Mobile robot path planning using fuzzy enhanced improved Multi-Objective particle swarm optimization (FIMOPSO)

•Efficient algorithm is needed for cracking Car-like Robot Path Planning problems.•An Improved Multi-objective PSO algorithm is proposed and compared with MOSPEA2.•Five objective functions are considered in solving multi-objective CRPP problems.•Totally six experiments were conducted to validate the...

Full description

Saved in:
Bibliographic Details
Published in:Expert systems with applications Vol. 198; p. 116875
Main Authors: Sathiya, V., Chinnadurai, M., Ramabalan, S.
Format: Journal Article
Language:English
Published: New York Elsevier Ltd 15.07.2022
Elsevier BV
Subjects:
Car-like robot path planning
Collision avoidance
Evolutionary algorithms
FIMOPSO
Inference
MOSPEA2
Multiple objective analysis
Non-holonomic and kinodynamic constraints
Obstacle avoidance
Particle swarm optimization
Path planning
Robots
Static and dynamic environments
Travel time
MOSPEA2
Car-like robot path planning
Non-holonomic and kinodynamic constraints
Static and dynamic environments
FIMOPSO
ISSN:0957-4174, 1873-6793
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Efficient algorithm is needed for cracking Car-like Robot Path Planning problems.•An Improved Multi-objective PSO algorithm is proposed and compared with MOSPEA2.•Five objective functions are considered in solving multi-objective CRPP problems.•Totally six experiments were conducted to validate the proposed strategy.•Moving obstacles around the robot and a dynamic goal point were considered. This paper introduces a method for car-like mobile robot path planning (CRPP). The robot works in both dynamic and static situations. The aim of this method is to explore the best safe path with minimum path length, minimum motor torque, minimum travel time, minimum robot acceleration and maximum obstacle avoidance. Kinodynamic and non-holonomic constraints related with car-like robot are considered. Fuzzy enhanced Improved Multi-objective Particle Swarm Optimization (FIMOPSO) algorithm is proposed to solve the CRPP problem. Fuzzy inference system is used for obstacle avoidance. In the proposed FIMOPSO, five improvements are made. Proposed technique is compared with Multi-objective Strength Pareto Evolutionary Algorithm 2 (MOSPEA2) technique. Experiments on a custom-made car-like robot are ensuring the quality of proposed technique. This research works show that proposed FIMOPSO is another alternative technique to CRPP problems. Paths dictated by FIMOPSO are safe, collision free, feasible, and possible and can be practically implemented. Fuzzy inference system works well for safe robot travel. FIMOPSO simulation paths are acceptable. Since, the deviation between experiment and simulation is less than 2%.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0957-4174
1873-6793
DOI:10.1016/j.eswa.2022.116875