Research Of Path Planning For Welding Robots Based On Hybrid Discrete Particle Swarm Optimization Algorithm

Welding is an important sector of industrial manufacturing and processing. Welding quality generally plays a decisive role for obtaining product quality. Robot could accomplish a variety of tasks with high strength, speed and precision, and the extensive applications of welding robot have greatly improved the production efficiency. Rational path planning of welding robot can greatly reduce production time of work station, shorten the overall working hours, improve production efficiency and reduce production costs. In the welding process, a robot could be allocated to a large number of solder joints, so a good path planning methods is of vital significance to the efficiency of welding robot.Particle swarm optimization algorithm (PSO) is a bionic optimization algorithm developed rapidly in the last ten years and has demonstrated excellent performance and great potential in solving many complex problems. This paper mainly studies three-dimensional path planning of a single welding robot which has been allocated to given welding task. Based on the comprehensive analysis of the current various path planning methods, this paper selects particle swarm optimization algorithm to solve welding robot path planning problem.First, an improved particle swarm optimization algorithm based on two kinds of chaotic maps is proposed because in some complex applications the traditional PSO was easy to fall into local minimum and cannot find the global optimum of the problem. Then the model of three-dimensional path planning problem, traveling salesman problem, is analyzed in this paper. By combining light ray optimization, genetic algorithm and chaos optimization strategies, a novel hybrid discrete PSO algorithm (HDPSO) is proposed and its performance in solving TSP problem is verified by simulation test. Afterwards performance of HDPSO in solving three-dimensional path planning problem is verified with2Body in White production cases. Finally, a welding robot path planning simulation system is designed by Visual C#2010, and its practicality is verified by testing a Body in White production case.