This paper studies the low-order dynamic modeling and active control of a flexible plate and provides experimental verification. First based on the input-output data of the system, the Markov parameters of the system are identified using the method of observer/Kalman filter identification (OKID). Then a low-order state-space model is built using the eigensystem realization algorithm (ERA). Finally, a linear quadratic Gaussian (LQG) controller is designed based on the low-order state-space model. Experimental results have proved the effectiveness and feasibility of the research. This paper studies the low-order dynamic modeling and active control of a flexible plate and can provide experimental verification. First based on the input-output data of the system, the Markov parameters of the system are identified using the method of observer / Kalman filter identification Then, a low-order state-space model is built using the eigensystem realization algorithm (ERA). Finally, a linear quadratic Gaussian (LQG) controller is designed based on the low-order state-space model. proved the effectiveness and feasibility of the research.