Elsevier

Neurocomputing

Volume 120, 23 November 2013, Pages 509-517
Neurocomputing

An improved genetic algorithm with co-evolutionary strategy for global path planning of multiple mobile robots

https://doi.org/10.1016/j.neucom.2013.04.020Get rights and content

Abstract

This paper presents a Co-evolutionary Improved Genetic Algorithm (CIGA) for global path planning of multiple mobile robots, which employs a co-evolution mechanism together with an improved genetic algorithm (GA). This improved GA presents an effective and accurate fitness function, improves genetic operators of conventional genetic algorithms and proposes a new genetic modification operator. Moreover, the improved GA, compared with conventional GAs, is better at avoiding the problem of local optimum and has an accelerated convergence rate. The use of a co-evolution mechanism takes into full account the cooperation between populations, which avoids collision between mobile robots and is conductive for each mobile robot to obtain an optimal or near-optimal collision-free path. Simulations are carried out to demonstrate the efficiency of the improved GA and the effectiveness of CIGA.

Introduction

Global path planning is a hot issue in many fields, especially in the application of robot navigation, which means to figure out an optimal or near-optimal collision-free path from a start location to a goal location in an environment with obstacles. The existing path planning methods are mainly divided into two major categories: traditional methods and intelligent methods. Prominent traditional planning methods include the global visibility graph algorithm [1] and potential field methods [2]. Intelligent planning methods include fuzzy logic [3], neural networks [4], [5], ant colony algorithms [6], genetic algorithms [7] and particle swarm optimization algorithms [8]. Each method has its own limitations and so far, there is not any method that can completely solve robot path planning. So in recent years, researchers have been persistently seeking for new more effective and efficient solutions or improving existing methods.

Genetic Algorithms (GAs) have been proven powerful for optimization problems thanks to their ability to find global optima and high parallelism [9], [10], [11]. Many research studies have been carried out in recent years which attempt to solve the robot path planning problem by using genetic algorithms. A path planning method for mobile robots based on an adaptive genetic algorithm is proposed in [12]. A dynamic robot path planning scheme for unknown environments is presented in [13]. Tsai et al. [14] proposed a parallel elite genetic algorithm for global robot path planning, and Chiu [15] presented a numerical assessment of path planning for an autonomous robot passing through multi-layer barrier systems using a genetic algorithm. For multiple robots path planning, an improved knowledge-based genetic algorithm was presented in [16], a knowledge-based genetic algorithm for on-line path planning of multiple mobile robots in a dynamic environment was proposed in [17]. Kala [18] presented a co-evolutionary genetic programming approach to solve multi-robot path planning, which involves a different source and goal for each robot.

However, conventional genetic algorithms has the disadvantages of slow convergence rate, local optimum and ignoring cooperation between populations [14]. Co-evolution is the process of mutual adaptation of two or more populations, and it is used to reflect the fact that all species are simultaneously co-evolving in a given physical environment [19]. Thus, a co-evolution mechanism is a potential way to improve on the shortcomings of the conventional genetic algorithms, and some researchers have applied it to solving constrained problems [20], [21] and multi-objective optimization problems [22], [23]. In this paper, to circumvent these shortcomings of conventional genetic algorithms, we present a co-evolution based improved GA, called the Co-Evolutionary Improved Genetic Algorithm (CIGA), and apply it to solving the global path planning problem of multiple mobile robots effectively.

The main contributions of this paper are (1) An improved GA is presented to solve the global path planning problem for a single mobile robot; (2) A CIGA based on the improved GA with a co-evolution mechanism is proposed for multiple mobile robots path planning. The proposed improved GA is helpful not only for avoiding local optima, but also for accelerating the convergence rate. The CIGA will be proven effective by experiments in a simulated environment.

The remaining part of this paper is organized as follows. Section 2 describes the process of the improved GA for solving global path planning. Section 3 presents CIGA which is based on the improved GA and describes how to apply CIGA to solving global path planning of multiple mobile robots. Section 4 shows the results of a simulation to demonstrate the performance of the proposed algorithms. Section 5 contains conclusion and remarks.

Section snippets

Improved GA for global path planning

In this section, an effective Improved Genetic Algorithm (IGA) for solving the global path planning problem for a mobile robot is presented. The proposed IGA is different from conventional GAs in its fitness function, selection operator and a novel operator named modification. In addition, other genetic operators, such as crossover and mutation, are also slightly modified. A more accurate fitness function which considers three different variables and distinction of feasible and infeasible paths

CIGA for global path planning of multiple mobile robots

CIGA needs to solve four problems consisting of algorithm initialization, evolution algorithm of each subpopulation, the design of subpopulation fitness function and the design of an information interaction operation. Normally, a conventional GA is used as the evolution algorithm for each subpopulation. In this paper, we use the improved GA instead, which can achieve better performance.

