Performance degradation will be caused by a variety of interfering factors for pattern recognition-based myoelectric control methods in the long term. This paper proposes an adaptive learning method with low computational cost to mitigate the effect in unsupervised adaptive learning scenarios. We presents a particle adaptive classifier (PAC), by constructing a particle adaptive learning strategy and universal incremental least square support vector classifier (LS-SVC). We compared PAC performance with incremental support vector classifier (ISVC) and non-adapting SVC (NSVC) in a long-term pattern recognition task in both unsupervised and supervised adaptive learning scenarios. Retraining time cost and recognition accuracy were compared by validating the classification performance on both simulated and realistic long-term EMG data. The classification results of realistic long-term EMG data showed that the PAC significantly decreased the performance degradation in unsupervised adaptive learning scenarios compared with NSVC (9.03% ± 2.23%, p < 0.05) and ISVC (13.38% ± 2.62%, p = 0.001), and reduced the retraining time cost compared with ISVC (2 ms per updating cycle vs. 50 ms per updating cycle).

Soft robot hands have the advantage of remarkable adaptability for grasping. Especially for the soft and fragile objects, soft fingers had presented their much excellent potential compared with their rigid counterparts. However, less degree of freedom, lower force output, lack of proprioception, and poor controllability still limit the application. Inspired by the anatomical structure of the human hand and following the idea of combining soft joints, rigid skeletons and embedded soft curvature sensors, modular dexterous hands composed of multijoint fingers are proposed in this study. Each finger has three quasi-joints, in which metacarpophalangeal soft-joint can realize adduction/abduction and bending motions, and distal two interphalangeal soft-joints are actuated by one actuator. Similar to human hand, soft-joint so-called quasi-joint has a short length of constant curvature segment. The integrated Indium Gallium Alloy sensors with Kelvin Bridge for proprioception can accurately detect joint angles, while closed-loop control based on proprioception was accomplished. Kinematics and statics modeling method of the rigid-soft finger is proposed. To further verify the performance of this design, prototypes of three-fingered and five-fingered hands are developed. The multifingered hands had demonstrated their capability of adaptive grasp and dexterous manipulation, while the force output of the three-fingered hand is up to 31.82 N, and 32 grasp types had accomplished by the five-fingered hand.

Elastic elements are commonly adopted to realize underactuation in the design of human-friendly prosthetic hands. The stiffness of these elastic elements, which is a key factor affecting the grasp performance of the underactuated finger, has not well addressed when considering both the stability and adaptability. In this study, an adaptive anthropomorphic finger that adopted a novel nine-bar mechanism is proposed. This nine-bar mechanism is integrated through a coupled four-bar mechanism and an adaptive seven-bar mechanism. The developed finger based on the nine-bar mechanism is able to improve the grasp stability in the global workspace under an extremely small spring stiffness. A quantitative analysis of the grasp stability was carried out. Comparative experiments on the grasps using the finger with/without adaptability were also performed. The results validated that our finger has a good stability when grasping the objects of different sizes.

Purpose: This paper aims to present the design and experiment of a modular multisensory prosthetic hand for applications. Design and experiment of a modular multisensory hand for prosthetic applications. Design/methodology/approach: This paper reveals more details focusing on the appearance, mechanism design, electrical design and control of the prosthetic hand considering anthropomorphism, dexterity, sensing and controllability. The finger is internally integrated with the actuator, the transmission mechanism, the sensors and the controller as a modular unit. Integrated with multiple sensors, the prosthetic hand can not only perceive the position, the contact force and the temperature of the environment like a human hand but also provide the foundation for the practical control. Findings: The experiments show that the prosthetic hand can accurately control the contact force to achieve stable grasps based on the sensors feedback and a simple and effective force-tracking impedance control algorithm. In addition, the experiments based on the cosmesis validate not only the cosmesis functionality but also the control performance for a prosthesis–cosmesis system. Practical implications: Because of the small size, low weight, high integration, modularity and controllability, the prosthetic hand is easily applied to upper-limb amputees. Meanwhile, the finger as a modular unit is easy to be fixed, maintained and applied to a partial upper-limb amputee. Originality/value: Each modular finger of the prosthetic hand integrated with the actuator, the transmission mechanism, the sensors and the controller as a whole can independently control the position and the force. The cosmetic glove design can provide pretty appearance without compromising the control performance.

The robotic breast ultrasound scanning (RBUS) system can help sonographer in the early screening of breast cancer. However, it still faces rigorous safety concerns, including non-smooth scanning paths and improper contact force regulation. In this letter, we propose a new path optimization method to improve the scanning smoothness, as well as, a dynamic contact regulation strategy considering both the breast deformation and ultrasound images. For the path planning, the anti-radial scanning approach is first adopted to generate some initial path points together with their normal vectors; then, a dual-objective optimization function is formulated, which considers both the acceleration continuity and deviation of the path, to filter out local path anomalies. For the force control, a contact force-strain regression model is first proposed and used to predict the desired force between the breast and probe at each path point. While within the adjacent path points, an updating algorithm is introduced to dynamically adjust the contact force in real-time by surveying the confidence of the ultrasound image. Our experimental results show that the proposed methods can make the scanning path smoother, and obtain intimate probe-tissue contact with less desired force.