This paper addresses trajectory planning and control for a dual-arm free-floating space robot (FFSR) based on predefinedtime stability. Firstly, a class of general nonsingular terminal sliding mode control strategy is proposed to achieve global predefined-time stability by switching a control parameter in a form of function. Then, a high-precision trajectory planning method is proposed for a FFSR using the proposed predefined-time stability theorem, which can make the end-effector and the base reach the desired pose within a prespecified time, and reduce the adverse effect on planning accuracy resulted from the singularity avoidance algorithm and the obstacle avoidance one. Subsequently, a trajectory tracking control strategy is proposed for a FFSR employing the proposed general nonsingular predefined-time terminal sliding mode, and the tracking errors can converge within a predefined time to an arbitrarily small neighborhood of zero in the presence of persistent disturbances. The proposed trajectory planning method has the advantage of pose feedback, so can be combined with the trajectory tracking controller in the control process to further enhance the control performance for the FFSR. Simulation results validate the effectiveness of the proposed methodologies.

This paper investigates the risk quantification for Unmanned Aircraft (UA) in urban environments, focusing on the safety of ground people. An assessment methodology is proposed to quantify the flying risk, which indicates the ground fatalities resulted from different potential risk sources. With the knowledge of UA’s specifications and ground environments, the flying risk of the target UA flying in the target area could be evaluated from the combination of results from independent assessment procedures focusing on multiple potential risk sources with specific safety metrics. A study case to assess the flying risk of the Talon and the DJI Inspire 2 flying in one piece of the region in Chengdu is presented. From the assessment result, the airspace management strategies for both Talon and DJI Inspires 2 could be easily developed to guarantee the safety of ground people, therefore, this risk quantification method could be a general tool to support decision-making in safety work.

In this paper, aiming at the problem of control and obstacle avoidance in quadrotor formation when mathematical modeling is not accurate, the artificial potential field method with virtual force is used to plan the obstacle avoidance path of quadrotor formation to solve the problem that the artificial potential field method may fall into local optimal. The adaptive predefined-time sliding mode control algorithm based on RBF neural networks enables the quadrotor formation to track the planned trajectory in a predetermined time and also adaptively estimates the unknown interference in the mathematical model of the quadrotor to improve the control performance. Through theoretical derivation and simulation experiments, this study verified that the proposed algorithm can make the planned trajectory of the quadrotor formation avoid obstacles and make the error between the true trajectory and the planned trajectory converge within a predetermined time under the premise of adaptive estimation of unknown interference in the quadrotor model.

This paper addresses a robust trajectory tracking controller for an underactuated quadrotor with external bounded disturbances and unknown inertia parameters. Different from most of the existing control algorithms, the proposed method does not adopt the dual-loop scheme in which the design is divided into position control and attitude control. Instead, command filter backstepping is employed to design the controller based on the integrated motion model such that the stability can be guaranteed strictly for the flight control system. Furthermore, adaptive compensation and robust compensation are introduced to deal with the uncertainty of the inertia parameters and the external bounded disturbances, respectively. Finally, a similar skew symmetric structure is chosen as the desired structure of the closed-loop system to facilitate the analysis of the stability of the integrated system. Stability and robust performance of the designed controller are verified by Lyapunov stability theorem. Simulations are provided to validate the proposed controller.

Gaze is a significant behavioral characteristic that can be used to reflect a person’s attention. In recent years, there has been a growing interest in estimating gaze from facial videos. However, gaze estimation remains a challenging problem due to variations in appearance and head poses. To address this, a framework for 3D gaze estimation using appearance cues is developed in this study. The framework begins with an end-to-end approach to detect facial landmarks. Subsequently, we employ a normalization method and improve the normalization method using orthogonal matrices and conduct comparative experiments to prove that the improved normalization method has a higher accuracy and a lower computational time in gaze estimation. Finally, we introduce a dual-branch convolutional neural network, named FG-Net, which processes the normalized images and extracts eye and face features through two branches. The extracted multi-features are then integrated and input into a fully connected layer to estimate the 3D gaze vectors. To evaluate the performance of our approach, we conduct ten-fold cross-validation experiments on two public datasets, namely MPIIGaze and EyeDiap, achieving remarkable accuracies of 3.11° and 2.75°, respectively. The results demonstrate the high effectiveness of our proposed framework, showcasing its state-of-the-art performance in 3D gaze estimation.

Human Pose Estimation (HPE) plays a critical role in medical applications, particularly within nursing robotics for patient monitoring. Despite its importance, HPE faces several challenges, including high rates of false positives and negatives, stringent real-time requirements, and limited computational resources, especially in complex backgrounds. In response, we introduce the HP-YOLO model, developed using the YOLOv8 framework, to effectively address these issues. We designed an Enhanced Large Separated Kernel Attention (ELSKA) mechanism and integrated it into the backbone network, thereby improving the model’s effective receptive field and feature separation capabilities, which enhances keypoint detection accuracy in challenging environments. Additionally, the Reparameterized Network with Cross-Stage Partial Connections and Efficient Layer Aggregation Network (RepNCSPELAN4) module was incorporated into the detection head, boosting accuracy in detecting small-sized targets through multi-scale convolution and reparameterization techniques while accelerating inference speed. On the COCO dataset, our HP-YOLO model outperformed existing lightweight methods by increasing average precision (AP) by 4.9%, while using 18% fewer parameters and achieving 1.4× higher inference speed. Our method significantly enhances the real-time performance and efficiency of human pose estimation while maintaining high accuracy, offering an optimal solution for applications in complex environments.