A Multi-Edge Task Unloading Method in Manufacturing Shops Based on Dynamic Cooperative Particle Swarm Optimization

Abstract

This invention discloses a method for offloading multiple edge tasks in a manufacturing workshop based on dynamic cooperative particle swarm optimization, relating to the field of edge computing and heuristic algorithms in intelligent manufacturing workshops. By introducing a hybrid initialization strategy, a dynamic subgroup cooperative update mechanism, and a mutation mechanism, this invention comprehensively improves the performance of particle swarm optimization algorithms in solving the multi-edge task offloading problem in manufacturing workshops from three levels: initialization, global search, and local search. It improves the quality of initial solutions and achieves dynamic adjustment of the evolution strategy, avoiding local optima in the optimization results, thereby further improving the efficiency of multi-edge task offloading in manufacturing workshops.

Images