微型电脑应用

为了降低单相接地故障对电网产生的负面影响，利用LightGBM(light gradient boosting machine)算法优化设计故障熄弧时刻识别方法。模拟单相接地故障产生与熄弧过程，从电流和电压2个方面，确定熄弧前后电网运行参数特征，以此作为熄弧时刻的识别标准。捕捉电网各区段线路运行数据，利用LightGBM算法确定故障位置，通过特征比对得出单相接地故障熄弧时刻识别结果。通过与传统识别方法的对比，发现优化设计方法的熄弧时刻识别误差降低了2.75 min,即优化设计方法在识别性能方面具有明显优势。

In order to reduce the negative impact of single-phase grounding fault on the power grid, LightGBM(light gradient boosting machine) algorithm is used to optimize the design of fault arc extinction time identification method. The process of single-phase grounding fault generation and arc extinction are simulated, and the characteristics of power grid operation parameters before and after arc extinction are determined from two aspects of current and voltage, which are used as the identification standard of arc extinction time. The line operation data of each section of the power grid are captured, LightGBM algorithm is used to determine the fault location, and the identification result of single-phase grounding fault arc extinction time are obtained through characteristic comparison. Compared with the traditional identification method, it is found that the identification error of the optimized design method is reduced by 2.75 min, that is, the optimized design method has obvious advantages in identification performance.