Simulation

Three experiments were made in order to demonstrate the feasibility and effectiveness of the proposed algorithms. They were compiled using Microsoft Visual C++ 6.0. The first experiment simulates mobile robot path planning using the improved GA. The second experiment compare its performance to that of the conventional GA and Ant Colony Optimization(ACO), while the third experiment simulates global path planning of multiple mobile robots using CIGA.

Conclusions and outlook

Global path planning of multiple mobile robots means to find a collision-free path for each robot while avoiding collisions between them. In this paper, an improved GA based on the conventional GA for global path planning was proposed. This was in turn used as the basis for CIGA, which utilizes a co-evolution mechanism, for global path planning of multiple robots. It has been shown that the proposed algorithms are capable of efficiently guiding the mobile robot travelling from the starting

Hong Qu received the B.S. degree and the M.S.and Ph.D. degrees in computer science and engineering from the University of Electronic Science and Technology of China, Chengdu, China, in 2000, 2003 and 2006, respectively. From March 2007 to February 2008, he was a Postdoctoral Fellow at the Advanced Robotics and Intelligent Systems Lab, School of Engineering, University of Guelph, Guelph, ON, Canada. Currently, he is a professor in the School of Computer Science and Engineering, University of

References (26)

  • Xiao-Ping Zeng et al.

    A dynamic chain-like agent genetic algorithm for global numerical optimization and feature selection

    Neurocomputing

    (2009)
  • Yong Zhang et al.

    Robot path planning in uncertain environment using multi-objective particle swarm optimization

    Neurocomputing

    (2013)
  • C.A.E. Poulos et al.

    Path planning for a mobile robot

    IEEE Trans. Syst. Man Cybern.

    (1992)
  • J. Barraquand et al.

    Numerical potential field techniques for robot path planning

    IEEE Trans. Syst. Man Cybern.

    (1992)
  • G.Z. Panagiotis et al.

    Motion control for mobile robot obstacle avoidance and navigationa fuzzy logic-based approach

    Syst. Anal. Model. Simulat.

    (2003)
  • Howard Li et al.

    Neural-network-based path planning for a multirobot system with moving obstacles

    IEEE Trans. Syst. Man Cybern.—Part CAppl. Rev.

    (2009)
  • Hong Qu et al.

    Real-time robot path planning based on a modified pulse-coupled neural network model

    IEEE Trans. Neural Networks

    (2009)
  • Shirong Liu, Linbo Mao, Jinshou Yu, Path planning based on ant colony algorithm and distributed local navigation for...
  • H.F.D. Alexandre et al.

    Multiobjective genetic algorithms applied to solve optimization problems

    IEEE Trans. Magn.

    (2002)
  • M.A. Bayir et al.

    Genetic algorithm for the multiple-query optimization problem

    IEEE Trans. Syst. Man Cybern.—-Part C: Appl. Rev.

    (2007)
  • M. Cmara et al.

    A single front genetic algorithm for parallel multi-objective optimization in dynamic environments

    Neurocomputing

    (2009)
  • Liu Chuangling et al.

    A path planning method based on adaptive genetic algorithm for mobile robot

    J. Inf. Comput. Sci.

    (2011)
  • Pu Shi, Yujie Cui, Dynamic path planning for mobile robot based on genetic algorithm in unknown environment, in: 2010...
  • Cited by (210)

    • A systematic review on recent advances in autonomous mobile robot navigation

      2023, Engineering Science and Technology, an International Journal
    View all citing articles on Scopus

    Hong Qu received the B.S. degree and the M.S.and Ph.D. degrees in computer science and engineering from the University of Electronic Science and Technology of China, Chengdu, China, in 2000, 2003 and 2006, respectively. From March 2007 to February 2008, he was a Postdoctoral Fellow at the Advanced Robotics and Intelligent Systems Lab, School of Engineering, University of Guelph, Guelph, ON, Canada. Currently, he is a professor in the School of Computer Science and Engineering, University of Electronic Science and Technology of China. His current research interests include neural networks, robot, neurodynamics, intelligent computation, and optimization.

    Ke Xing He is a postgraduate student of computer science and engineering, University of Electronic Science and Technology of China. His current research interests relate to genetic algorithm, ant colony optimization and mobile robot path planning.

    Takacs Alexander He is a final year student of the theoretical computer science master's program at the School of Computer Science and Communication of the Royal Institute of Technology, Stockholm. He is currently writing his thesis at the University of Electronic Science and Technology of China, Chengdu. His research interests include shortest path ranking and related graph problems.

    This work was supported by National Science Foundation of China under Grant 61273308 and the Fundamental Research Funds for Central Universities under Grant ZYGX2012J068.

    View full text