Study Of The Comprehensive Assessment Method For The Power Transformer Condition In Service With Fuzzy Theory

The reliability of power equipment is vital to power system's safe operating. It is imperative forthe CBM under the extension of power grid and the reform of electric power system. It is also atrend of developing power grid. As the ultimate element in the electricity supply chain, the safeoperating of transformer is close correlative with the safetyof the power system. So it is important tofind out the state of running transformer. At present, more and more attention has been paid to thecondition assessment of power transformer. Although some achievements have been obtained, thereare still needs for the further investigation to this field. Therefore, on base of consulting a greatamount of material, a deep research on the condition assessment method of the transformer usingfuzzy comprehensive evaluation is done in this paper.　Appropriate status parameter was chosen to assess the condition of the transformer bycombining the existent achievement and the expert experience, considering the practicalmaneuverability. So electrical test, state parameter of dissolved gases in oil, parameter of oilcharacteristic and other factors (include the working atmosphere, operational data, maintenancerecord, operational aspect of accessories section ),etc ,are selected as the evaluation factors of fuzzycomprehensive evaluation to build model, which is advantageous to add more effective and sensitivestate parameter in the sequel; At the same time the influence degree to the state of the powertransformer is different according to each evaluation factor, adopt the improved analytic hierarchyprocess (IAHP ) to confirm the weight distribution of every factor in models; according to thecharacteristic of evaluation factors, the relative impairment grade was introduced to reflect theinfluence degree between factors and state of power transformer, then the membership gradefunction of the relative impairment grade to the state of transformer is confirmed by fuzzymathematics; To the state parameter of dissolved gases in oil, because of its one's own characteristicand importance, carry on state parameter reasoning by putting up the neural network to train themembership grade function, which have achieved the preferable result and have worth popularizing.Through the examination of actual transformer show that the assessment method fortransformer condition is reasonable and satisfying. This method has the feature of good practicability,extension and flexibility, offering a way of getting the state of transformer and a guidance oftransformer maintenance.