In this paper, a novel micro vibration reconstruction method is proposed to mitigate the influence of in-range large-scale dynamic clutters for radar sensing. In order to understand the impact of dynamic clutters on the target signal, a bi-exponential radar signal model is introduced to explicitly characterize the superposition mechanism between the target and clutter vector signals. Theoretical analysis shows that strong dynamic clutters would overwhelm the target signal and further distort the sensing result when using conventional demodulation methods. To solve this problem, a vector separation technique is proposed via accurate estimation of both vector signals and proper phase unfolding. Leveraging an Impulse-Radio Ultra-Wideband (IR-UWB) radar, experiment results show that the proposed technique could effectively recover the micro mechanical vibration in the presence of a strong in-range dynamic clutter.