水轮机故障噪声特征分析与识别方法.pdfVIP

水声工程电声技术

nderwaterAcousticEngineeringudioEngineering

文献引用格式：刘皓.水轮机故障噪声特征分析与识别方法[J].电声技术，2025，49（4：70－72.

LIUH.CharacteristicsanaandidentificationmethodoffaultnoiseinwaterturbinelJj.Audi，49(4)：

70-72.

中图分类号：TL375.2文献标识码：AD0I:10.16311/j.audioe.2025.04.021

水轮机故障噪声特征分析与识别方法

刘皓

（宜昌利民电建工程有限公司，湖北宜昌443003)

摘要：分析水轮机故障噪声特征，提出一种基于声学特征分解与协同辨识的水轮机故障识别方法。通过分布式声压传感器阵

列实现机械噪声与流体动力噪声的物理分离。结合多分辨率分析与空时域增强技术，构建故障特征解耦框架。采用动态权

重分配与多维度决策融合机制，建立从声学特征到故障模式的端到端映射模型。实验结果验证了该方法在复杂流场环境下

故障识别的准确性与实时性。

关键词：水轮机；故障噪声特征；故障识别方法

CharacteristicsAnalysisandIdentificationMethodofFaultNoiseinWaterTurbine

LIUHao

(YichangLiminElectricConstructionEngineeringCo.,Ltd.,Yichang443003,China)

Abstract:Analyzedthenoisecharacteristicsofwaterturbinefaultsandproposedawaterturbinefaultidentificationmethod

basedonacousticfeaturedecompositionandcollaborativeidentification.Thephysicalseparationofmechanicalnoiseandfluid

dynamicnoiseisachievedthroughadistributedsoundpressuresensorarray.Afaultfeaturedecouplingframeworkisconstructed

bycombiningmulti-resolutionanalysisandspatiotemporalenhancementtechnology.Adynamicweightallocationandmulti-

dimensionaldecisionfusionmechanismareadoptedtoestablishanend-to-endmappingmodelfromacousticfeaturestofault

modes.Theexperimentalresultsverifytheaccuracyandreal-timeperformanceofthismethodinfaultidentificationincomplexflow

fieldenvironments.

Keywords:waterturbine;faultnoisecharacteristics;fault