基于灰色支持向量机的输电线路覆冰厚度预测模型

为了降低输电线路覆冰事故对电网安全造成的严重影响,对输电线路覆冰厚度进行预测将能够有效地指导电网抗冰工作。提出了基于灰色支持向量机的输电线路覆冰厚度短期预测模型,分析了样本中脏数据的剔除及数据预处理方法,通过模型预测值与实测数据的对比验证了该模型的准确性和适用性,根据模型预测的线路最大覆冰厚度值对现场观冰、冰情预警以及开展交直流融冰提供策略指导。将该模型与传统的支持向量机和广义回归神经网络覆冰预测模型进行了对比,结果表明,该模型平均误差为0.325 mm,平均绝对百分误差仅为2.61%,适用于输电线路覆冰厚度短期预测。在易覆冰地区,应用该预测模型能够更好地指导输电线路抗冰工作。

Prediction Model for Icing Thickness of Power Transmission Line Based on Grey Support Vector Machine

In order to reduce the impact of icing accidents on transmission lines, prediction of icing thickness on transmission lines is very effective in anti-icing work of power grid. A short-term prediction model based on grey support vector machine for icing thickness of transmission lines is proposed. The methods of dirty data elimination and data preprocessing are analyzed. The accuracy and applicability of proposed model are verified by comparison between model predictions and measured data. Predicted maximum ice thickness can provide guidance on monitoring icing condition, early warning and AC/DC ice melting work. Compared with support vector machine (SVM) and generalized regression neural network prediction model, the proposed method has average error of 0.325 mm, and average absolute error of 2.61%, which is suitable for short-term prediction of icing thickness of transmission line. The application of the prediction model can guide the transmission line ice-resistant work in the ice area.