This paper studied an integrative fault diagnostic system on the power transformer. On-line monitor items were grounded current of iron core, internal partial discharge and oil dissolved gas. Diagnostic techniques wer...This paper studied an integrative fault diagnostic system on the power transformer. On-line monitor items were grounded current of iron core, internal partial discharge and oil dissolved gas. Diagnostic techniques were simple rule-based judgment, fuzzy logistic reasoning and neural network distinguishing. Considering that much faults information was interactional, intellectualized diagnosis was implemented based on integrating the neural network with the expert system. Hologamous integrating strategies were materialized by information-based integrating monitor devices, shared information database on several levels and fusion diagnosis software along thought patterns. The expert system practiced logic thought by logistic reasoning. The neural network realized image thought by model matching. Creative conclusion was educed by their integrating. The diagnosis example showed that the integrative diagnostic system was reasonable and practical.展开更多
The conventional fault analysis method based on symmetrical components supposes that the three-phase parameters of un-transposed transmission line are symmetrical in case of fault. The errors caused by the method with...The conventional fault analysis method based on symmetrical components supposes that the three-phase parameters of un-transposed transmission line are symmetrical in case of fault. The errors caused by the method with the symmetrical distributed parameter circuit model as the equivalent circuit of the un-transposed ultra high voltage(UHV) transmission line were studied under both normal operation and fault,and the corresponding problems arising were pointed out. By contrast with electromagnetic transient and power electronics(EMTPE) simulation results with the asymmetrical distributed parameter circuit model of un-transposed line, it is shown that the conventional method cannot show the existence of negative and zero sequences before fault happening and there are many errors on voltage and current after fault happening which are different with fault types. The error ranges of voltage and current are 2.13%-81.13% and -7.82%- -86.15%, respectively.展开更多
文摘This paper studied an integrative fault diagnostic system on the power transformer. On-line monitor items were grounded current of iron core, internal partial discharge and oil dissolved gas. Diagnostic techniques were simple rule-based judgment, fuzzy logistic reasoning and neural network distinguishing. Considering that much faults information was interactional, intellectualized diagnosis was implemented based on integrating the neural network with the expert system. Hologamous integrating strategies were materialized by information-based integrating monitor devices, shared information database on several levels and fusion diagnosis software along thought patterns. The expert system practiced logic thought by logistic reasoning. The neural network realized image thought by model matching. Creative conclusion was educed by their integrating. The diagnosis example showed that the integrative diagnostic system was reasonable and practical.
文摘The conventional fault analysis method based on symmetrical components supposes that the three-phase parameters of un-transposed transmission line are symmetrical in case of fault. The errors caused by the method with the symmetrical distributed parameter circuit model as the equivalent circuit of the un-transposed ultra high voltage(UHV) transmission line were studied under both normal operation and fault,and the corresponding problems arising were pointed out. By contrast with electromagnetic transient and power electronics(EMTPE) simulation results with the asymmetrical distributed parameter circuit model of un-transposed line, it is shown that the conventional method cannot show the existence of negative and zero sequences before fault happening and there are many errors on voltage and current after fault happening which are different with fault types. The error ranges of voltage and current are 2.13%-81.13% and -7.82%- -86.15%, respectively.
