Research On Intelligent Decision For The Prevention And Control Of Icing Power Grid

Snow rainfall and freezing disasters have made power system suffered a great loss. Ice coating on power transmission line in winter has brought a series of problems, it has also gotten more and more attentions. Decision making problems about how to make use of new ways of anti-icing and de-icing is very important. This paper focuses on prevention and prediction of the power grid icing, configuration of DC de-icing devices and the de-icing sequence for the icing grid. A standardized workflow is completed, which is from the pre-disaster prediction to the de-icing in disaster. Firstly, this paper uses the Back Propagation(BP) neural network to predict the power grid icing based on the historical 60-year icing data in Hunan province, and the icing degree is also determined by the number of days. Secondly, about the icing degree which is in severe icing and above, this paper makes deep study for the dynamic configuration of the DC de-icing devices and proposes optimal configuration of DC ice-melting devices. Based on the optimal configuration, this paper uses Hunan icing grid area as an example and gets the optimal types and number of ice melting device. Finally, taking icing severity factor, path factor and specific line factor which influence the ice-melting sequence of icing grid into consideration, this paper uses the scoring model to build a decision model of icing grid ice-melting sequence. This model solves the sequential decision problem of transmission line when the icing grid is under the de-icing. The ice-melting sequence in Hunan power grid is used as an example to demonstrate the practicality and feasibility of the model.