Keywords: 
• 三谱切片 /
• 波动特性 /
• 流型识别 /
• 流型动力学

Abstract: A new method of nonlinear analysis is proposed by combining the sliced trispectrum method with the fluctuation characteristics, and the influences of the length of the time sequence and fast Fourier transform (FFT) on the fluctuation characteristic value of sliced trispectrum are discussed. It is found that in the absence of noise, the length of sequence and FFT are proportional to the vector distance value;in the presence of noise, it is approximate inversely related. In order to test the anti-noise ability and characterize complexity ability of the fluctuation characteristics method, the proposed method is applied to fractal sequence (Brown), chaotic sequence (Lorenz) and periodic sequence (sine signal). The results show that compared with other power spectral method, the method of the fluctuation characteristics has good noise immunity of fractal sequence, and the anti-noise ability is relatively weak with periodic sequence;but the fluctuation characteristic theory of sequence internal complexity representation has a good effectiveness. On this basis, the differential pressure signals of nitrogen-water two-phase flow in small rectangular channel (w×h=2 mm×0.81 mm) are studied. By analyzing the differential pressure sliced trispectrum of flow patterns, the secondary coupling phenomenon of main oscillation mode of different flow patterns is established. The fluctuation characteristic values of the sliced trispectrum of different flow patterns are extracted to accurately identify the typical flow patterns of small channel nitrogen-water two-phase flow. At the same time, the fluctuation characteristic theory can be used to provide a useful exploration for the further investigation of flowing mechanism of multi-phase flows.