Modified Genetic Algorithm based on A* Algorithm of Multi Objective Optimization for Path Planning

Abstract

A new hybrid approach algorithm based on modified Genetic Algorithm (GA) and modified the search algorithm (A*) and has been developed to solve the Multi objectives global path planning (MOPP) problem for mobile robot navigation in complex environment with static distributed obstacles. The aim of this combination is to improve GA efficiency and path planning performance. Hence, several genetic operators are proposed based on domain-specific knowledge and characteristics of path planning to avoid falling into a local minimum in complex environment and to improve the optimal path partly such as deletion operator and enhanced mutation with basic A*. In addition, the proposed approach is received an initial population from a classical method or modified A*. The objective function for the proposed approach is to minimizing travelling distance, smoothness and security, without collision with any obstacle in the robot workspace. The simulation results show that the proposed approach is able to achieve multi objective optimization in complex static environment efficiently. Also, it has the ability to find a solution when the environment is complex and the number of obstacles is increasing.  Index Terms-multi objectives optimization, path planning, mobile robot, static complex environment, GA, A*.

Scalable Many-Objective Pathfinding Benchmark Suite

“…Weise et al used a customised NSGA-II algorithm for a multi-objective generation of wiring harnesses [26] and optimised length and the maximum heat for a wire which is confronted in a path. Oleiwi et al used a hybrid approach by modifying a genetic algorithm based on the A* algorithm [27]. Changan and Quiongbing used genetic algorithms and customised operators to work on the pathfinding problem [6], [7].…”

“…The angle θ is obtained by extending the line between two nodes and measuring the angle to the third node. Similar to [27], [14], we invert a • b = a b cos(θ): IV. BENCHMARK TEST SUITE Given the above model, we can generate several problem instances.…”

Different Nature-Inspired Techniques Applied for Motion Planning of Wheeled Robot: A Critical Review

“…Genetic Algorithms are a special class of Evolutionary Algorithms (EA) that utilizes approaches which is eager through evolutionary biology alike as inheritance, mutation, selection, and crossover [17]. Oleiwi and Roth, [18] have considered a complex terrain consists of static distributed obstacles, and figured out a multi-object global route selection challenge for wheeled robot navigation.…”

Multi Objective Optimization of Trajectory Planning of Non-holonomic Mobile Robot in Dynamic Environment Using Enhanced GA by Fuzzy Motion Control and A*

“…The remaining parameters are defined as follow [16,17], Where, F 1 (P) is the total length of path and criteria of path, F 2 (P) is the path smoothness, F 3 (P) is the path clearance or path security. ) P ( F 4 represents the total consumed time for robot motion and it can obtained by:…”

“…In this paper, we extended our approach in [16,17], Here we address the multi objective optimization problem (MOOP) of mobile robot navigation and obstacles avoidance. The formulation of the problem is concerned with finding three important steps.…”