基于原始回归算法的高压电力计量数据智能校正方法

Intelligent Correction Method for High Voltage Power Measurement Data Based on Primitive Regression Algorithm

高压电力计量数据在采集、传输以及处理过程中可能会产生误差,若这些误差得不到校正,则会影响到电力系统的稳定性和可靠性,因此研究基于原始回归算法的高压电力计量数据智能校正方法。将估算功率因数与负载率作为异常指标,通过异常指标突出高压电力计量数据等异常信息;采用原始回归算法确定高压电力计量数据的偏差,以回归系数梯度下降法减小数据偏差,实现高压电力计量数据的智能校正。实验结果表明,分别在异常数据干扰以及噪声干扰的环境下获取测试样本,新方法能够在出现干扰因素后快速实现高压电力计量数据的智能校正,且校正结果具有较高的准确性,可以为实际工程应用提供有力支持。

There may be errors in the collection,transmission,and processing of high-voltage power metering data.If these errors are not corrected,they will affect the stability and reliability of the power system.Therefore,an intelligent correction method for high-voltage power metering data based on the original regression algorithm is studied.Using estimated power factor and load rate as abnormal indicators,highlighting abnormal information in high-voltage power metering data through abnormal indicators.The original regression algorithm is used to determine the deviation of high-voltage power measurement data,and the regression coefficient gradient is used to reduce the data deviation,achieving intelligent correction of high-voltage power measurement data.The experimental results show that the new method can quickly achieve intelligent calibration of high-voltage power measurement data in environments with abnormal data interference and noise interference,and the calibration results have high accuracy,providing strong support for practical engineering applications.