论文部分内容阅读
本文应用时间序列分析方法对一个转子—轴承—基础框架系统模型的无阻尼固有频率和阻尼比进行识别,用伪随机信号对上述模型进行激振,在 PS—600数据处理系统上进行处理,用 FORTRAN—PS600语言编制了计算程序,用频率分段处理的方法得到振动系统的特征值,其结果与传统的正弦激振和曲线拟名的结果作了比较,两者吻合。为了进一步模拟工程实际情况,撤去激振系统,将轴运转,在稳态下记录框架和轴承座响应,同样用本方法进行参数识别,并将结果与锤击法及上述方法所得结果作比较,证实了本方法能成功地“在线”识别振动系统在运行工况下的真实动力特性,并体现了本方法的可靠性和优越性。
In this paper, the time-series analysis method is used to identify the undamped natural frequency and damping ratio of a rotor-bearing-foundation frame system model. The pseudo-random signal is used to excite the above model and processed on the PS-600 data processing system. The program of FORTRAN-PS600 is programmed, and the eigenvalues of the vibration system are obtained by the method of frequency segmentation. The results are compared with the results of the traditional sinusoidal excitation and curve fitting. In order to further simulate the actual situation of the project, the excitation system is removed, the shaft is operated, the frame and the bearing seat are recorded under steady state, and the parameters are also identified by this method. The results are compared with those obtained by the hammering method and the above method. It is proved that the method can successfully identify the real dynamic characteristics of the vibration system under operating conditions “online” and reflects the reliability and superiority of the method.