A shortest path algorithm of long-period cyclic fully connected layer graph based on Dijkstra algorithm

• Published in: International Conference on Advanced Mechatronic Systems pp. 323 - 327
• Main Authors: Mi, Shanghua, Ji, Zengwei, Zheng, Hao, Gao, Yicong
• Format: Conference Proceeding
• Language: English
• Published: IEEE 10.12.2020
• Subjects: algorithm; Approximation algorithms; cyclic directed fully connected layer graph; global shortest path; Heuristic algorithms; Mechatronics; Optimization; Search problems; Shortest path problem; Viruses (medical)
• ISSN: 2325-0690
• DOI: 10.1109/ICAMechS49982.2020.9310152

For the shortest path problem of cyclic directed fully connected layer graph in several cycles, there is no excellent general solution algorithm in the literature. In this paper, we address this challenge by proposing a long-period decomposition algorithm based on the Dijkstra algorithm. This method employs periodic search primitives for continuous unidirectional path distance optimization. Thus the global shortest path of the long-period cyclic directed fully connected layer graph can be obtained by selecting primitives and decomposing repetitive cycles.. Experimental results demonstrate that state-of-the-art performance has achieved by our method in solving the optimal energy consumption of repeated picking and placing operations of industrial robots. Moreover, it can significantly reduce the computational complexity of the optimization process compared to the global solution using Dijkstra algorithm.