Abstract: In order to reduce production cost and increase productive efficiency, working positions
of assembling robots need to be optimized to reduce the number of robots and the times of changing
working positions. A hybrid optimization algorithm combining clustering algorithm and simulated
annealing algorithm was proposed to optimize working positions of robots. Firstly, processing
objects with common available working positions of robots were clustered by the clustering
algorithm in order to reduce the scale of the problem. Then, a simulated annealing algorithm with
multi-solution parallel search was employed to search for the minimum working positions of robots
available for processing all objects. Testing results show that this algorithm is capable of improving
optimization result significantly as well as accelerating the convergence of the algorithm
remarkably.

Key words: robot, working position, clustering algorithm, simulated annealing algorithm