深度学习在电网智能调控系统中应用研究

基于对人工智能深度学习的研究，将该技术与电力智能调控深入融合，实现以数据、知识为基础的实时在线分析控制型电力调控。首先，分析深度学习的发展及其对电网调度的影响，针对深度学习在出力预测、状态估计和故障诊断方面的应用进行说明；然后，提出基于人工智能深度学习的电力调控系统构架，分析其与传统调控系统的区别，突出了该系统所实现的功能；接着,采用基于多核图像处理器（MGP）系统,阐明了所提出的智能调控系统实现的方法；最后，通过分析人工智能化调控系统的应用案例，说明该系统具有广泛的应用性。

Research on the Application of Deep Learning Theory in Power Grid Intelligent Dispatching

Based on deep learning theory,a new power system dispatch mechanism which is highly integrated with the power intelligence is proposed,featured by data based and knowledge based dispatch.Firstly,the development of deep learning and its impact on power grid dispatching are analyzed,and the application of deep learning on output forecast,state estimation and fault diagnosis are explained.Secondly,the architecture of power replacement system based on artificial intelligence deep learning is proposed,and the difference between it and traditional minimization system is analyzed,while highlighting the system functions.Thirdly,the multicore image processor(MGP)based system is used as the method of intelligent alternative system implementation proposed in this paper.Finally,this paper analyzes the application case of artificial intelligence solution system.