基于Q-Learning算法的产业互联协同调度研究

Research on Industrial Interconnection Cooperative Scheduling Based on Q-learning Algorithm

针对产业互联中的跨界、业务资源流通滞后现象,研究产业互联下多区域、跨边界协同调度问题。根据产业互联模式特征,引入强化学习Q-Learning算法,在此基础上构建了供需双侧协同规划调度实施方案,阐释了Q-learning算法在推动产业互联模式中的调度技术应用细节,并以泛在电力网作为案例对象,探讨所提技术方案的实操性。结果表明:强化学习算法具有良好的适用性,能有效地协同泛在电力网资源供需调度,为产业互联在产业制造应用方面提供新视角。

In view of the phenomenon of cross-boundary and lagging flow of business resources in industrial interconnection,the problem of multi-regional and cross-boundary cooperative scheduling under industrial interconnection was studied.According to the characteristics of industrial interconnection model,the reinforcement learning Q-Learning algorithm was introduced.On this basis,the implementation scheme of supply and demand bilateral collaborative planning and scheduling was constructed to explain the application details of Q-learning algorithm in promoting industrial interconnection mode.Besides,a case study based on the ubiquitous power network was conducted to discuss the practicability of the proposed technical scheme.The results show that the reinforcement learning algorithm has good applicability and effectively coordinates the supply and demand scheduling of ubiquitous power network resources,and provides a new perspective for the application of industrial interconnection in industrial manufacturing.