Intelligent Modelling of Electromechanical Piezoelectric Actuator

This research models a piezoelectric actuator using ARX model with particle swarm optimization method. System modeling is a crucial step in engineering for understanding system behavior and implementing control strategies especially when employing intelligent and precise modeling techniques. This research contributes by introducing a new type of modeling in this field by employing the Autoregressive with eXogenous inputs (ARX) model and the estimation parameters calculated by PSO method. In current work, 0 to 35 V as sinusoidal frequency wave at 1.2 Hz supplied to piezoelectric actuator (PZT) to collect the vibration as an output for the system using an accelerometer sensor. Then, double integration was employed to collect the displacement data. One side of the PZT was fixed, while the other was left free. MATLAB program 2022b employed to build the ARX model and manage the particle swarm optimization (PSO) variables such as swarm size and iterations to obtain the transfer function that represents dynamic behavior through input and output readings. The results appeared effectively of PSO and representing the accurate transfer function in continuous and discrete times. The PZT validated by recorded the minimum mean square error (MSE) up to 3.447×10-6 and the system behavior was within the confident limit at 95 % before and after the modeling process. The developed model can be used to design a robust vibration controller in the future or for energy harvesting.