The performance of the atomic force microscope (AFM), a nanoscale imaging tool, is significantly influenced by the dynamics of the piezoelectric tube scanner (PTS). The limitations of the PTS are its nonlinear behavior in the form of hysteresis and creep effects, low mechanical resonance frequency, and cross-coupling between its two axes of motion. The key focus of this dissertation is to mitigate the limitations of the PTS using controllers to increase the scanning speed. Broadly speaking, this thesis reports four significant contributions for improved scanning using AFMs with piezo tube scanners.The first and unique contribution of this thesis is the design and experimental implementation of a model predictive control (MPC) scheme to achieve accurate tracking for improved image quality at high scanning speeds. Close tracking of the reference signal ensures compensation of the creep and hysteresis effects in the PTS. The MPC control scheme is commonly used when tracking a reference trajectory is the primary goal. This capability of MPC controllers has motivated our research to indirectly compensate the creep and hysteresis effects of the PTS to improve the imaging performance of AFMs. The design of this controller is based on an identified single-input single-output (SISO) model of the PTS, and the AFM's improved imaging quality and scanning speed is illustrated for the design using the proposed control scheme. To evaluate the overall performance improvement achieved by the proposed scheme, an experimental comparison of scanned images from the proposed controller and the existing AFM proportional-integral (PI) controller is undertaken, the results of which verify the efficacy of the proposed design.Tube resonance is one of the major limitations for fast image scanning using AFMs. To overcome this, a damping compensator or notch filter introduced with the proposed MPC controller achieves a high closed-loop bandwidth and significant damping, and enables a reference triangular signal to be tracked. This is the second contribution of this dissertation. By using the proposed controller, the vibration effects in scanned images at high scanning speeds are compensated. Comparisons of the performance of the MPC controller and the modified MPC scheme with a damping compensator are presented.Another reason for the poor performance of the PTS for nanopositioning is that any movement in its x- or y-axis direction produces an undesirable motion in the z-direction called the cross-coupling effect. A multi-input multi-output (MIMO) MPC controller is proposed to compensate for the cross-coupling effect and this is the third contribution of this research. The design is based on an identified MIMO model of the PTS that includes the cross-coupling dynamics. Using the proposed controller a significant reduction in the cross-coupling is achieved and the imaging speed is enhanced up to 125 Hz.Also, a barrier in achieving high scanning speeds is the use of the zig-zag raster pattern scanning technique. In this thesis, a high-speed sinusoidal scanning method, i.e., a spiral scanning method, with an improved MIMO MPC scheme with a damping compensator for faster scanning is proposed and this is the final contribution of this research. The spirals produced have particularly narrow-band frequency which changes slowly over time, thereby making it possible for the scanner to achieve improved tracking and continuous high-speed scanning rather than being restricted to the back-and-forth motions of raster scanning. The experimental results show that, using the proposed method, the AFM is able to scan a 6 μm radius image within 1.42 s with a quality better than that obtained using the conventional raster pattern scanning method.
展开▼
