An energy distribution theory was presented based on regular evolvement of energy fraction of acoustic signals with fluidization velocity. Wavelet packet analysis was used in processing the acoustic signals originated from particle impact on the wall of a fluidized bed. A new criterion of judging incipient fluidization (Umf) velocity and minimum turbulent velocity (Umt) was proposed according to the energy distribution theory. Experiments were performed with five groups of high density polyethylene (PE)particles and one binodal PE to acquire incipient fluidization velocity and minimum turbulent velocity by using the criterion. The feasibility of this method in obtaining characteristic fluidization parameters was further verified by comparing it to results from the pressure drop method and the empirical value from industry.