| Transmission linesicing isaffected by many factors,and lines load is increased while the icing occurring.Many accidents,like break,tower collapseand flashover were resulted from serious icing.As a result,a huge economic loss to society would be caused.So,it's very important to establish a transmission lines icing predictivemodel,which is based on On-line monitoring data and will provide early icing warning,protecting the power system's operation safely and reliably,and it has a giant engineering practical significance.Research contents are as follows:Firstly,the grey relation analysis' s DEN-relation model and improved slope relation model were studied comparatively.As a result,the important influence factor is environment temperature,the sequence of gray correlation is:environment temperature> conductor temperature >environment humidity >environment windspeed.Direct evidence for the theoretical analysis of transmission lines icing forecast were provided byGrey relational analysis results.Secondly,genetic algorithms and fuzzy logic theoretical basis,key technologies,advantages and disadvantages,and the integration of the two theories were focoused on.On this basis,a kind oflines icingforecast model method based on genetic algorithm and fuzzy logic was put forward in this paper.The prediction results show that when the ice thickness is between 0 ~ 5 mm,the average relative error is 0.0165%,when the ice thickness is between 5 ~ 10 mm,the average relative erroris-0.165%,when the ice thickness is between 10 ~ 18 mm,the averagerelative error is 3.34%.Finally,a medium climatic chamber was designed and built.It can be simulated the transmission lines icing tests under different weather conditions,different line conditions,and analyzed the icing mechanism.Anicing forecast model field validation was also carried out,the field validation results show that: the model error is within ± 2mm;the relative error does' t exceed 20%.There is a large relative error in ice thickness of about 1mm. |
PDF Full Text Request