Extract and Explore Chaos Characters of an Abnormal Turbo-Unit

The nonlinear characteristic behaviors of failure turbo-unit are analyzed and researched in detail. Comparing and exploring the correlation dimensions and the largest Lyapunov exponents that extracted from time series signals of the rotating machinery at different running states deeply, we find that the information blended in the signals are very useful to explode the working state of the rotating machinery. Different running state of turbo-unit shows different correlation dimension and different largest Lyapunov exponent distinctly. Our research authenticates that if we use only one numerical characteristic between fractal dimension and Lyapunov exponent, the running state, about the failure turbo-unit, that is deduced by us could be incorrect. This result was validated by locale practicality. Hence, we present our conclusion that making use of correlation dimension and the largest Lyapunov exponent combined with other information of the complex rotary dynamical system is a better way to diagnose turbo-unit at any running time. This approach is a new and a better way to research other failure complex mechanism.