Traveling Salesman Problem Multi-destination Route Recommendation System Using Genetic Algorithm and Google Maps API

Abstract

Google Maps does not provide route recommendations if users want to find the shortest route from multiple destinations or stop destinations, or more than two destinations. Departing from the shortcomings of Google Maps which cannot sort the sequence of multi-destination routes with the shortest distance, the researcher created an innovation with a genetic algorithm in solving the problem of the Traveling Salesman Problem category. The processes in the genetic algorithm of solving the Traveling Salesman Problem include data collection from primary document sources, ETL implementation, genetic algorithm implementation, and genetic algorithm testing with comparison algorithms. The data to be used in this study is primary data from the day tour package "Banyuwangi City Tour" from PT. LINTASNUSA TOURISM PRIMARY. This research produces recommendations for destination routes with the shortest real-time travel distance with short computational time. The genetic algorithm that has been programmed will be compared with other Traveling Salesman Problem solving algorithms, namely Nearest Neighbor and Brute Force. Based on the results of testing with primary data, the genetic algorithm is proven to be able to solve the Traveling Salesman Problem with the shortest average distance and the same as the solution of the Brute Force algorithm, which is 42.759 kilometres. The genetic algorithm also successfully recommended destination routes with shorter real-time travel distances or more optimal solutions compared to the Nearest Neighbor algorithm, but the genetic algorithm took 0.9 seconds slower computational time than the Nearest Neighbor algorithm.