In view of the fact that Fractional Fourier Transform (FRFT) has large computational complexities when the Chirp signals are detected blindly. This paper proposes a blind detection method for Chirp signal parameter estimation. Firstly, the Chirp signal is roughly estimated by window-length changing phase difference method and the least square method is used to fit and obtain the rough estimation value of the Chirp signal frequency modulation slope. Then, the Chirp signal is precisely estimated by FRFT, and the order of FRFT is obtained by Bisection method, which further reduces the computational complexity of the algorithm. The simulation result shows that the method of detecting the Chirp signal by using the modified window-length changing phase difference method and the FRFT at low Signal to Noise Ratio (SNR) has higher precision estimation value, and its normalization mean square error is close to Cramer-Rao lower bound. The method compared with FRFT exhaustive detection greatly reduces the computational complexity and has good engineering application significance.