A Knowledge-Driven Cooperative Meta-Heuristic Algorithm for Multiplex Industrial Chain Networks Design with Multiple Supply Cycles

The current industrial chain structure is gradually transitioning from a single industrial chain network to multiplex industrial chain networks. The design of multiplex industrial chain networks is crucial for resource allocation within the industrial chain. In this paper, the design problem of multiplex industrial chain networks with multiple supply cycles (MICND-MSC) is studied, and the impact of cross-chain supply of multiple raw materials by raw material manufacturing enterprises is considered in the problem. The mixed-integer linear programming (MILP) model of MICND-MSC is constructed, and a knowledge-driven cooperative meta-heuristic algorithm (KDCMA) that integrates the framework of the artificial bee colony (ABC) algorithm and reinforcement learning mechanism is proposed to address the MICND-MSC. Multiple heuristic methods are designed to generate potential solutions for the initial population. The neighborhood structures for different production stages are utilized in the KDCMA to explore the solution space of MICND-MSC. The reinforcement learning mechanism is employed to learn empirical knowledge of neighborhood structures to guide the search process. The experimental results indicate that the KDCMA is a potential algorithm to address the MICND-MSC.