Research On Fault Diagnosis And Condition Based Maintenance Of Power Equipments

Power transformers are usually installed outdoors, In most cases, large-scale power transformers are operated ceaselessly all day without monitoring by human in bad environment . Because of operating outdoors, they are monitored less than Secondary equipments which are installed indoors. At present, the quantity and the capacity of transformers are growing gradually with the rapid development of electric power industry. Power transformers play an important role in power system . Therefor, it is helpful and necessary to strengthen the research on power transformers intelligent fault diagnosis and condition based maintenance(CBM).This thesis deeply analyzese Thesis 11 and Thesis 12. In this thesis, information processing technology and artificial intelligence are used to analyze fault diagnosis of power transformers and decision support systems of the power transformers maintenance based on condition. The major contents are shown as follows: 1,The abnormality and its type that is available in transformer operation are investigated. It is noted that there exist several methods frequently used for fault identification and fault property judgment. The transformers fault diagnosis system which is brought forward by Thesis 8 is deepened in this thesis;2,One side due to multiplicity of transformers fault sign and positions it is difficult to comprehensively confirm transformer fault component only using neural network, the other side abundant experience is accumulated during the long-term monitoring and maintenance to transformer. It is necessary to transform these experience into fault diagnosis rules. Therefor this thesis try to apply decision tree ID3 arithmetic to confirm transformer fault position. Based on the entropy theory , in virtue of the locale experience of operation and diagnosis, applying conclusion of the routine electric examination and using exterior information, this arithmetic forms the fault diagnosis rule of transformer ;3,After transformer fault diagnosis ,usually experts need to bring forward some opinion and advice about the maintain methods. Transformer maintenance is a complex job that is involved in economic profit, societal effect,etc.Due to involved in too much factors, it is very difficult for decision-maker to make a correct choice. Therefor this thesis applies analytic hierarchy process(AHP) to set up a decision support system. This system can choose the best scheme from many primary schemes, give decision-maker helpful theory support.4,A software system for fault diagnosis of electrical equipment is developed with JAVA technology. This system can manage the nameplate data, forecast the dissolved gases in oil content, diagnose fault of the devices according on-line monitoring and preventive electric experiment.It is constructed on enterprise Intranet and is transplantable, extensible and easy for maintenance.