Transformer Inner Temperature Rise Calculation And Operation Scheme Optimization

Transformer top oil temperature(TOT)and winding hot spot temperature(HST)is the key factors to estimate the thermal state of the transformer.Calculation and prediction of TOT and HST is important to safe operation and estimating insulation life of transformer.In this paper,a lot of work has been done which focus on transformer inner temperature rise calculation,cooling effiency calculation and operation modes optimization considering insulation life loss.This paper has established thermal models for traditional oil-immersed transformer and one new split type cooling transformer.For the TOT calculation of traditional oil-immersed transformer,this paper proposed two models:point prediton model and interval prediction model.Firstly,combining the advantages of.semi-physical model and data-driven model,a combined model based on kernel extrame learning machine(KELM)error prediction and correction is established.The results of case study demonstrate that the accuracy of the proposed model is better than single semi-physical model and data-driven model.Secondly,a TOT interval prediction model based on KELM and Bootstrap method is established.The upper bound and lower bound the prediction interval can be used as the conservative and optimistic estimation of the TOT.The Conservative prediction result avails the safe operation of transformer.At last,thermal model for split type cooling transformer is established.Split type cooling transformer is ued in underground substation.The transformer body is underground while the radiator is overground.Traditional thermal models do not apply to split type cooling transformer due to structural difference.Based on thermal principle analysis and comparasion with traditional transformer,a thermal circuit model for ONAN split type cooling transformer is established.The temperature rise experiment data are used to verify the model.In this paper,a transformer cooling effiency calculation method based on reverse solution of thermal resistance is established.On-line measured TOT data and PSO algorithm are use to slove the actual thermal resistance,and the actual thermal resistance is compared with factory test thermal resistance.The comparation result and variation tendency of actual thermal resistance are used to estimate the cooling capability,diagnose the cooling system faults and provide information for transformer operation and overhauling.In this paper,reducing the transformer thermal life loss is taken as one of the object of economic operation.Combined with traditional economic operation object and based on double period controed strategy,PSO algorithm is used to solve the best switching time of standby transformer.This paper provides a reference method for transformer economic operation from the perspective of reducing the thermal life loss.