Abstract:Duffing oscillator is extremely sensitive to the periodic signal and has strong immunity to noise. So it is used for the detection of weak periodic signal in strong noise background and the measurement of its parameters. In order to further improve its detection performance, a new method of Duffing oscillator weak signal detection based on signal preprocessing is proposed and is proved in theory. The new method firstly processes the signal by segmentation, stacking, extension then sends the processed signal to Duffing oscillator to detect. Simulation results show that the proposed method can obtain a large SNR gain, when the segmentation number is three, the lower limit of detection SNR reduces about 2dB, when the number of segmentation is five, the lower limit of detection SNR reduces about 3dB.