基于贝叶斯算法的电力变压器全寿命周期故障概率预测

Power transformer life cycle fault probability prediction based on Bayesian algorithm

为实现对电力变压器全寿命周期故障行为的准确预测,针对基于贝叶斯算法的电力变压器全寿命周期故障概率预测方法展开研究。根据电信号增益量,计算变压器评估指标的权重水平,联合贝叶斯算法,确定健康指数参量的数值情况,完成对电力变压器全寿命周期的评估。分别从全寿命周期成本、变压器故障成本两个角度着手,求解故障概率系数表达式,完成电力变压器全寿命周期故障概率预测模型的建立。对比实验结果表明,在贝叶斯算法作用下,可靠度指标计算数值体现了明显增大的变化趋势,与深度学习算法相比,该模型能够更加准确地预测电力变压器的全寿命周期故障表现行为,符合实际应用需求。

In order to realize the accurate prediction of the fault behavior of power transformers in the whole life cycle,a method for predicting the failure probability of power transformers in the whole life cycle based on Bayesian algorithm is studied.According to the gain of the electrical signal,the weight level of the transformer evaluation index is calculated,and the Bayesian algorithm is combined to determine the value of the health index parameter to complete the evaluation of the full life cycle of the power transformer.Starting from the two perspectives of life cycle cost and transformer fault cost,the expression of failure probability coefficient is solved,and the establishment of the failure probability prediction model of power transformer life cycle is completed.The comparative experimental results show that under the action of the Bayesian algorithm,the calculated value of the reliability index shows a significantly increasing trend.Compared with the deep learning algorithm,the model can more accurately predict the life cycle fault behavior of power transformers,in line with practical application requirements.