FL-MTSP: a fuzzy logic approach to solve the multi-objective multiple traveling salesman problem for multi-robot systems
Ref: CISTER-TR-180101 Publication Date: Dec 2017
FL-MTSP: a fuzzy logic approach to solve the multi-objective multiple traveling salesman problem for multi-robot systemsRef: CISTER-TR-180101 Publication Date: Dec 2017
This paper considers the problem of assigning target locations to be visited by mobile robots. We formulate the problem as a multiple-depot multiple traveling salesman problem (MD-MTSP), an NP-Hard problem instance of the MTSP. In contrast to most previous works, we seek to optimize multiple performance criteria, namely the maximum traveled distance and the total traveled distance, simultaneously. To address this problem, we propose, FL-MTSP, a new fuzzy logic approach that combines both metrics into a single fuzzy metric, reducing the problem to a single-objective optimization problem. Extensive simulations show that the proposed fuzzy logic approach outperforms an existing centralized Genetic Algorithm (MDMTSP_GA) in terms of providing a good trade-off of the two performance metrics of interest. In addition, the execution time of FL-MTSP was shown to be always faster than that of the MDMTSP_GA approach, with a ratio of 89 %.
Published in Soft Computing, Springer, Volume 21, Issue 24, pp 7351-7362.