In many active noise and vibration control systems it is desired to reduce the vibration and also the noise emitted by flexible structures. With this objective the task of the controller is to control the maximum number of modes allowed by the control system's limitations. Some of the key parameters in the control system design for this purpose are the location, number and size of actuators and sensors on a flexible structure. In this paper a simply-supported thin plate with laminated piezoelectric sensors and actuators is studied for noise and vibration attenuation. For this purpose, a performance index based on Hankel singular values of the system is selected. The resulting nonlinear optimization problem is solved using a new particle swarm optimization (PSO) algorithm. The results are compared with a genetic algorithm and verified with a simple state-feedback controller.