烟气轮机是炼化企业生产中重要的核心设备,保障烟气轮机的安全可靠运行,实现科学维护十分重要。对多传感器信息融合技术作了简要介绍,给出了一种神经网络与D-S证据理论相结合的诊断方法,并提出了基于多传感器信息融合技术的烟气轮机故障诊断模型。通过在转子实验台上的实验表明,采用该方法提高了故障诊断系统的灵活性及故障诊断的效率和准确性。
Stack gas turbine is core equipment in production, so scientific maintenance of the turbine appears to be important to ensure its safe and reliable operation. The muhi-sensor information fusion algorithm is briefly introduced ;a neural network with the D-S evidence of a combination of diagnostic methods is provided ; and a multi-sensor fusion-based algorithm fault diagnosis of the gas turbine model is proposed, Experiment result shows that the method enhances the system flexibility,the efficiency and accuracy of the fault diagnosis.