Measurement Of Capacitance And Dielectric Loss And Evaluation Of Health Condition Of Power Capacitors

Power capacitors are widely used as reactive power compensation equipment in the power system,which have the advantages of low cost and easy operation and play an important role in improving power factor,ensuring voltage quality,reducing power loss,improving system power transmission capacity and enhancing system stability.The relatively complex working environment of capacitors may cause various early failures.If the failures develop to a certain extent,they will threaten the safety of the capacitors and the normal operation of the power system.Therefore,in order to discover the early faults of the capacitor in time and ensure the normal operation of the power system,it is necessary to study the online monitoring and health assessment of the capacitorAmong the key parameters of the capacitor,the capacitance and the dielectric loss angle can reflect various early failures of the capacitor.By measuring these two key parameters online,the health state of the capacitor can be reflected in real time.Existing methods mainly measure the capacitance and dielectric loss angle through the steady-state waveform.The measurement accuracy is affected by asynchronous sampling,and due to the limited information in the steady-state waveform,it is difficult to take the error of the transformer into account during the measurement,which is not conducive to the actual site application.Compared with the steady-state waveform used in the existing method,the frequency information in the transient waveform is more abundant.In this thesis,the transient waveform is analyzed by the parameter identification method,and the capacitance and the dielectric loss angle are measured under the premise of considering the transformer error.Regarding the evaluation of the health status of capacitors,related researches face difficulties in quantifying the health status of capacitors.In this thesis,the condition number is used to provide a quantitative basis for the evaluation of the health status of capacitors.In order to make full use of the rich frequency information in the transient waveform and reduce the influence of the transformer error and asynchronous sampling on the measurement accuracy,the thesis studies the measurement algorithm based on the power electronic disturbance signal.By installing the thyristor on the secondary side of the PT to produce the disturbance signal,it first solve the system harmonic impedance through the regularization algorithm,and then substitute the system harmonic impedance into the equivalent circuit considering the transformer error,establish the capacitance and dielectric loss angle measurement equations,solved by combining genetic algorithm with LM method.It is shown that this algorithm has high accuracy and low initial iteration value requirements,which is basically not affected by frequency fluctuations and harmonics,but has high sampling frequency requirements.Compared with the two representative methods,the measurement accuracy is not affected by error limit of transformer.In order to solve the problem of algorithm based on the power electronic disturbance signal that the thyristor needs to be installed and the measurement process is complicated,the thesis studies the capacitance and dielectric loss angle measurement algorithm based on the capacitor switching transient.The equation is established according to the transient state of capacitor switching,the capacitance and dielectric loss angle are solved by the interior point method,the influence of noise is reduced by adding the wavelet denoising,and a measurement frequency selection method based on sensitivity analysis is proposed to further improve measurement accuracy.It is shown that the measurement accuracy of this algorithm is basically not affected by frequency fluctuations and harmonics.Compared with the algorithm based on power electronic disturbance signals,it is greatly affected by the initial value of the iteration,but it reduces the requirement for sampling frequency and increases the anti-noise ability.Compared with the two existing methods based on steady state,this algorithm is not only less affected by the error of the transformer,but also can improve the accuracy by selecting the measurement frequency.In order to establish a quantitative basis for evaluating the health status of capacitors,the thesis studies an evaluation method based on conditional number.The matrix is used to describe various breakdown conditions of capacitors in series,parallel and complex wiring.The relationship between the condition number and the health status of the capacitor is analyzed,and the evaluation basis based on the condition number is established to sort the threat levels and then score the health of the capacitor.It is shown that by comparing the theoretical value and the measured value of the capacitor value,the current breakdown status of the capacitor can be effectively determined,and the evaluation result of the health status of the capacitor based on the condition number is consistent with the field operation experience.The thesis uses the rich frequency information of the transient waveform to measure the capacitance and dielectric loss angle online.The measurement accuracy is basically not affected by asynchronous sampling.Compared with other representative methods,the influence of the error of the transformer is significantly reduced.After obtaining the measurement result of the capacitance,the breakdown state of the capacitor can be determined.The thesis uses the evaluation based on the condition number to evaluate the health status of the capacitor under different breakdown conditions.The research work in this thesis can provide a reference for the online monitoring of capacitors and the assessment of health status,and it has a positive significance in ensuring the normal operation of capacitors and improving the reliability of power system.