Enhanced Odd-Harmonic Repetitive Control of Nanopositioning Stages Using Spectrum-Selection Filtering Scheme for High-Speed Raster Scanning

Odd-harmonic repetitive control (ORC) has been successfully applied to improve the triangular trajectory tracking performance of nanopositioners. However, the conventional ORC tends to amplify the tracking errors at frequencies other than the odd-harmonic components, mainly the even harmonics of the fundamentals of the intended triangular trajectory to be tracked. Due to the influence from the hysteresis nonlinearity of the piezoelectric actuator, this would result in significant tracking errors. To overcome this limitation, this article proposes an enhanced odd-harmonic repetitive control (EORC) using the spectrum-selection filtering scheme to improve the loop-shaping property of the ORC. This effectively eliminates the problem of amplifying the tracking errors while preserving the advantages of the conventional ORC, such as fast convergence speed and low computation cost. The EORC is combined with a proportional-integral tracking controller to improve the tracking performance. The controller design, stability analysis, and performance evaluation are presented. The experimental results demonstrating the effectiveness of the proposed EORC-based control scheme are presented showing the excellent tracking of triangular trajectories with fundamental frequencies up to 1000 Hz. Moreover, a reduction in rms tracking errors by up to 52% is achieved. Note to Practitioners -The high-speed atomic force microscopy (AFM) plays an increasingly vital role in observing and manipulating objects at the nanoscale. In the raster scanning of AFMs, the most challenging issue is the triangular trajectory tracking of nanopositioning stages with high precision. Although the odd-harmonic repetitive control (ORC) has been successfully applied to improve the tracking performance, the conventional ORC would amplify the tracking errors distributed at the frequencies other than the odd harmonics, especially those at the even harmonics, resulting from the inherent complicated hysteresis nonlinearity. This problem of amplifying the tracking errors would lead to larger tracking errors, deteriorating the performance of AFM imaging. To address this issue, this article proposes an enhanced ORC using the spectrum-selection filtering scheme to improve the loop-shaping property of the ORC, so as to eliminate the problem of amplifying the tracking errors while preserving the advantages of the conventional ORC, such as fast convergence speed and low computation cost. The experimental results show