Multi-agent scheduling of social manufacturing services considering crowdsourced transportation time

Scheduling of manufacturing services is an important element in social manufacturing, and a good scheduling solution can improve resource utilization and productivity. Meanwhile, under the orientation of Industry 5.0, issues such as the enhancement of scheduling system toughness have become a research hotspot. Therefore, this research paper proposes a multi-agent based toughness scheduling system scheme for social machining tasks under crowdsourced transportation. The system employs several different Agents for the management and monitoring of the task scheduling process, adopts the improved artificial hummingbird algorithm (IAHA) for the two-phase toughness scheduling strategy with static pre-planning and dynamic adjustment, and predicts the crowdsourced transportation time by using the LSTM. The IAHA uses the initialization rules with global selection, local selection, and inverse learning strategies combined with the Levy flight and the monomorphic search strategies. The simulation cases with improved standard Mk arithmetic are used for testing and comparison experiments with other algorithms are conducted to verify the rationality of the scheduling system model and the feasibility and high efficiency of the method in this paper. And it can effectively cope with the random disturbances such as the failure of the machine, the insertion of new tasks and the weather change of transportation to ensure the production schedule and realize the requirements of a certain degree of Industry 5.0 resilience.