A Knowledge-guided Pareto-based Evolutionary Algorithm for Welder-robot Cooperation Scheduling in Distributed Welding Shop under Industry 5.0

Moving towards human-centric manufacturing, Industry 5.0 prioritizes the synergistic cooperation between humans and robots, shifting scheduling from relying exclusively on machine resources to actively incorporating worker resources. Meanwhile, with the ongoing development of globalization, distributed welding shops have become prevalent in the real-world manufacturing scenario. Under Industry 5.0, an efficient scheduling of welder-robot cooperation mode in a distributed welding shop can improve productivity and flexibility. However, the existing literature lacks discussions on welder-robot cooperation scheduling in the distributed welding shop. Thus, this paper focuses on a welder-robot cooperation distributed welding shop scheduling problem (WCDWSP) aimed at optimizing both makespan and the total cost (TTC). To address this issue, a mathematical model and a knowledge-guided Pareto-based evolutionary algorithm (KPEA) are proposed. The novelties of KPEA are outlined as follows: 1) A hybrid initialization is designed to yield a high-quality initial population; 2) An evolutionary strategy is proposed to enhance KPEA's exploration capability; 3) Based on eight efficient neighborhood structures, a local search in the framework of variable neighborhood search (VNS) is devised to improve KPEA's exploitation capability; 4) By leveraging problem knowledge, a novel resource adjustment strategy is presented to optimize TTC without prolonging makespan. Finally, a series of comparative experiments is carried out to evaluate the effectiveness of KPEA. By benchmarking against six other algorithms, the results consistently highlight KPEA's superior performance over the competing methods.