In view of the time-frequency representation (TFR) reconstruction of the micro-motion signals under conditions of incomplete data, a reconstruction method of micro-motion is proposed based on the adaptive parameter estimation. Firstly, the missing micro-motion TFR reconstruction problem is modeled on the LP norm minimization sparse reconstruction problem, and by introducing Hadamard product parameter (HPP), the LP norm minimization sparse reconstruction problem is transferred to a joint minimization problem with multiple L2 norm, and solved by using iterative Tikhonov regularization. Simultaneously, the regularization parameter is estimated adaptively in each iteration based on reconstruction results. Finally, the amplitude decay of the reconstructed TFR is reduced by the de-biasing process. Compared with the traditional micro-motion echo time-frequency representation reconstruction method, the proposed method avoids the disadvantage of setting the regularization parameter manually, and the reconstructed TFR is more complete. The effectiveness and robustness of the proposed method is verified by simulation and measured data processing