Due to the special working condition of the wind turbine, it is very difficult to detect the weak intermittent fault growing state. In order to solve this problem, a kind of weak fault amplitude change detection method is proposed in this paper. First, the relationship between the amplitude of weak intermittent fault and the built-in signal amplitude of the large scale periodic status in the fractional order chaotic system is found through the experiment. Second, density mean algorithm was used to calculate poincare section data points of critical large-scale periodic state of Duffing system. The outliers are obtained and the system is judged to be in a large scale periodic state, and then automatic detection of different amplitudes of weak fault is completed. Last, the simulation results show that the method is feasible and reliable. This method provides a new idea for the theoretical research of intermittent fault.