In order to improve the denoising effect of low frequency oscillation signals and provide the accurate and reliable data for detection and analysis of low frequency oscillation signals, the denoising method based on tunable Q-factor wavelet transform via sparse representation is given on the analysis of tunable Q-factor wavelet transform and sparse representation theories. Firstly, the tunable Q-factor wavelet transform is adopted to perform the signal sparse decomposition for the noisy low frequency oscillation signals, and the initial wavelet coefficients are obtained; secondly, the BP denoising algorithm is used to optimize the obtained wavelet coefficients; lastly, the optimized wavelet coefficients are reconstructed, then the low frequency oscillation signal without noisy is obtained. After the computer simulation, the result demonstrates that this method is superior to the current widely used wavelet soft-threshold and hard-threshold in denoising effect and reliability.