忆阻神经网络在智能电网监控系统中的应用

Application of memristive neural networks in smart grid monitoring system

随着能源危机的到来,智能电网技术成为世界各国所关注的重点。而与此同时,智能发电、输电、变配电、用电以及智能调度等各个环节所取得的技术性突破也为智能电网的大规模实现提供了可能。然而,智能电网结构复杂,电气设备分布广泛,应用需求多,这些特点都给电网运行的安全性和可靠性提出了挑战。针对智能电网的安全运行问题,提出了将忆阻器与人工神经网络相结合,构建出基于忆阻神经网络的智能电网运行状态监控系统,从而可以利用忆阻器所具有的记忆功能,节省人工神经网络的权值传输时间,提高神经网络的数据训练效率,保证了监控系统的及时性和有效性。

With the coming of the energy crisis, smart grid technology has become the focus concern of the world. While at the same time, technological breakthroughs in all aspects of intelligent power generation, transmission, substation, power distribution, power consumption and intelligent scheduling make it possible for large-scale realization of the smart grid. However, the structure is complex and the electrical equipments are widely distributed in smart grid. Meanwhile, there are a lot of application requirements. These features give most challenge of safety and reliability for the power＇s running. For the problem that safe operation of the smart grid, the combination of memristor and artificial neural networks was proposed to build a smart grid operation status monitoring system based on memristive neural networks. The system could use the memristor＇s memory function to save the value of artificial neural network transmission time and improve the efficiency of neural network training data. So it could ensure the timeliness and effectiveness of the monitoring system.