A Novel Collaborative Evolutionary Algorithm with Two-Population for Multi-Objective Flexible Job Shop Scheduling

Abstract

Job shop scheduling (JS) is an important technology for modern manufacturing. Flexible job shop scheduling (FJS) is critical in JS, and it has been widely employed in many industries, including aerospace and energy. FJS enables any machine from a certain set to handle an operation, and this is an NP-hard problem. Furthermore, due to the requirements in real-world cases, multi-objective FJS is increasingly widespread, thus increasing the challenge of solving the FJS problems. As a result, it is necessary to develop a novel method to address this challenge. To achieve this goal, a novel collaborative evolutionary algorithm with two-population based on Pareto optimality is proposed for FJS, which improves the solutions of FJS by interacting in each generation. In addition, several experimental results have demonstrated that the proposed method is promising and effective for multi-objective FJS, which has discovered some new Pareto solutions in the well-known benchmark problems, and some solutions can dominate the solutions of some other methods.

A Strengthened Dominance Relation NSGA-III Algorithm Based on Differential Evolution to Solve Job Shop Scheduling Problem

“…Xu et al [14] presented a three-layer coded hybrid algorithm to solve MO-JSSP, accounting for job outsourcing and carbon emission, effectively addressing pressing environmental issues. Wang et al [15] proposed a multiobjective dual-population algorithm capable of continuously improving scheduling solutions during optimization interaction. Based on previous research, it has been found that many algorithms contain a significant number of simple crossover and mutation operations, resulting in inefficient search patterns.…”

“…The dominance region of x consists of two parts, as given by Eq. (15). Individuals within the dominant region will be selected for the next iteration.…”

A review of reinforcement learning based hyper-heuristics

“…With the capability to explore complex search spaces, EAs offer remarkable adaptability and parallelization ease ( Young et al, 2015 ). At present, EAs find extensive application across diverse research domains, including sequence optimization ( Lutz et al, 2023 ; Junior et al, 2023 ), scheduling ( Chen et al, 2023 ; Wang, Li & Gao, 2023 ; Wang et al, 2023 ), object recognition ( Afif et al, 2023 ; Zhang, Li & Qi, 2023 ).…”

Research on integrated scheduling of equipment predictive maintenance and production decision based on physical modeling approach

“…Production decisions are primarily concerned with the rational allocation of resources and the sequencing of processes to optimise the desired production goals. Rational production decisions can improve efficiency, reduce costs and energy consumption, and ensure product quality and delivery [20].…”