Simulated kalman filter with modified measurement, substitution mutation and hamming distance calculation for solving traveling salesman problem
Background - Combinatorial optimization is a problem that consists of finding the optimum combination of each component from a finite set of objects. Most metaheuristic algorithms are designed for continuous optimization problem. Modification or enhancement is needed for continuous algorithm to solv...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | en en |
| Published: |
Springer International Publishing
2022
|
| Subjects: | |
| Online Access: | https://umpir.ump.edu.my/id/eprint/32897/2/1.1%20Simulated%20kalman%20filter%20for%20solving%20travelling%20salesman%20problem%20based%20on%20numerical%20ordering.pdf https://umpir.ump.edu.my/id/eprint/32897/13/Simulated%20kalman%20filter%20with%20modified%20measurement.pdf https://umpir.ump.edu.my/id/eprint/32897/ https://doi.org/10.1007/978-981-19-2095-0_27 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Background - Combinatorial optimization is a problem that consists of finding the optimum combination of each component from a finite set of objects. Most metaheuristic algorithms are designed for continuous optimization problem. Modification or enhancement is needed for continuous algorithm to solve combinatorial optimization problem.
Purpose – The purpose of the research is to solve Travelling Salesman Problem (TSP) using Simulated Kalman Filter (SKF) algorithm and single-solution SKF (ssSKF) algorithm based on numerical ordering technique.
Methodology/approach – The aim of the TSP is to find the optimal path between the starting city to the end city. The solution of TSP can represent into a sequence of permutation number where each number act as a city. In this research, each solution is encoded as n-number vector, each within the [-1,1] range, where n refers to the number of cities. To determine the travel sequence, the encoded solution is sorted in ascending order. The city with the smallest number will be visited first. This ordered series defines the travel sequence.
Findings – The results of the SKF algorithm and the ssSKF algorithm will be evaluated to decide which algorithm is better at solving this type of problem. The results of the experiment show that the SKF algorithm does better than the ssSKF algorithm in solving TSP.
Research limitations – The experiment of this research only uses 15 types of TSP function which are TSP functions that are evaluated using Euclidean distance.
Originality/value – One of the approaches that is yet to be applied to the SKF and ssSKF algorithms for solving combinatorial optimization problems is numerical ordering. The technique will be extended to the SKF and ssSKF algorithms to solve TSP in this study. |
|---|