Model for Forecasting of Electricity Losses During Transmission and Distribution in an Electricity System

摘要

The forecasting of losses in a country’s electrical system is an important factor for its optimal management. It allows for the planning of the appropriate amount of electricity that must be produced in order to meet the needs of the economic sector and household customers. A portion of the electricity losses can be forecasted and assessed. This includes losses caused by technical components for transmission and distribution of electricity, losses from power lines formed by the length, crosssection, or resistance of conductors, the corona effect, losses due to no load or light load of the transformers, energy consumption for the needs of the individual units in the electricity system, etc. Another portion of the losses is not subject to accurate calculation. These are losses from accidents on significant lines, which may be due to wear and tear of the equipment being used, natural disasters, changes in the prevention programs, or even illegal or dishonest consumption of electricity, etc. These losses cannot be forecasted by standard statistical methods or wellknown formulas. This article presents a study for forecasting electricity losses by using neural networks. Experiments have been conducted related to the forecast of the losses of the electricity system in Bulgaria.