We present an experimental study on characterizing the dynamic behavior of flow velocity field during combustion noise from the viewpoints of statistical complexity and complex-network theory, involving the detection of a precursor of blowout. The multiscale complexity-entropy causality plane clearly shows the possible presence of two dynamics, (1) stochastic dynamics in the injector rim region and (2) noisy chaos in the shear layer between a vortex breakdown bubble in a wake of a centerbody and the outer recirculation region in a dump plate. The turbulence network is used to study the vortical interactions during combustion noise. The weighted permutation entropy incorporating the amplitude information of pressure fluctuations is useful for capturing a precursor of blowout.

我们从统计复杂性和复杂网络理论的角度，介绍了表征燃烧噪声过程中流速场动态行为的实验研究，其中包括检测井喷的前兆。多尺度复杂性-熵因果关系平面清楚地显示了两种动力学的可能存在，（1）喷油器边缘区域中的随机动力学，以及（2）中心体和外部涡流破裂后的气泡之间的剪切层中的嘈杂混沌转储板中的再循环区域。湍流网络用于研究燃烧噪声过程中的涡旋相互作用。结合了压力波动的幅度信息的加权置换熵对于捕获井喷的前兆很有用。