Yumin Ma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yumin Ma has a total of 42 co-authors. Here are the top co-authors by number of shared publications:
- Fei Qiao
Tongji University
22 shared publications - Juan Liu
College of Electronic and Information Engineering, Tongji University, Shanghai, China
10 shared publications - Kuo-Yi Lin
Department of Industrial Engineering and Management, National Kaohsiung University of Science and Technology, No. 415, Jiangong Road, Sanmin District, Kaohsiung City, Taiwan
8 shared publications - Li Li
Department of Control Science and Engineering, College of Electronics and Information Engineering, State Key Laboratory of Autonomous Intelligent Unmanned System, Frontiers Science Center for Intelligent Autonomous Systems, Ministry of Education, Shanghai Research Institute for Intelligent Autonomous Systems, and Shanghai Institute of Intelligent Science and Technology, Tongji University, Shanghai, China
8 shared publications Qingyun Yu
Tongji University
8 shared publications- Jiaxuan Shi
Shanghai Research Institute for Intelligent Autonomous Systems, TongJi University, Shanghai, China
4 shared publications
Latest Publications
- Open Access
- 21 May 2026
- Published by Elsevier in SSRN Electronic Journal
Dynamic Scheduling of Aircraft Pulsed Assembly Lines under Rework Events: A QMIX-GA-Based Multi-Agent Reinforcement Learning Method
Aircraft assembly is a zero-defect-oriented manufacturing process with strict safety, reliability, and delivery requirements. In practice, rework events are major corrective responses to assembly defects, especially when high-value components cannot be scrapped. However, inserting rework events into aircraft pulsed assembly lines may disrupt production plans, increase takt instability, and create new challenges for dynamic scheduling and collaborative decision-making. To address this problem, this paper investigates dynamic scheduling of aircraft pulsed assembly lines under rework events and proposes a rework-aware distributed scheduling method based on QMIX-GA. A hierarchical multi-agent framework is developed, in which each workstation is managed by a workstation agent composed of a Scheduling Agent, a Task Sequencing Agent, and a Worker Dispatching Agent. QMIX coordinates intra-workstation operation sequencing and worker dispatching through an event-driven multi-agent decision process, while a genetic algorithm evaluates cross-workstation rework transfer feasibility and guides tail-end rework handling. Thus, rework impact is first absorbed through intra-workstation rescheduling, and cross-workstation collaboration is activated only when local adjustment cannot maintain takt feasibility. Experiments on five benchmark datasets under diverse rework scenarios show that the proposed method reduces completion-time mean, variance, and tardiness rate, verifying its effectiveness.
- 21 May 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDynamic Scheduling of Aircraft Pulsed Assembly Lines under Rework Events: A QMIX-GA-Based Multi-Agent Reinforcement Learning Method
Aircraft assembly is a zero-defect-oriented manufacturing process with strict safety, reliability, and delivery requirements. In practice, rework events are major corrective responses to assembly defects, especially when high-value components cannot be scrapped. However, inserting rework events into aircraft pulsed assembly lines may disrupt production plans, increase takt instability, and create new challenges for dynamic scheduling and collaborative decision-making. To address this problem, this paper investigates dynamic scheduling of aircraft pulsed assembly lines under rework events and proposes a rework-aware distributed scheduling method based on QMIX-GA. A hierarchical multi-agent framework is developed, in which each workstation is managed by a workstation agent composed of a Scheduling Agent, a Task Sequencing Agent, and a Worker Dispatching Agent. QMIX coordinates intra-workstation operation sequencing and worker dispatching through an event-driven multi-agent decision process, while a genetic algorithm evaluates cross-workstation rework transfer feasibility and guides tail-end rework handling. Thus, rework impact is first absorbed through intra-workstation rescheduling, and cross-workstation collaboration is activated only when local adjustment cannot maintain takt feasibility. Experiments on five benchmark datasets under diverse rework scenarios show that the proposed method reduces completion-time mean, variance, and tardiness rate, verifying its effectiveness.
