In this paper, a modified repetitive control (MRC) based approach is developed for high-speed tracking of nanopositioning stages. First, the hysteresis nonlinearity is decomposed as a periodic disturbance over a linear system. Then, the MRC technique is utilized to account for the periodic disturbances/errors caused by the hysteresis and dynamics behaviors. The developed approach provides a simple and effective hysteresis compensation strategy, avoiding the constructions of hysteresis model and its inversion. Besides, with improved loop-shaping properties, the MRC can alleviate the nonperiodic disturbance amplification problem of the conventional repetitive control. Finally, the effectiveness and performance of the developed MRC-based approach are verified by the experimental results on a custom-built piezo-actuated stage in terms of hysteresis compensation, disturbance rejection and tracking accuracy.
Deep learning-based classification algorithms are promising in gesture recognition with soft e-skin patches. However, the reported algorithms usually require large amount of training data, resulting in the time-consuming data collection process. In this paper, we present a deep transfer learning-based adaptive strategy for accurate gesture recognition of a soft e-skin patch with reduced training data and time. To this end, we first train a base neural network as the general feature extraction network. Next, we transfer the front layers of the pre-trained base network to target networks of new gesture recognition tasks. Further, we apply the fine-tune technique to refine the copied parameters. Finally, with our custom-built soft e-skin patch, we experimentally verify the developed strategy on two typical transfer cases, termed as the user transfer case (Case I) and the gesture transfer case (Case II). The experimental results show that, to ensure the stable accuracy of 95 %, the training data with and without the adaptive strategy are 1,312 vs 10,912 for Case I, and 8,192 vs 12,032 for Case II, respectively. In this sense, the training time of target networks can be reduced by 62.96 % for Case I and 34.20 % for Case II, respectively. This work shows the potential to promote the widespread application of e-skins in human computer interaction.
In raster scanning applications of atomic force microscopies, precisely tracking periodic triangular trajectories is the major objective of nanopositioning stages. Considering the fact of periodic operations, the repetitive control technique becomes promising and has been recently developed to reduce tracking errors. In our new experiments, it is found that, with triangular reference input, the hysteresis nonlinearity mainly affects the system at the odd harmonics of the input signal. In this sense, an odd-harmonic repetitive control (ORC) strategy is proposed to handle the hysteresis nonlinearity, with the hysteresis treated as the odd-harmonic periodic disturbance. Therefore, it avoids the modeling and inverting of the complex hysteresis nonlinearity. Another benefit of the developed ORC strategy is that it can also account for the tracking errors caused by the linear dynamics effect. To verify the effectiveness of the ORC strategy, real-time experiments are performed on a custom-built piezo-actuated nanopositioning stage. Experimental results show that the developed ORC strategy achieves precise tracking of 1562.5-Hz triangular trajectory with the hysteresis nonlinearity mitigated to a negligible level, which demonstrates the feasibility and effectiveness of the developed ORC strategy on hysteresis compensation during high-speed raster scanning.
A robust control approach with the inverse backlash compensation is presented for a class of non-linear systems preceded by unknown asymmetric backlash non-linearity. Firstly, the analytical expressions of the inverse compensation error for an asymmetric backlash are obtained by introducing new indicator functions, which make it possible to design a corresponding controller for the asymmetric input backlash. With the developed compensation error expression, conventional robust control approaches can be utilised to deal with such a non-smooth non-linear system. As an illustration, a robust adaptive control strategy is applied to demonstrate the approach. The developed control laws ensure the robust inverse compensation and achieve tracking within a desired accuracy. Finally, simulations performed on an unstable and uncertain non-linear system illustrate and clarify the effectiveness of the developed approach.
