基于覆冰周期分段建模预测的研究

Study on the Prediction of Segmented Model Based on Icing Period

输电线路覆冰现象在冬季比较普遍,已成为威胁架空输电线路安全运行的重要因素。物理模型和经验模型是建立输电线路覆冰模型的2种思路。一个完整的覆冰周期变化复杂,目前建立的经验模型均是针对整个覆冰周期,建模精度不高。根据大量覆冰数据聚集分析,将覆冰周期分为4段:覆冰快速增长期、覆冰不稳定增长期、覆冰稳定期和覆冰脱落期。提出了针对不同时期变化特征建立不同模型进行覆冰预测,对比了整体建模预测结果与分段建模预测结果,结果表明分段建模预测精度优于整体建模精度。

Transmission line icing is a common phenomenon in China, which has become an important factor to thz^aten the safe operation of overhead transmission lines. Transmission line modeling has two thoughts： physical models and experiential models. So far, none of models is proved successfully. The whole icing period changes complexly, experiential models are based on the whole period and precision of models is low. The whole icing period is divided into four sections： the icing period of rapid growth, the icing period of unstable growth, the stable icing period and the icing period of drop. It is put forward to build different forecast models according to different characteristics of different models. Compared with overall modeling, it is concluded that segmented modeling is better than overall modeling.