论文部分内容阅读
为及时可靠地检测出气动失稳现象,保障压气机的安全工作,采用Laplace小波构建原子函数库,引入稀疏分解算法提取阻尼比小于0的信号成份,通过与预设阈值比较内积值的大小,进行气动失稳信号的检测。结合在线检测系统的特点,将Laplace小波蜕化为衰减正弦函数,限定了参数取值范围;并仅进行一步匹配追踪,由此减少了稀疏分解的计算量。采用压气机试验台实测数据验证,该方法可以在压气机进入喘振前49~154ms准确发出报警信号。
In order to detect the aerodynamic instability timely and reliably and ensure the safety of the compressor, the Laplacian wavelet is used to build the atomic function library. The sparse decomposition algorithm is introduced to extract the signal components with damping ratio less than 0. By comparing the inner product value with the preset threshold , The pneumatic instability signal detection. Combined with the characteristics of the on-line detection system, the Laplace wavelet is degenerated into a decaying sine function, which limits the range of parameter values. Only one-step matching pursuit is used to reduce the computational complexity of sparse decomposition. The test data from the compressor test bench verify that this method can accurately send an alarm signal 49 ~ 154ms before the compressor enters surge.