欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对航空发动机的振动信号，提出了特征基函数ICA提取和SVM相结合的航空发动机故障诊断方法。首先利用ICA从混合振动信号中提取源信号，再利用ICA进一步提取源信号的特征基函数，将特征基函数的频域特性——峰值频率和拐角频率作为特征样本数据用于支撑向量机进行模式识别。采用该方法对某航空涡扇发动机的振动信号进行了分析。分析结果表明，该方法比直接使用SVM的故障诊断准确率高，证明了基函数提取特征的有效性。
英文摘要：
For the vibration signals of aero--engine, a fault diagnosis method based on independent component analysis （ICA） and support vector machine （SVM） is proposed. This feature extraction model is based on using ICA algorithm twice. First, separate the statistical independent source signals of different parts from mixed observing signals, and then extract feature basis functions from the single source signal. Afterwards, the frequency characteristics of the feature basis function, peak frequency and corner frequency, has been introduced as feature sample value into support vector machine for pattern recognition. This method has been used to analyze the vibration signals of a certain turbofan engine. The result shows this method has higher fault diagnosis accuracy comparing with using SVM directly, and proves the validity of using basis functions to extracted features.
同期刊论文项目
同项目期刊论文
期刊信息
位置：