PROJECT TITLE :

Neural-Network-Based Nonlinear Model Predictive Control for Piezoelectric Actuators

ABSTRACT:

Piezoelectric actuators (PEAs) are widely used in nanotechnology due to their characteristics of fast response, massive mass ratio, and high stiffness. However, hysteresis, which is an inherent nonlinear property of PEAs, greatly deteriorates the management performance of PEAs. During this paper, a nonlinear model predictive control (NMPC) approach is proposed for the displacement tracking drawback of PEAs. 1st, a “nonlinear autoregressive–moving-average with exogenous inputs” (NARMAX) model of PEAs is implemented by multilayer neural networks; second, the tracking management drawback is converted into an optimization drawback by the principle of NMPC, and then, it's solved by the Levenberg–Marquardt algorithm. The most distinguished feature of the proposed approach is that the inversion model of hysteresis is now not a necessity, which avoids the inversion imprecision drawback encountered within the widely used inversion-primarily based control algorithms. To verify the effectiveness of the proposed modeling and management methods, experiments are created on a business PEA product (P-753.1CD, Physik Instrumente), and comparisons with some existing controllers and a commercial proportional–integral–spinoff controller are conducted. Experimental results show that the proposed scheme has satisfactory modeling and management performance.