Oil-immersed Transformer Hot State Research And Cooling Efficiency Evaluation

Power transformer work as the expensive and pivotal equipment of the power system,its operating state has a significant effect on reliability and operational cost of the power system.Transformer internal temperatures determine the load capacity and performance of the transformer insulation system,is the key indicator examining its stable operation,and transformer cooling equipment works as the key structure to achieve cooling the transformer,whose cooling performance directly affects the internal thermal condition of the transformer.Therefore,research on transformer internal temperature rise and cooling capacity and cooling efficiency has great significance for regulating operation,reducing loss,monitoring and preventing the thermal troubles and extending the equipment life of the transformer substation system.The paper firstly carries through algorithm prediction and thermal circuit calculation for the internal temperature of the integrated and split-type transformer,then deduces and calculates the utilization efficiency of cooling equipment,at last the evaluation method of the cooling equipment capacity is established.Aiming at the widely used oil-immersed transformer,the paper's goal is predicting the internal temperature accurately,hybrid kernel extreme learning machine is made use of to achieve the temperature prediction.The model takes delay effect of influence factors on the target temperature and ambient wind speed into consideration roundly,different from the traditional prediction methods,takes advantage of particle swarm optimization to train the model and optimize kernel parameters and output weights of the prediction model.The positive and negative errors of the predicted values are processed with different allowable limits to make the predicting top oil temperature results are much more conservative and accurate during the training process of the machine learning.Domestic and overseas research lacks thermal circuit calculation of the split-type transformer,basing on the principle of thermoelectric equivalence,lumped parameter and energy conservation,the paper establishes the thermal circuit calculation model of the split-type transformer internal temperature and improve the calculation model according to the oil viscosity,real time loading and special loss function.The relevant parameters of the model are obtained by fitting according to search algorithm and measured data,another set of data is used to verify the model.The accuracy and feasibility of the model is verified in comparison with traditional methods,providing an idea that the standard guide of the split-type transformer thermal circuit calculation is drawn up.Effective and reliable transformer cooling equipment is crucial for safely operation of the transformer.The paper studies the transformer cooling problems from cooling capacity and cooling efficiency of cooling equipment,these two perspectives.The ratio of actual cooling power and cooler's rated cooling capacity is firstly defined as the cooling efficiency,the actual cooling efficiency is calculated according to classical equivalence mathematic model of the three-winding transformer heat power,guide rule of the cooler,energy conservation and temperature and structure data.Case study of the oil and water heat exchanger is carried out,the calculation results make great sense for regulating operation of the cooling equipment.At last,the approximate exponential function of the transformer internal temperature rise is set up in terms of heat transferring theory and energy conservation,method of undetermined coefficients is utilized to seek the time constant of the temperature rise function and steady temperature value to by fitting temperature rise data.Both of the indicator and temperature rise curve are used to evaluate cooling capacity variation of the cooling equipment.The instance analysis of split-type transformer and integrated transformer demonstrate the feasibility of the method.