Abstract
A novel algorithm (CP-QRRT*) is proposed for the path planning tasks of hyper-redundant manipulators (HRMs) in confined spaces, addressing the issues of unmet joint angle constraints, redundant planning paths, and long planning times present in previous algorithms. First, the PSO algorithm is introduced to optimize the random sampling process of the RRT series algorithms, enhancing the directionality of the random tree expansion. Subsequently, the method of backtracking ancestor nodes from the Quick-RRT* algorithm is combined to avoid getting trapped in local optima. Finally, a constraint module designed based on the maximum joint angle constraints of the HRM is implemented to limit the path deflection angles. Simulation experiments demonstrate that the proposed algorithm can satisfy the joint angle constraints of the HRM, and the planned paths are shorter and require less time.