Hybrid Assembly Path Planning for Complex Products by Reusing a Priori Data

Abstract

Assembly path planning (APP) for complex products is challenging due to the large number of parts and intricate coupling requirements. A hybrid assembly path planning method is proposed herein that reuses a priori paths to improve the efficiency and success ratio. The assembly path is initially segmented to improve its reusability. Subsequently, the planned assembly paths are employed as a priori paths to establish an a priori tree, which is expanded according to the bounding sphere of the part to create the a priori space for path searching. Three rapidly exploring random tree (RRT)-based algorithms are studied for path planning based on a priori path reuse. The RRT* algorithm establishes the new path exploration tree in the early planning stage when there is no a priori path to reuse. The static RRT* (S-RRT*) and dynamic RRT* (D-RRT*) algorithms form the connection between the exploration tree and the a priori tree with a pair of connection points after the extension of the exploration tree to a priori space. The difference between the two algorithms is that the S-RRT* algorithm directly reuses an a priori path and obtains a new path through static backtracking from the endpoint to the starting point. However, the D-RRT* algorithm further extends the exploration tree via the dynamic window approach to avoid collision between an a priori path and obstacles. The algorithm subsequently obtains a new path through dynamic and non-continuous backtracking from the endpoint to the starting point. A hybrid process combining the RRT*, S-RRT*, and D-RRT* algorithms is designed to plan the assembly path for complex products in several cases. The performances of these algorithms are compared, and simulations indicate that the S-RRT* and D-RRT* algorithms are significantly superior to the RRT* algorithm in terms of the efficiency and success ratio of APP. Therefore, hybrid path planning combining the three algorithms is helpful to improving the assembly path planning of complex products.

Recent advances in Rapidly-exploring random tree: A review

“…[ 160 ], the nominal mean value of the stochastic control distribution in the model predictive path integral is provided by RRT, leading to satisfactory control performance in both static and dynamic environments without any parameter fine-tuning. Other applications include manipulator dynamic obstacle avoidance [ 161 ], hybrid assembly path planning for complex products [ 162 ], 10-DOF rover traversing over 3D uneven terrains [ 163 ], UAV path planning [ 77 , 151 , 164 ], electric inspection robot navigation [ 165 ], cobot in dynamic environment [ 166 ], underground vehicles [ 167 ], automated guided vehicle [ 145 ], mining truck [ 143 ], redundant robots [ 168 ].…”

An Integrated RRT*SMART-A* Algorithm for solving the Global Path Planning Problem in a Static Environment

“…RRT*-Smart is an enhanced version of RRT* that works in the same way as RRT* for finding the initial path [14,20]. Besides that, it undertakes a path optimization procedure once an initial path has been determined.…”

“…Path planning is the basis of a robotic navigation, namely the ability to plan a path without collisions with obstacles in the environment [6][7][8]. Although there are several strategies with different exploration representations, grid-based algorithms [9][10][11] and sampling-based algorithms [12][13][14][15] are in great demand because of their ease of implementation. For example, there are A* and RRT* algorithms for these types, respectively.…”

Functional Representation of the Intentional Bounded Rationality of Decision-Makers: A Laboratory to Study the Decisions a Priori

“…Before choosing between discrete alternatives, the decision-makers have a starting point, prior knowledge/beliefs about the expected performance of alternatives, and the possibility of processing signals (information) to improve their prediction [18]. However, processing signals until certainty is achieved about the performance of each alternative is costly, due to inherent human limitations, which means that in practice some uncertainty persists about the alternatives' performances when a decision-maker chooses between them.…”