Abstract:The calculation of reasonable value of synchronous line loss ratio is an important part of the delicacy management of line loss in power distribution station. It provides effective technical support for locating the stations with abnormal line loss, and then helping to reduce the energy loss accurately. This paper decomposes the complex influencing factors, reduces the dimensionality of multi-dimensional influencing factors, and establishes a synchronous line loss ratio calculation model based on characteristic parameters such as metering deviation, transmission loss under balanced full load conditions, load characteristic and basic loss. Based on big data analysis, a "dual learning" solution algorithm for characteristic parameters is proposed. Solve the characteristic parameters of each station, and through the correlation analysis of the same type of stations, the reasonable interval of characteristic parameters of each type of stations is solved, so as to discriminate the stations with abnormal line loss and the learn the reasonable values of parameters of these stations. Finally, the intelligent computing of benchmark value and reasonable interval of synchronous line loss rate of each station under different power supply are realized. The rationality and effectiveness of the algorithm model are proved by actual applications.