Yongmin Zhong

Flexure mechanisms, precision actuation, and micro/nanoscale sensing and control are the fundamental building blocks for precision engineering, micro/nanopositioning systems, and instrumentation for micro/nanoscale manufacturing. Furthermore, a growing number of applications require both millimeter-scale workspace range and nanometer-scale resolution from a single positioning stage. However, achieving simultaneous long-range displacement and nanometer-scale precision positioning remains a significant challenge in micropositioning systems. This paper presents a novel hybrid flexure-based mechanism that integrates a voice coil motor (VCM) for a millimeter-scale positioning of a long-range motion stage, with a piezoelectric actuator (PEA) for the high-resolution positioning stage. The two stages are not connected directly: the piezoelectric actuator drives a bridge-type displacement amplifier whose output reaches the platform through decoupling parallelogram flexures, mechanically isolating the precision motion stage from the long-range motion input. Under closed-loop control, the coupled displacement of the hybrid motion is approximately 0.15%, compared with about 0.3% when the long-range stage is driven alone by the VCM, improving positional accuracy and stability during combined motion. This architecture provides a continuous transition between coarse and fine motion positioning without exposing the piezoelectric stack to adverse loading effects. Analytical modeling, finite element analysis, and experimental validation show that the system delivers a travel range of 2 mm with a resolution of 10 nm in the long-range stage and 5 nm in the high-precision stage. Experimental results show that the hybrid mode improves sensing and control specification for the rise time from 32.40 ms to 18.90 ms, the settling time from 51.90 ms to 32.10 ms, and the overshoot from 24.65% to 5.70% relative to long-range motion actuation alone.

Purpose: Virtual reality-based robotic surgery training has received significant attention in recent years due to its numerous advantages, notably improved safety, an enhanced learning experience, and reduced cost. The visual realism and user immersiveness of such platforms are enhanced through the integration of appropriate soft tissue deformation models. Methods: This study presents a modified mass-spring-damper framework designed to provide stable, realistic soft tissue deformation while maintaining real-time performance, even for high-density mesh models. The proposed framework extends the conventional mass-spring model by introducing two kinds of spring-damper elements: deformation and restoring components. Optimization of the model parameters is performed through a combination of analytical derivation and empirical tuning. Various numerical simulation studies are performed to assess model restoring capability, numerical stability, and real-time performance. Results: The results show that the model produces physiologically realistic deformation responses, regains its initial shape characteristics when the external force is removed, and provides a stable response in real-time simulation. A high performance rate of 171.11 frames per second is achieved on high-density mesh models consisting of approximately 29,754 vertices. Moreover, the deformation solver consistently maintains an average update frequency of 2828.14 Hz, with a mean step time of 0.354 ms, demonstrating its real-time capability. Additional experiments involving synthetic tissue and a surgical end-effector validate the tissue response to external forces. Conclusion: The ability of the proposed framework to deliver real-time performance on high-density mesh models highlights its suitability for haptic-enabled robotic surgical training environments that demand both computational efficiency and visual realism.

Engineering computer graphics education frequently exhibits a gap between procedural CAD software (e.g., CATIA 2022) training and the strategic engineering reasoning required by industrial practice. This paper documents a holistic redesign of two advanced CAD courses. The study is framed within the Scholarship of Teaching and Learning (SoTL) tradition as a practitioner-led reflective case study. The redesign integrates four pedagogical mechanisms within an enterprise-CAD context: authentic problem-based learning, dual-layered asynchronous video scaffolding, software-agnostic heuristics (including pre-modelling cognitive mapping), and cognitive apprenticeship. The analysis triangulates three institutional data sources: quantitative Course Experience Survey indicators, qualitative student response themes, and twelve consecutive years of cohort-level academic performance records (2013–2024). The 2022 intervention iteration coincided with a marked elevation in academic performance. Grades reached approximately two standard deviations above the historical baseline. Concurrently, qualitative themes highlighted perceived industrial relevance and platform-portable confidence. However, performance in the post-intervention iterations (2023 and 2024) partially regressed. While scores remained above the historical mean, they did not sustain the 2022 peak. This pattern indicates partial sustainment, rather than evidence of a stable or definitive sustained pedagogical effect. This case is reported as descriptive rather than inferential. While the observed patterns align strongly with theoretical predictions, they do not establish definitive causal effects. Ultimately, the primary contribution of this study lies in documenting the integrated operationalization of these four mechanisms. Furthermore, it highlights longitudinal pedagogical sustainability as a critical, under-examined dimension that single-iteration evidence systematically obscures.

To achieve autonomous and reliable all-weather cross-domain aerospace navigation, this study proposes an adaptive fault-tolerant federated Kalman filter (AFTFKF) for an INS/SRNS/CNS integrated navigation system to enhance system robustness against measurement outliers. First, a noise estimator based on maximum likelihood estimation (MLE) and aided by a sequential probability ratio test (SPRT) is introduced to handle slowly growing outliers. Second, a double residual-based Chi-square test (DCST) information factor is designed to mitigate the impact of inaccurate local state estimation in subsystems under abruptly changed outliers. Finally, the SPRT-MLE-based noise estimator and the DCST-based information factor are integrated into the federated Kalman filter framework to construct the complete AFTFKF. Simulation results demonstrate that the proposed method achieves superior accuracy and strong stability for SINS/SRNS/CNS integrated navigation in the presence of outliers.

Kalman filter is commonly used in INS/GNSS (inertial navigation system / global navigation satellite system) integration to achieve navigation solutions. However, GNSS signals are often vulnerable to adverse environmental conditions, leading to measurement outliers that cause changes to noise statistics. These variations degrade the Kalman filtering estimation and consequent INS/GNSS integrated navigation solution. This paper investigates a new kernel function-weighted robust Kalman filter for INS/GNSS integrated navigation system against time-varying measurement noise statistics. Assuming the measurement noise covariance follows the inverse Wishart distribution, a state estimation equation is formulated to handle constant or slightly varying measurement noise statistics. To extend this state estimation equation for handling abruptly varying measurement noise statistics, weights are constructed based on the Gaussian kernel function with an adaptive bandwidth. The proposed robust Kalman filter can be simply implemented by plugging the state estimation equation into the Kalman filter framework. Simulations and experiments along with comparison analyses exhibit the superiority of the proposed filter for INS/GNSS integrated navigation against both slightly and abruptly varying measurement noise covariances.