. As standard problems, we have chosen TSP with different stations visitedonce. This research presents the development of solution engine based on local search method knownas Greedy Method and with the result generated as the initial solution, Simulated Annealing (SA) andTabu Search (TS) further used to improve the search and provide the best solution. A user friendlyoptimization program developed using Microsoft C++ to solve the TSP and provides solutions tofuture TSP which may be classified into daily or advanced management and engineering problems