Method for simulating and predicting low-voltage line loss based on machine learning

Abstract

The invention relates to a method for simulating and predicting low-voltage line loss based on machine learning. The method comprises the following steps: step S100, obtaining a low-voltage line losshistorical vector HL = (h1, h2,..., hn) of the to-be-predicted area, wherein hi = (ti, ai) is a low-voltage line loss element in the HL, the value of i is 1-n, ti is the predicted value of the ith low-voltage line loss performed in a first mode in the history of the to-be-predicted area, and ai is the measured value of the ith low-voltage line loss in the history of the to-be-predicted area; stepS200, if any hi in the HL exists, predicting the low-voltage line loss of the to-be-predicted area in a first mode; step S300, if one hi exists in the HL, executing the step S300; wherein D1 is a preset first threshold value; S300, if max (Hi)-max (min (Hi), D1) is smaller than or equal to lambda * D1, and lambda is a preset coefficient, predicting the low-voltage line loss of the to-be-predictedarea in a machine learning mode; otherwise, executing the step S400; and step S400, determining that abnormal low-voltage line loss elements exist in the HL.

Description

Technical field [0001] The invention relates to the information application field of power systems, and in particular to a method for simulation and prediction of low-voltage line loss based on machine learning. Background technique [0002] Line loss (referred to as line loss) refers to the loss of energy dissipated in the form of heat, that is, the active power consumed by resistance and conductance. In the power system, according to the different transmission voltages, the lines can be divided into high-voltage lines and low-voltage lines. The line loss of high-voltage lines far exceeds that of low-voltage lines. Therefore, the field of low-voltage line loss has become a research focus. [0003] Theoretically, the low-voltage line loss can be calculated according to the parameters of the power supply equipment and the operation mode of the power grid, power flow distribution and load conditions. For example, the article "Influencing Factors of Low-Voltage Line Loss and Research...

AI Technical Summary

Problems solved by technology

However, due to the complexity of the low-voltage line in the power supply area and the existence of electricity theft, the theoretical calculation method has a large error in some areas, and the calculation method of the line loss in time has a high volatility

Xu Hui et al. proposed in "Automatic Calculation Model of Distribution Network Line Loss Based on Machine Learning" to use active power supply, reactive power supply, total capacity of distribution transformers, and total line length as the input of the line loss SVR model Variables, through machine learning, simulated prediction of line loss in two cases of classification and non-classification, and achieved certain results, but the prediction accuracy in specific cases is still not accurate enough, and the running speed is relatively slow in large-scale prediction. Slow, low predictive efficiency