基于油色谱监测数据的变压器动态可靠性分析

动态可靠性是电力变压器进行短期风险评估、检修决策等的依据.文中首先分析能反映变压器可靠性的因素,选择油中溶解气体分析(DGA)数据中特征气体含量、气体总量产气速率、设备役龄为关键影响因素.然后采用最小二乘支持向量机作为动态可靠性模型,进行变压器动态故障率预测.最后用算例分析了影响因素和模型的合理性,并与采用马尔可夫状态空间模型计算的故障率结果进行了比较.结果表明,设备役龄是影响变压器内部潜伏性故障率的重要因素,最小二乘支持向量机方法作为变压器动态可靠性模型具有计算速度快、监测信息的识别度高的优点.

Analysis of Dynamic Reliability of Transformer Based on Monitoring Data of Oil Chromatography

Dynamic reliability is the basis of short-time risk evaluation and maintenance decisionmaking for transformers. The paper first analyzes the factors that influence the failure rate of transformer by considering the dissolved gas analysis (DGA) characterized gas content, velocity of total gas production and enlistment age as the main influencing factors. Then the paper establishes dynamic reliability model based on least-square support vector machines (LS-SVM) to predict the dynamic failure rate of the transformer. Finally, a specific case is studied to prove the rationality of chosen factors and model. The result calculated by the proposed method is then compared with that by Markov method, proving that the enlistment age is the important factor of internal latent failure of transformer, and the transformer dynamic reliability model based on LS-SVM has the advantages of faster calculation and higher recognition of monitoring information.