The aero-engine electrostatic monitoring technology shows a high fault warning capability. However, the raw electrostatic signal usually contains unexcepted noise. To improve the accuracy of the information extraction, the electrostatic signal should be denoised. This study firstly introduces the principle of electrostatic monitoring technology and analyzes the sources and principal components of the noise. To address the problem of coupling noise filtering of electrostatic signal, the theories of sparse signal representation and empirical mode decomposition are introduced. The basis and criteria for mode functions selection are investigated, and a joint denoising algorithm and process based on the mode function optimized reconstruction and sparse representation are proposed. The actual electrostatic signal acquired in the turbofan engine test run is used to evaluate the denoising effectiveness of the proposed algorithm. Meanwhile, it is compared with other classic methods. The results show that the proposed method can effectively remove the random noise and power frequency interference while remaining the useful particle signal to a large degree. The abnormal signal can be extracted well when the number of sparse iterations is set from 20 to 50.