基于贝叶斯网络的智能电能表可靠性预计研究

可靠性作为智能电能表正常工作的重要指标,是现阶段电能质量研究的重点问题,针对智能电能表可靠性预计的重要性为切入点,讨论电能表可靠性的研究过程,以贝叶斯网络为依据,分析了传统的贝叶斯网络的可靠性预计双向推理的方法,并且同时结合故障树的模型理论,在可靠性研究的求解过程中引入了新的桶消元法对贝叶斯网络进行计算简化,从而快速得到可靠性预计结果。最后通过提出的方法,应用于电能表可靠性预计实例当中,计算的结果满足GJB/Z299C-2006标准在误差范围内的要求,从而证明,该方法简化预计过程的同时准确的预计出可靠性结果。

Research on reliability prediction of smart meter based on Bayesian network

The reliability,as an important indicator of the normal work of smart meter,is the focus of power quality research at this stage of the problem;this paper aims at the importance of smart meter reliability prediction as the starting point to discuss the reliability of the energy meter research process.The reliability prediction bi-direction reasoning in traditional method of Bayesian network is analyzed,combined with the model theory of the fault tree,a new bucket elimination method is introduced to simplify the calculation of Bayesian network in the process of solving the reliability problem,and thus the expected results of reliability is obtained quickly.Finally,the method proposed in this paper is applied to the case of reliability prediction of electric energy meters.The results of the calculation meet the requirement of GJB / Z 299C-2006 standard in the error range,which proves that the method can simplify the prediction process and accurately predict the reliability at the same time.