On-line Monitoring System of Customer Electricity Safety Based on Power Grid Enterprise

Smart grid, Client monitoring, Data model, Complex scenarSmart grids can accurately monitor the power consumption behavior of each grid node and each user. It can effectively ensure the two-way flow of information and power of all nodes in the whole transmission and distribution process and is an automatic power transmission network. The event-based method collects different types of actual load data and preprocesses them. Then, it detects the state of the event, and extracts the characteristic information for load identification. In the process of event detection, the data of this method is vulnerable to noise pollution. It can not monitor the power consumption information of electrical equipment in complex scenarios. In this paper, the variational mode decomposition is applied to obtain the data power allocation of NILM. The effective information of NILM is retained. Combined with the convolutional neural network model to realize load identification, the rationality and effectiveness of the proposed scheme in a complex operating environment are verified by example analysis. The simulation results indicate that the hybrid model can overcome the low efficiency of the single model. Finally, the superiority of the proposed scheme is proved by comparing the method based on the conditional generation countermeasure network and the method based on the holey residual network. Besides improving the load monitor accuracy of the NILM in a complex scene, the invention expands the universality and the expandability of the NILM by establishing a combined experiment, which improves the load monitoring efficiency and increases the practical use value.