A Comparative Analysis of Optimization Heuristics Algorithms as Optimal Solution for Travelling Salesman Problem

Abstract

Travelling Salesman Problem (TSP) is considered non-deterministic polynomial time hard (NP hard) problem that cannot be solved traditionally especially when the number of cities increase. Therefore, Heuristic Algorithms are feasible solution to such type of problem. TSP is a representative of a larger class of problems known as combinatorial optimization problems. In TSP, if a Salesman wants to sell goods in different cities, he leaves a city and visits each other cities exactly once and returns back to the starting city. People may want to plan for the fastest or the most economical method to their destinations. The research aims to examine and develop effective and efficient optimization technique to get a shortest or suboptimal path. Google map uses Dijkstra's Algorithm as its fast-finding algorithm which is reported to have problem in searching for all route within a limited location. On the other hand, Ant Colony Optimization Algorithm can contribute effectively in solving lots of problems including shortest path problems, particularly, where other algorithms are inefficient. Both ACO and Dijkstra's Algorithm for simulations with a given routes(length) gave better result than random generation of routes for given cities. Results from the simulation experiment for ACO shows the Best Routes and total length for the best routes while results from Dijkstra's Algorithm show Minimum Cost for source node and destination node, for PSO gives better result with slower convergence. More iterations lead to get accurate results especially PSO Algorithm.