Contamination Grade Prediction Of The Catenary Insulator Based On The Characteristic Parameters Of The Leakage Current

With the striding development of electrical railway in our country,the new period of “eight vertical and eight horizontal” high speed railway which has been included in planning,the increase in the mileage of high speed railway makes the operating environment of the catenary insulator more complex.The research shows that the surface contamination grade and the wetting of the insulator contamination layer are closely related to its operating environment,and the contamination flashover accident caused by polluting greatly affected the safe operation of the railway.The main reason for contamination flashover of the insulator is the generation and increase of the leakage current,and the leakage current is related to the contamination grade of the insulator.Therefore,it is an important way to help develop the cleaning and maintenance plan of the catenary insulator and ensure the stability of the electrical railway power supply by researching the state information contained in the leakage current signal and then establishing the contamination grade prediction model by using these information as input and using contamination grade as output and using the online monitoring data to predict the contamination grade on the surface of the insulator.It is pointed out in this thesis that the leakage current as a dynamic parameter is a perfect parameter to characterize the contamination grade of the insulator through the study on the reason of surface discharge and the flashover mechanism of the insulator and compare the parameters which can characterize the operating state of the insulator.So the artificial contamination test for the catenary composite insulator(FQX-25/120)is carried out,as well as the leakage current signal and the information of ambient temperature and humidity are collected and stored by the leakage current on-line monitoring device which researched and developed by research group.By comparing the leakage current data with the experimental phenomena and the ambient humidity,it is found that the change of the amplitude of leakage current is not obvious with the increase of the contamination grade under low humidity condition.Under saturated humidity condition,the amplitude of leakage current and the content of pulse and harmonic in the leakage current waveform increase with the increase of the contamination grade.Then,the study of the leakage current data is based on the partition theory which is mature and widely recognized in the field of electric power network.It is pointed out that the leakage current waveform of the catenary composite insulators also has three zones: the safe zone,the forecast zone and the danger zone,and the partition threshold values of these three zones are 10 mA and 30 mA.The study on the characteristics of leakage current in different zones shows that the root mean square of the danger zone,the pulse time domain entropy of the leakage current and the total harmonic distortion(THD)of the forecast zone can be used as the characteristic quantity to characterize the contamination grade of the insulator.Finally,the prediction model of the contamination grade of catenary insulator based on radial basis function(RBF)neural network is established,the particle swarm optimization(PSO)is used to optimize the parameters of the neural network to improve the accuracy of model prediction.Training and validating the model using 400 sets of leakage current characteristic data(the RMS of the danger zone,the pulse time domain entropy of the leakage current and the THD of the forecast zone),and the predicted results are compared with the results of the RBF neural network without optimization.The results show that the prediction model has a good effect on predicting the contamination grade of the catenary insulator,and it can provide some guidance for the cleaning and maintenance of the catenary insulator.