Welding Path Planning Of Automotive Trim Based On Improved Fruit Fly Optimization Algorithm

With the development of economy and the improvement of living standards,China's automobile manufacturing industry has developed vigorously.In order to increase production and quality,domestic enterprises have adopted robots for welding operations.In order to complete the welding task in a shorter time,it is necessary to optimize the welding path of the robot and allocate the welding task reasonably.In traditional automobile manufacturing industry,robot welding path planning is completed manually,which not only consumes manpower and time,but also results in welding paths that often do not meet the requirements of welding.Therefore,the research on welding path planning is of great significance to ensure production safety,improve product quality and productivity.First of all,this thesis select and improve the fruit fly optimization algorithm.According to the production process and structural characteristics of the flexible welding production line of automobile parts,after comparing various optimization algorithms,the fruit fly optimization algorithm was selected to solve the welding path planning problem.The original fruit fly optimization algorithm does not have global search ability and is easy to fall into local optimum.In this thesis a Modified Global Fruit Fly Optimization Algorithm(MGFOA)is proposed by adding a uniform candidate solution generation mechanism,adaptive adjustment of flight radius and elite fruit flies.Based on 12 Benchmark functions,the comparison between MGFOA and other optimization algorithms is carried out to verify the effectiveness and superiority of the improved fruit fly optimization algorithm.Secondly,the MGCOA is discretized and used to solve the single robot path planning problem.The TSP mathematical model was established,and the improved fruit fly optimization algorithm was discretized by designing flight operators and elite fruit flies.The discrete improved fruit fly optimization is compared with genetic algorithm,particle swarm algorithm and simulated annealing algorithm based on TSBLIB database.Then taking the front door and the back door of the automobile as the welding object,the planning simulation of the welding path planning of a single robot is realized,and the effectiveness of the discrete improved fruit fly optimization algorithm is verified.Finally,the problem of multi-robot collaborative welding path planning is studied.Analyze the constraints and path planning goals of multi-robot collaborative welding.A dynamic planning multi-fruit fly group collaborative planning algorithm was designed.The solution process is decomposed into three stages: initial distribution of welding spots,coordinated planning of multiple fruit fly groups and anti-collision detection.The welding point adjustment operator and the anti-collision method with the same direction welding are designed.With the front door and the rear door of the car as the welding objects,the welding path planning for the four robots is realized.The simulation results demonstrate the effectiveness of the proposed dynamic programming multi-fruit fly group collaborative planning algorithm.