基于贝叶斯网络的变风量末端故障诊断方法

Fault diagnosis method of variable air volume terminals based on Bayesian network

针对压力无关再热型变风量末端的15种典型故障,提出了一种基于贝叶斯网络的故障诊断方法。根据某建筑实际运行系统建立了Dymola仿真模型,并基于模拟故障数据对所提出的诊断方法进行了验证。结果表明,该方法对于绝大多数故障都可以成功检测并分离,有着较高的准确性和可靠性,并且可较好地应对实际工程中存在的数据问题,将实时故障诊断的应用场景进一步推广。

Aiming at 15typical faults of the pressure independent variable air volume(VAV)terminal with reheat coil,proposes a Bayesian network-based fault diagnosis method(FDD).Establishes a Dymola simulation model based on an actual VAV system,and verifies the performance of the proposed method based on simulated fault data.The results show that this method performs well at:(1)detecting and isolating most faults with high accuracy and reliability,(2)dealing with the data problems existing in actual engineering,(3)further popularizing the application of real-time fault diagnosis.