Line Loss Rate Prediction Method Based on Deep Learning with Long Short Term Memory

The line loss rate plays an important guiding role in the power system management and operation. Furthermore, the line loss rate prediction is extremely necessary for power supply manager to set energy-saving goal. The development of machine learning and artificial intelligence technology provides an effective way for improving the line loss rate prediction accuracy. In this paper, a deep learning framework based on Long-Short Term Memory for line loss rate prediction is proposed to improve the accuracy where the following influencing factors, including power supply, power factor, total harmonic distortion, unbalanced degree and load shape coefficient, are considered. To be specific, in our work, the above five factors are input and the predicated line loss rate is output. In other words, the model based on Long-Short Term Memory is employed to calculate the nonlinear relationship between the input and output, where the simplified relationship are built firstly, and then, more implicit relationship which is more complicated and difficult arc further calculated. Finally, the test collected from a power supply company of china is applied to confirm the effectiveness of the proposed methodology. The tests result illustrates that the accuracy of line loss rate can be greatly guaranteed.