Intelligent diagnostics of electrical networks

Abstract

As follows from the conducted analysis [1-2], artificial intelligence has a great impact on modern energy as a whole, and it has also reached the electric power sector, especially "intelligent" networks with a high degree of automation, which closely interact with renewable energy sources. Here, artificial intelligence has gained widespread use in order to predict the level of illumination of a photovoltaic panel and estimate the output power of solar power plants. Considering the fact that artificial intelligence is a powerful and popular tool that is widely used in renewable energy (in particular, in photovoltaic), it is important to understand to what extent this tool can be used when creating a forecast of the generation of electrical energy at the output of a photovoltaic plant. It is becoming clear that with the help of artificial intelligence, it is necessary to increase both energy efficiency and accuracy values in the power grid, since electronic computing machines can process more data than an operator can in a given time period. When diagnosing the quality of electrical energy in a photovoltaic plant, it is important to observe certain provisions, namely: - adequate, for a specific task, the time of control. As a rule, this parameter must be installed in the system in advance; – determination of the number of electrical equipment and/or power system nodes for monitoring; – assignment of the limit level of parameters for measurement; – the choice of the method for performing the analysis of the measured data; – choosing the type and location for saving the received data, also here it is worth providing for compatibility with other devices in the electrical network, for example, control or signaling devices. In order to indicate the main tasks for the diagnostics system in the photovoltaic plant, which will include artificial intelligence, a structural diagram was created that indicates what tasks must be done in each link of the electrical network. It is worth noting that the structure of the diagnostic system can be divided into several components according to their physical location in the system under study. So, for example, sensors are responsible for all data collection in the power grid. Of greater interest are the links that perform monitoring in the power grid, as well as develop conclusions based on the conducted monitoring and accumulate databases for decision-making. An artificial neural network is responsible for fulfilling these requirements, and its data set for training and retraining can serve as a database.