Research on Anti-stealing Algorithm of Distributed Photovoltaic Power System Based on Power Big Data

DPV (Distributed Photovoltaic Power System) is an important development direction in the field of renewable energy, which helps to reduce greenhouse gas emissions, improve the sustainability of energy supply and provide consumers with more energy choices. Power BD (Big Data) technology, as a BD technology oriented by the power industry under the background of the new era, is widely used in anti-theft investigation and work, which has good accuracy, convenience and scientificity, and is of great significance to anti-theft investigation. In this paper, a DPV anti-stealing algorithm is established by BD mining. Based on RF (Random forest), these features are used to identify the behavior of stealing electricity and find the features related to stealing electricity. An anti-stealing model is built by LR (Logical regression) algorithm, and the suspected stealing degree of users is calculated. The research results show that RF algorithm performs better on data sets, with the highest accuracy and recall of 0.84 and 0.89 respectively, and F1 also reaches 0.81. The results show that the anti-stealing prediction method based on electric power BD proposed in this paper is efficient and feasible in identifying suspected stealing users.