This paper introduces a computerized on-line partial discharge （PD） monitoring and diagnostic system for transformers. The system, which is already in use in a power station, uses wide-band active transducers and a ...This paper introduces a computerized on-line partial discharge （PD） monitoring and diagnostic system for transformers. The system, which is already in use in a power station, uses wide-band active transducers and a data acquisition unit with modularized and exchangeable components. The system software is a power equipment monitoring and diagnostic system, which is based on the component object model, and was developed for monitoring multiple parameters in multiple power supply systems. The statistical characteristics of PDs in power transformers were studied using 7 experimental models for simulating PDs in transformers and 3 models for simulating interfering discharges in air. The discharge features were analyzed using a 3-D pattern chart with a three-layer back-propagation artificial neural network used to recognize the patterns. The results show that PDs in air and oil can be distinguished. The model can be used for interference rejection on-line monitoring of partial discharge in transformers.展开更多
基金Supported by the National Natural Science Foundation of China, the Northeastern Electric Power Corp. (Group) (No. 59637200), and the National Foundation of USA
文摘This paper introduces a computerized on-line partial discharge （PD） monitoring and diagnostic system for transformers. The system, which is already in use in a power station, uses wide-band active transducers and a data acquisition unit with modularized and exchangeable components. The system software is a power equipment monitoring and diagnostic system, which is based on the component object model, and was developed for monitoring multiple parameters in multiple power supply systems. The statistical characteristics of PDs in power transformers were studied using 7 experimental models for simulating PDs in transformers and 3 models for simulating interfering discharges in air. The discharge features were analyzed using a 3-D pattern chart with a three-layer back-propagation artificial neural network used to recognize the patterns. The results show that PDs in air and oil can be distinguished. The model can be used for interference rejection on-line monitoring of partial discharge in transformers.
