基于MO-SAC算法的工业软件组件边缘集群多目标决策部署
Multi-Objective Decision-Making Deployment on Edge Clusters of Industrial Software Components Based on MO-SAC Algorithm
DOI: 10.12677/csa.2025.154089, PDF,    国家自然科学基金支持
作者: 徐沛沅, 王 涛：广东工业大学自动化学院，广东 广州；广东省信息物理融合系统实验室，广东 广州；程良伦：广东省信息物理融合系统实验室，广东 广州；广东工业大学计算机学院，广东 广州
关键词: 工业物联网；边缘集群；分布式部署；MO-SAC算法；Industrial Internet of Things； Edge Cluster； Distributed Deployment； MO-SAC Algorithm
摘要: 随着工业4.0的发展，工业物联网(IIoT)中软件组件的分布式部署已成为提高生产效率和灵活性的关键手段，通过在边缘端进行数据处理和分析，能够实现更低的延迟和更高的可靠性。尽管分布式部署带来了诸多优势，但仍面临任务需求复杂，节点资源受限，响应时间要求严格、集群负载不均衡以及环境动态变化等挑战，限制了其在实际工业环境中的应用。为了在任务响应时间和受限的资源分配中寻找最优解，并保证负载的均衡与稳定，本文提出了一种工业软件组件应用程序分布式边缘部署的系统架构，并且利用改进的MO-SAC算法对工作流进行分布式部署，引入温度系数，鼓励模型探索更多更广的部署策略，避免陷入局部最优。大量仿真结果表明MO-SAC可以在有效降低任务响应时间的同时最小化资源消耗并且使集群的负载更加均衡。
Abstract: With the development of Industry 4.0, the distributed deployment of software components in the Industrial Internet of Things (IIot) has become a key means to improve productivity and flexibility, enabling lower latency and higher reliability through data processing and analysis at the edge. Despite the many advantages of distributed deployment, it still faces challenges such as complex task requirements, limited node resources, strict response time requirements, unbalanced cluster loads, and dynamic changes in the environment, which limit its application in the actual industrial environment. In order to find the optimal solution in the task response time and limited resource allocation, and ensure the balance and stability of the load, this paper proposes a system architecture for distributed edge deployment of industrial software component applications, and uses the improved MO-SAC algorithm to distribute the workflow, introduces the temperature coefficient, encourages the model to explore more and wider deployment strategies, and avoids falling into the local optimum. A large number of simulation results show that MO-SAC can effectively reduce the task response time while minimizing resource consumption and making the load of the cluster more balanced.
文章引用：徐沛沅, 王涛, 程良伦. 基于MO-SAC算法的工业软件组件边缘集群多目标决策部署[J]. 计算机科学与应用, 2025, 15(4): 162-176. https://doi.org/10.12677/csa.2025.154089

参考文献

[1] Urrea, C. and Benítez, D. (2021) Software-Defined Networking Solutions, Architecture and Controllers for the Industrial Internet of Things: A Review. Sensors, 21, Article 6585. [Google Scholar] [CrossRef] [PubMed]
[2] Bourechak, A., Zedadra, O., Kouahla, M.N., Guerrieri, A., Seridi, H. and Fortino, G. (2023) At the Confluence of Artificial Intelligence and Edge Computing in IoT-Based Applications: A Review and New Perspectives. Sensors, 23, Article 1639. [Google Scholar] [CrossRef] [PubMed]
[3] Hassan, N., Gillani, S., Ahmed, E., Yaqoob, I. and Imran, M. (2018) The Role of Edge Computing in Internet of Things. IEEE Communications Magazine, 56, 110-115. [Google Scholar] [CrossRef] 
[4] Chen, B., Wan, J., Celesti, A., Li, D., Abbas, H. and Zhang, Q. (2018) Edge Computing in IoT-Based Manufacturing. IEEE Communications Magazine, 56, 103-109. [Google Scholar] [CrossRef] 
[5] Qi, Q., Zhang, L., Wang, J., Sun, H., Zhuang, Z., Liao, J., et al. (2020) Scalable Parallel Task Scheduling for Autonomous Driving Using Multi-Task Deep Reinforcement Learning. IEEE Transactions on Vehicular Technology, 69, 13861-13874. [Google Scholar] [CrossRef] 
[6] Walia, G.K., Kumar, M. and Gill, S.S. (2024) AI-Empowered Fog/edge Resource Management for IoT Applications: A Comprehensive Review, Research Challenges, and Future Perspectives. IEEE Communications Surveys & Tutorials, 26, 619-669. [Google Scholar] [CrossRef] 
[7] Tan, H., Han, Z., Li, X. and Lau, F.C.M. (2017). Online Job Dispatching and Scheduling in Edge-Clouds. IEEE INFOCOM 2017-IEEE Conference on Computer Communications, Atlanta, 1-4 May 2017, 1-9.[CrossRef] 
[8] Zhang, Y., Du, P., Wang, J., Ba, T., Ding, R. and Xin, N. (2019) Resource Scheduling for Delay Minimization in Multi-Server Cellular Edge Computing Systems. IEEE Access, 7, 86265-86273. [Google Scholar] [CrossRef] 
[9] Chiang, Y., Zhang, T. and Ji, Y. (2019) Joint Cotask-Aware Offloading and Scheduling in Mobile Edge Computing Systems. IEEE Access, 7, 105008-105018. [Google Scholar] [CrossRef] 
[10] Orhean, A.I., Pop, F. and Raicu, I. (2018) New Scheduling Approach Using Reinforcement Learning for Heterogeneous Distributed Systems. Journal of Parallel and Distributed Computing, 117, 292-302. [Google Scholar] [CrossRef] 
[11] Liu, H., Ma, Y., Chen, P., Xia, Y., Ma, Y., Zheng, W., et al. (2020) Scheduling Multi-Workflows over Edge Computing Resources with Time-Varying Performance, a Novel Probability-Mass Function and DQN-Based Approach. In: Lecture Notes in Computer Science, Springer, 197-209. [Google Scholar] [CrossRef] 
[12] Xiong, X., Zheng, K., Lei, L. and Hou, L. (2020) Resource Allocation Based on Deep Reinforcement Learning in IoT Edge Computing. IEEE Journal on Selected Areas in Communications, 38, 1133-1146. [Google Scholar] [CrossRef] 
[13] Haarnoja, T., Zhou, A., Hartikainen, K., et al. (2018) Soft Actor-Critic Algorithms and Applications.
[14] Fan, Z., Su, R., Zhang, W. and Yu, Y. (2019) Hybrid Actor-Critic Reinforcement Learning in Parameterized Action Space.
[15] Haarnoja, T., Zhou, A., Abbeel, P., et al. (2018) Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor. 2018 International Conference on Machine Learning, Stockholm, 10-15 July 2018, 1861-1870.
[16] Figurnov, M., Mohamed, S. and Mnih, A. (2018) Implicit Reparameterization Gradients. Advances in Neural Information Processing Systems, 31.