Assessment of MV XLPE cable aging state based on PSO-XGBoost algorithm

作者：Pan, Qiaosheng; Zhang, Chi; Wei, Xinyuan; Wan, Aode; Wei, Zeping^*

来源：ELECTRIC POWER SYSTEMS RESEARCH, 2023, 221: 109427.

DOI：10.1016/j.epsr.2023.109427

• Cable aging status assessment
• Extreme gradient boosting algorithm
• Particle swarm optimization algorithm

Cable aging is one of the main security risks to power systems. With the widely used cables in power systems, the accurate assessment of cable aging status is increasingly important. This study proposes an efficient assessment model based on the PSO-XGBoost algorithm, which integrates the particle swarm optimization (PSO) algorithm and the extreme gradient boosting (XGBoost) algorithm. The XGBoost model is established to assess the cable aging status with the inputs of partial discharge, operating life, corrosion condition and load condition. The PSO algorithm automatically optimizes parameters during XGBoost model training. Then, the standard performance evaluation metrics of the proposed assessment model are compared with four advanced classification models. The accuracy, precision, recall and F1-score of the assessment model are above 98%, indicating that the proposed PSO-XGBoost model can accurately assess the cable aging state. Furthermore, these calculation results of the proposed model are better than the other four benchmark models, which shows that the proposed model performs better in cable aging status assessment than the existing models.