Minimum-cost rapid-growing random trees for segmented assembly path planning

In this paper, the segmented assembly path-planning problem is proposed to reveal an integrated assembly path-planning task. We define the two phases of an assembly path-planning task and address their differences. A new method based on rapid-growing random trees (RRT) is presented for obtaining a minimum-cost path and making the total cost of the solution path remain constant for solving the same path-planning tasks in different iterations. The approach is also able to guarantee a solution with smooth rotation motions. We test this algorithm in various examples with good performance. And we also discuss the impact of the random sample set size of the algorithm in path optimization and execution time in various situations including segmented assembly path-planning task. Based on the results from the tests, suggestions for selecting a proper random sample set size are given.