Research Status and Prospect of Deep Learning in Secondary State Monitoring of Smart Substation

The network of the secondary system of the smart substation realizes the sharing and interaction of equipment information, and also brings a large amount of secondary system status data. Conventional secondary condition monitoring methods have deficiencies in processing big data. As a research hotspot of artificial intelligence, deep learning has strong data mining capability that meets the needs of state monitoring of smart substation secondary systems. In this context, the paper first outlines the basic ideas of deep learning and the typical structure of several commonly used models. Secondly, the concept, monitoring object and index selection of secondary system monitoring in smart substation are discussed. The advantages and disadvantages of using conventional methods and deep learning in communication network condition monitoring and secondary device status evaluation are then analyzed. Finally, combined with the current research and application status of deep monitoring of secondary system status in smart substation, the future development prospects are prospected.

I. Introduction

The wide application of deep learning in power systems is a hot research topic in the field of artificial intelligence [1]. The origin of its development can be summarized as two points: on the one hand, with the rapid construction of the smart grid and the development of the ubiquitous power Internet of Things, the massive data generated by the operation and maintenance of the power system meets the needs of deep learning; On the other hand, the rapid development of computer science has promoted the rapid development of deep learning theory, and the ability of modern computers to process big data is changing with each passing day, so it also lays a solid foundation for the application of deep learning in power systems [2].