27 November 2025 in Robotica
Lower limb-assisted exoskeletons can provide payloads and support, but the hip joints of current lower limb-assisted exoskeletons have problems such as single mechanisms, few degrees of freedom, and limitations in the types of gaits that can be acted upon. To address these problems, a novel double-cam hip joint assist mechanism is proposed for walking, running, and carrying gait situations. The double cam and two rectangular compression springs with different stiffness are used to satisfy the differences in exoskeleton assistance requirements in multiple gaits. First, biomechanical simulation of walking, running, and carrying movements is carried out with OpenSim to obtain the hip joint angle and torque data, and then the structural design of the assist mechanism is carried out. Hip joint angle planning and contour solving are carried out for the contour lines of the cams, so that the cam contour lines are assist according to different hip joint angles, and the stiffness of compression springs is determined by D’Alembert’s principle, and the assist torques are analyzed. Meanwhile, a human–machine coupling model was established for theoretical analysis. Finally, the muscle power change curves were exported using OpenSim, and the assistance was verified by comparison.
1 December 2025 in Physica Scripta
Manual carrying tasks subject the spine to high external loads, resulting in excessive joint torques and an increased risk of injury. Wearable exoskeletons provide an effective way to solve this problem. However, most existing exoskeletons are designed with rigid linkages, which restricts the natural movement of the human body. B ased on the unique anatomical structure of the human spine and dynamic bending mechanisms, this study proposes a novel multi-jointed passive bionic spinal exoskeleton characterized by a five-segment serial elastic unit structure. This design significantly enhances compliance and flexibility in sagittal, coronal, and transverse plane movements, enabling the exoskeleton to better match the natural activities of the human spine. First, we performed a biomechanical simulation of the carrying maneuvers using OpenSim to obtain data on the flexion angles of the major joints. Second, we conducted the structural design of the bionic spine, determined the stiffness of its elastic elements using the static method, and derived its kinematic model using the D-H coordinate analysis method. Concurrently, a human-machine coupled model was established for dynamic analysis. Finally, OpenSim was used to derive the muscle power change curves for a comparative verification of the exoskeleton’s good assisting effect.
19 June 2026 in Physica Scripta
In order to improve the smoothness of walking and human–machine coordination for exoskeleton robots and their wearers, this study proposes an exoskeleton motion intent recognition method based on multi-source information fusion. By integrating surface electromyography (sEMG) signals and inertial measurement unit (IMU) data, a dual-stream convolutional neural network (CNN) combined with the ReliefF algorithm (Two Stream Deep CNN-ReliefF) was constructed to improve the feature extraction model. Additionally, an improved sparrow search algorithm-optimized support vector machine was employed for motion pattern classification. Experimental results demonstrate that the proposed multi-source information fusion method exhibits a significant advantage in motion pattern recognition accuracy compared to traditional feature extraction methods using a single signal source (sEMG or IMU). Specifically, this method achieved a maximum average classification accuracy of 99.56%, which is 7.83%, 0.96%, and 0.54% higher than the traditional feature extraction methods using single sEMG signals, single IMU signals, and sEMG-IMU fusion, respectively. This study provides an efficient motion intent recognition solution for material handling exoskeleton robots, which is of great significance for improving human-machine collaborative operation efficiency and reducing occupational health risks for workers.