Abstract: The manufacturing system is undergoing a profound transformation towards a customer-centric flexible manufacturing model, progressively evolving into multi-stage integration and distributed networking. This evolution necessitates integrated decision-making to address complex multi-dimensional objectives. Traditional scheduling methods, which rely on algorithms or machine intelligence, face limitations when it comes to addressing nonlinear constraints and accurately quantifying intricate business-related lower-bound requirements. Additionally, decision-makers often face challenges in anticipating and intervening to mitigate potential sunk costs. To address these challenges, a human-machine interactive scheduling technology for complex manufacturing systems is proposed. The proposed approach aims to enable flexible control over optimization through human-machine interaction mechanisms. By combining business requirements with decision-maker preferences, the solution space is pre-optimized and partitioned to improve efficiency, achieving scheduling and optimization through human-machine collaboration. For an appliance manufacturing system with intensive production and diverse delivery requirements, a two-stage interactive scheduling method is proposed. This approach balances business conflicts and reflects decision-maker preferences through constraint boundary adjustments. A constraint programming model is developed, along with neighborhood structures tailored to different objective orientations. Finally, the proposed method is validated using historical case data from a home appliance enterprise. The results demonstrate that the proposed scheme exhibits high reliability and gains user acceptance, confirming its effectiveness.

Key words: human-centric smart manufacturing, complex manufacturing system, human-machine interactive scheduling, human-machine collaborative optimization, flexible constraints