Research On The Algorithm Of PCB Borads Automatic Soldering Robot Path Optimization

Equipment manufacturing industry is the core and foundation of industry. The PCB industry is very large in scale but not strong in china. The PCB soldering robot is one of the high-end production equipment. It is a time-consuming job to manually teach PCB soldering robot by skilled technicians for the best soldering path. An application system is developed to solve this problem that computer program can read PCB project file automatically, using intelligent optimization algorithms of automatic optimization of PCB soldering path. It will be of great significance to reduce time-consuming of PCB soldering and enhance PCB industry productivity.In this paper, the mathematical model of the path optimization problem of the automatic soldering robot is established. By in-depth analysis of PCB board production line soldering process, corresponding to the initial setting, the mathematical model of PCB soldering path optimization can applied to different applications. Moreover, through the analysis and comparison of path optimization algorithm, we choose ant colony algorithm to solve the problem.In the research of optimization algorithm, firstly, we propose an improved small window ant colony algorithm——small window of random ant colony algorithm, since small window ant colony algorithm has poor adaptability and fall into local optimum easily. We bring the scale of the problem and randomness in the small window ant colony algorithm to enhance the robustness of the algorithm. Through eight sets of standard test function, it is proved that the algorithm can effectively improve the ability of searching and void early maturing algorithm and trapping in a local optimum. Secondly, for solving the deficiency lying of cross links found in the result of small window of random ant colony algorithm, the method is introduced to eliminate the cross links in the optimization algorithm. Analyzing test results, obtaining and selecting a better way to eliminate cross links through the study of the number and location of the eliminate crossover point. A new hybrid ant colony algorithm is developed to settle the soldering robot path optimization problem, its effectiveness and feasibility is proved by simulation tests.Finally the application system of soldering robot path optimization is developed, the system program is written by C++, its function include:automatically read PCB project file; data acquisition of solder joint;adjust the parameter optimization algorithm by the man-machine interface; using the hybrid ant colony algorithm optimizing path; output best soldering path. The manual teaching soldering path link can bereplaced by soldering robot path optimization application system.