POD方法在转子系统降维中的应用(英文)

将POD(Proper Orthogonal Decomposition)方法应用到转子动力系统中.建立一端松动的7个自由度非线性转子模型,考虑到混沌信号中包含各种不稳定的周期轨道,因此含有较多的系统信息,利用POD方法从混沌信号中获得一组POMs,并将原系统投影到该组POMs(Proper Orthogonal Modes)上,得到原系统4个自由度的近似等效模型,通过降维前后模型的定性性质比较,包括轴心轨迹、相图、分岔图等的比较可以看出降维后系统较好的保持了原系统的动力学特性,说明POD方法对于该模型降维是有效的.

Application of POD method to reduce dimensions of rotor system

The POD(Proper Orthogonal Decomposition) method was applied in a rotor bearing system.For a simple rotor-bearing system with pedestal looseness at one support,considering that a chaotic orbit visits many unstable periodic orbits and hence should contain more information about the system than a single periodic orbit,after applying the POD method,the POMs(Proper Orthogonal Modes) were obtained from the chaotic signals.Then four-degree-of-freedom reduction model was obtained by projecting the system on these POMs.The reduced model was validated using qualitative mean(including comparisons of the phase portraits and the orbits of shaft center and the occurrence of the bifurcations).The results confirm the validity of the POD method.