The repetitive control (RC) has been employed for high-speed tracking control of nanopositioning stages due to its abilities of precisely tracking periodic trajectories and rejecting periodic disturbances. However, in digital implementation, the sampling frequency should be integer multiple of the tracking frequency of the desired periodic trajectory. Otherwise, the rounding error would result in a significant degradation of the tracking performance, especially for the case of high input frequencies. To mitigate this rounding effect, the fractional repetitive control (FRC) technique is introduced to control the nanopositioning stage so as to precisely track high-frequency periodic inputs without imposing constraints on the sampling frequency of the digital control system. The complete procedure of controller design and implementation is presented. The techniques to deal with the problems of non-minimum phase system and fractional delay points number are described in detail. The proposed FRC is plugged into the proportional-integral control, and implemented on a custom-built piezo-actuated nanopositioning stage. Validation experiments are conducted, and the results show that the tracking errors caused by the rounding effect in the traditional RC approach are almost completely eliminated, when tracking sinusoidal waveforms with frequencies from 1000 Hz to 1587.3 Hz under the sampling frequency of 50 kHz, where the fractional parts being rounded vary from 0 to 0.5.
In this paper, a fast automatic precision approaching system is developed for electrochemical nanofabrication using visual and force-displacement sensing. Before the substrate is fabricated, the template should approach the substrate accurately to establish the initial gap between the template and substrate. During the approaching process, the template is first quickly moved towards the substrate by the stepping motor until a specified gap is detected by the visual feedback. Then, the successive approach using the switch of macro-micro motion with a force-displacement sensing module is triggered to make the template contact with the substrate to nanometre accuracy. The contact force is measured by the force-displacement sensing module which employs the high-resolution capacitive displacement sensor and flexure compliant mechanism. The high sensitivity of this capacitive displacement sensor ensures high accuracy of the template-substrate contact. The experimental results show that the template can reach the substrate accurately and smoothly, which verifies the effectiveness of the proposed approaching system with the visual and the force-displacement sensing modules.
This paper presents a new damping control scheme for piezo-actuated nanopositioning stages with recursive delayed position feedback (RDPF). The RDPF is proposed to attenuate the resonant mode of the nanopositioning stage in the inner feedback loop, which results in a neutral-type time-delay system. To realize the pole placement of this system, a new numerical integration method is proposed to determine the rightmost pole and select the parameters of the RDPF. Then, a high-gain proportional-integral (PI) controller is designed in the outer loop to minimize the tracking errors caused by the hysteresis nonlinearity and modeling uncertainties. To validate the effectiveness of the proposed approach, comparative experiments are conducted on a piezo-actuated nanopositioning stage. Experimental results demonstrate that the proposed approach improves the control bandwidth of the system from 32.5 Hz (with the PI controller) and 687 Hz (with the conventional delayed position feedback based controller) to 793 Hz.
Mainly composed of electrical motors and sophisticated mechanical components, existing neuroprosthetic hands1,2 are typically heavy (>400 g) and expensive (>USD 10,000), and they lack the compliance and tactile feedback of human hands. These limitations hamper neuroprosthetic hands’ innovation and broad utility for amputees3-5. Here we report the design, fabrication and applications of a lightweight (292 g) and potentially low-cost (component cost below USD 500) soft neuroprosthetic hand with simultaneous myoelectric control and tactile feedback. The soft neuroprosthetic hand consists of five soft fingers and a palm to give six active degrees of freedom under pneumatic actuation, four electromyography sensors that measure the surface electromyogram signals to control the hand to deliver four common grasp types, and five hydrogel-elastomer capacitive sensors on the fingertips that measure the touch pressure and elicit electrical stimulation on the skin of the residual limb. The soft finger is made of a fiber-reinforced elastomeric structure embedded with rigid segments to mimic the soft-joint/rigid-bone anatomy of the human finger. We use a set of standardized tests6 to compare the speed and dexterity of the soft neuroprosthetic hand and a conventional rigid neuroprosthetic hand7 on two transradial amputees. The soft neuroprosthetic hand gives overall superior performances to the rigid hand. We further demonstrate that one transradial amputee wearing the soft neuroprosthetic hand can regain the versatile hand functions with primitive touch sensation and real-time closed-loop control in daily activities such as handling tools, eating, shaking hands, petting animals, and recognizing touch pressure. This work not only represents a new paradigm for designing soft neuroprosthetic devices but also opens an avenue to widespread applications of lightweight, low-cost, and compliant hand replacements for amputees.