- 3 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 23 (15455955) , 1284-1300
Game Theory-Based Production-Maintenance Collaborative Scheduling Using Imitation-Enhanced Alternating-Training Reinforcement Learning
The collaborative organization of production scheduling and machine maintenance is crucial for achieving effective manufacturing. However, since these two activities are typically managed by different self-interested managers with distinct optimization needs in practice, their collaboration faces the challenge of simultaneous consideration and balanced optimization of both parties’ interests. To this end, this study investigates a novel game theory-based production-maintenance collaborative scheduling problem, wherein the typical objectives individually concerned by two activities are explicitly considered and the decision interactions between activities are modeled as a non-cooperative stochastic game. Nash equilibrium within the game model provides the ideal solution for balancing the interests of both parties. To solve this problem, an imitation-enhanced alternating-training reinforcement learning method is presented. In this method, an imitation learning-based initialization mechanism is designed to accelerate the training of two game agents, and an alternating training mechanism are developed to facilitate two agents efficiently learning the optimal responses to each other’s behaviors and ultimately achieving the decisions that converge to Nash equilibrium. The superiority of the proposed method is verified through comprehensive experiments.
- 3 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 23 (15455955) , 1284-1300
Game Theory-Based Production-Maintenance Collaborative Scheduling Using Imitation-Enhanced Alternating-Training Reinforcement Learning
The collaborative organization of production scheduling and machine maintenance is crucial for achieving effective manufacturing. However, since these two activities are typically managed by different self-interested managers with distinct optimization needs in practice, their collaboration faces the challenge of simultaneous consideration and balanced optimization of both parties’ interests. To this end, this study investigates a novel game theory-based production-maintenance collaborative scheduling problem, wherein the typical objectives individually concerned by two activities are explicitly considered and the decision interactions between activities are modeled as a non-cooperative stochastic game. Nash equilibrium within the game model provides the ideal solution for balancing the interests of both parties. To solve this problem, an imitation-enhanced alternating-training reinforcement learning method is presented. In this method, an imitation learning-based initialization mechanism is designed to accelerate the training of two game agents, and an alternating training mechanism are developed to facilitate two agents efficiently learning the optimal responses to each other’s behaviors and ultimately achieving the decisions that converge to Nash equilibrium. The superiority of the proposed method is verified through comprehensive experiments.
- Top Cited Papers
- 4 March 2025
- Published by Elsevier in Advanced Engineering Informatics
- Vol. 65
Production-logistics collaborative scheduling in dynamic flexible job shops using nested-hierarchical deep reinforcement learning
- 4 March 2025
- Published by Elsevier in Advanced Engineering Informatics
- Vol. 65
Top Cited PapersProduction-logistics collaborative scheduling in dynamic flexible job shops using nested-hierarchical deep reinforcement learning
- 29 November 2024
- Published by Elsevier in Expert Systems with Applications
- Vol. 264
A new data-driven production scheduling method based on digital twin for smart shop floors
- 29 November 2024
- Published by Elsevier in Expert Systems with Applications
- Vol. 264
A new data-driven production scheduling method based on digital twin for smart shop floors
- 14 July 2024
- Published by Elsevier in Computers & Industrial Engineering
- Vol. 194
A semi-supervised production scheduling method based on co-training deep neural network for smart shop floors
- 14 July 2024
- Published by Elsevier in Computers & Industrial Engineering
- Vol. 194
A semi-supervised production scheduling method based on co-training deep neural network for smart shop floors
- 7 August 2023
- Published by Springer Nature in Neural Computing and Applications
- Vol. 35 (30) , 22281-22296
Double deep Q-network-based self-adaptive scheduling approach for smart shop floor
- 7 August 2023
- Published by Springer Nature in Neural Computing and Applications
- Vol. 35 (30) , 22281-22296
Double deep Q-network-based self-adaptive scheduling approach for smart shop floor
- 21 May 2023
- Published by Springer Nature
- p. 135-182
Data-Driven Release Control of Semiconductor Manufacturing System
- 21 May 2023
- Published by Springer Nature
- p. 135-182
Data-Driven Release Control of Semiconductor Manufacturing System