Prospecting and Fraud Combat Actions Identification to Commercial Losses

The traditional commercial losses prospect processes have not been effective. This paper presents a network inspection methodology using measurements at some strategic network points. Customers´ loads and distribution transformers are provided for a priori, based on existing load typology of a distribution utility and based on a social research considering financial reserve and consumption habits. These predictions are represented as trapezoidal fuzzy numbers. Commercial losses are detected by the difference between power supply output measure reading and the distribution transformers loads sum added with the ohmic losses. The commercial losses are estimated to be solving a linear fuzzy system, which calculates the transformers and distribution customers hourly demand correction initially referred to inaccurately by load typology. The transformers and customers with larger correction values are selected for inspection.