Xinyi Yu

It has been a critical challenge to provide reliable performance constraints while ensuring safe stability for robotic systems, particularly in relatively complex human-robot interaction tasks. To overcome the limitation in traditional controllers where control accuracy is highly dependent on the dynamic model, a novel neural network-based data-driven sliding mode controller (NNDSMC) is proposed. It optimizes control strategies in real time by constructing a dynamic data model while considering interaction forces. On this basis, a data-driven prescribed performance control (DPPMC) method is proposed based on quadratic programming (QP), where faster transient response and greater steady-state tracking accuracy are guaranteed with the designed data-driven high-order control barrier function (DHCBF). To further realize compliant human-robot interaction, a novel data-driven adaptive admittance controller is developed, which combines the flexibility of model predictive control with the fast convergence of DPPMC. The whole design process does not involve any model information. The feasibility of the proposed method is validated through the Franka-Panda robot with seven degrees of freedom.

Hysteresis is very common in nonlinear compliant actuators and cannot be neglected. In this type of actuators, the hysteresis effect is associated with applied torque, leading to multiloop asymmetry curves. Although many methods have been developed to describe analytically hysteresis behavior, modeling of nonlinear hysteresis with multiloop asymmetry curves remains very challenging due to the complexity of nonlinear curves. In this article, a novel phase-dependent data-driven hybrid modeling method is proposed to describe accurately this complex hysteresis. The multiloop asymmetry curve is first segmented into ascending, descending, and transition phases, and the transition phase is further separated into a trend curve and a quasi-symmetrical looped curve. A data-driven hybrid model, combining an elementary function and a Maxwell-slip based model, is developed to fit these two separated components. The proposed method was verified by experiments on a nonlinear compliant actuator. The results show that the proposed hysteresis model outperforms a modified Bouc-Wen model, a power-function-based model, and a Fourier neural operator-based model in terms of modeling accuracy. The performance of this method is further evaluated on two additional datasets: a public dataset on material pressure tests and a dataset collected from a variable stiffness actuator, verifying its versatility across different applications.

Object navigation is a core capability of embodied intelligence, enabling an agent to locate target objects in unknown environments. Recent advances in vision-language models (VLMs) have facilitated zero-shot object navigation (ZSON). However, existing methods often rely on scene abstractions that convert environments into semantic maps or textual representations, causing high-level decision making to be constrained by the accuracy of low-level perception. In this work, we present 3DGSNav, a novel ZSON framework that embeds 3D Gaussian Splatting (3DGS) as persistent memory for VLMs to enhance spatial reasoning. Through active perception, 3DGSNav incrementally constructs a 3DGS representation of the environment, enabling trajectory-guided free-viewpoint rendering of frontier-aware first-person views. Moreover, we design structured visual prompts and integrate them with Chain-of-Thought (CoT) prompting to further improve VLM reasoning. During navigation, a real-time object detector filters potential targets, while VLM-driven active viewpoint switching performs target re-verification, ensuring efficient and reliable recognition. Extensive evaluations across multiple benchmarks and real-world experiments on a quadruped robot demonstrate that our method achieves robust and competitive performance against state-of-the-art approaches.The Project Page:https://aczheng-cai.github.io/3dgsnav.github.io/

The Contrastive Language-Image Pretraining (CLIP) model has been widely used in various downstream vision tasks. The few-shot learning paradigm has been widely adopted to augment its capacity for these tasks. However, current paradigms may struggle with fine-grained classification, such as satellite image recognition, due to widening domain gaps. To address this limitation, we propose retrieval-enhanced visual prompt learning (RePrompt), which introduces retrieval mechanisms to cache and reuse the knowledge of downstream tasks. RePrompt constructs a retrieval database from either training examples or external data if available, and uses a retrieval mechanism to enhance multiple stages of a simple prompt learning baseline, thus narrowing the domain gap. During inference, our enhanced model can reference similar samples brought by retrieval to make more accurate predictions. A detailed analysis reveals that retrieval helps to improve the distribution of late features, thus, improving generalization for downstream tasks. RePrompt attains state-of-the-art performance on a wide range of vision datasets, including 11 image datasets, 3 video datasets, 1 multi-view dataset, and 4 domain generalization benchmarks.

An analytical method is presented to find the stability domain of PID controller parameters for nonlinear time-delay systems by utilizing the Volterra series and its generalized frequency response function (GFRF), and all the parameters in the derived region can stabilize the system. According to the conditions for closed-loop L ₂ stabilization, the procedures are derived to determine the PID controller stabilization region. The simulation results show the accuracy of the stability domain. Meanwhile, this proposed method preserves the nonlinear characteristics and is suitable for nonlinear time-delay systems with arbitrary-order nonlinearity. It also brings convenience to the tuning of PID parameters, providing a new idea for the practical design of nonlinear time-delay systems.