Covert communication can achieve the purpose of hiding information transmission behavior without being detected by Willie, greatly improving the security of wireless communication, and has a broad application prospect in military and civilian. In order to solve the problem that the existing weighted fractional Fourier transform (WFRFT) covert schemes are not strong in anti-scanning performance and easy to be cracked by violence, which brings challenges to the covert performance of the system, this paper proposes a chaotic sequence-driven dynamic multi-layer generalized multi-parameter WFRFT (ML-GMPWFRFT) covert communication method. The method constructs ML-GMPWFRFT on the basis of traditional WFRFT system, and then uses chaotic sequence to spread spectrum and dynamically determine the type combinations of ML-GMPWFRFT, thus realizing covert communication with stronger scanning resistance ability. The simulation results show that the signal processed by this method has the characteristics of Gaussian-like constellation distribution and low energy, which means good covert performance. And compared with the existing covert communication methods, this method enhances the complexity and anti-interception performance of the system, protecting the covert system better.