Along with the fast deployment of bike-sharing systems (BSS) nowadays is the bike imbalance problem induced by the demand variability. In this study, we design a user-based bike rebalancing (UBR) strategy to solve the bike imbalance problem in a free-floating bike sharing system (FBSS), where the user arrival and incentive budget allocation of the FBSS are both dynamic. Specifically, we design a dynamic bidding-model-based incentive mechanism (BIM) to progressively determine the incentive price, based on which users are assigned the rebalancing task. The proposed BIM is proved to be budget feasible, incentive compatible, and competitive. We use a set of bike sharing data to numerically verify the effectiveness of BIM in improving the service level of FBSS. Furthermore, we find that the BIM outperforms the post-price-based incentive mechanism (PIM), with its advantage negatively correlated with user irrationality. Some managerial insights and potential research opportunities are provided for urban planners, policy makers and BSS practitioners.