In this paper, a novel design concept for active self-adaptive metamaterial (ASAMM) plates is proposed based on an active self-adaptive (ASA) control strategy guided by the particle swarm optimization (PSO) technique. The ASAMM plates consist of an elastic base plate and two periodic arrays of piezoelectric patches. The periodic piezoelectric patches place on the bottom plate surface act as sensors, while the other ones attached on the top plate surfaces act as actuators. A simplified plate model is established by the Hamilton principle. By assuming a uniform or constant plate thickness, the plane wave expansion (PWE) method is adopted to calculate the band structures. The finite element method (FEM) using 2D plate and 3D solid elements is also used to calculate the band structures and the transmission spectra or frequency responses. The conventional displacement, velocity and acceleration feedback control methods are introduced and analyzed. Then, a novel ASA control strategy based on combining the displacement and acceleration feedback control methods and guided by the PSO technique is developed. Numerical results will be presented and discussed to show that the proposed ASAMM plates can automatically and intelligently evolve different feedback control schemes to adapt to different stimulations on demand. Compared to the conventional metamaterial (MM) plates, the proposed ASAMM plates exhibit improved and enhanced band-gap characteristics and suppression performance for flexural waves at frequencies outside the band-gaps