It is very challenging for robots to perform grinding and polishing tasks on surfaces with unknown geometry. Most existing methods solve this problem by modeling the relationship between the force sensing information and surface normal vectors by analyzing the forces on special end tools such as spherical tools and cylindrical tools and simplified friction model. In this paper, we propose a normal vectors learning method to simultaneously control end-effector force and direction on unknown surfaces. First, the relation that mapping the force sensing information to the surface normal vectors is learned from the demonstrated data on the known plane using locally weighted regression. Next, the learned relation is used to estimate surface normal vectors on the unknown surface. To improve the force control precision on the unknown geometry surface, the adaptive force control is developed. To improve the direction control precision due to friction, the iterative learning control is developed. The proposed method is verified by comparative simulations and experiments using the Franka robot. Results show that the end-effector can be controlled perpendicular to the surface with a certain force.

This article focuses on a multibody model of a new passive articulated suspension tracked robot. The model of the vehicle is developed using the multibody software MSC Adams (acronym for automatic dynamic analysis of mechanical systems) in conjunction with the Adams Tracked Vehicle toolkit. The various subsystems that make up the vehicle assembly are described, with particular attention to the modeling strategies of the swing arms that constitute the suspension system and the characterization of the track belt. The multibody model allows the system performance to be evaluated in advance for fine-tuning of the design parameters and using different configurations, for example, with active and locked suspension. Efforts are also presented to validate the multibody twin against the real prototype, showing a good agreement with field experiments. Once validated, the digital twin is used to assess the performance of the innovative suspension system in various challenging environments.

Recently, deep learning methods have achieved considerable performance in gesture recognition using surface electromyography signals. However, improving the recognition accuracy in multi-subject gesture recognition remains a challenging problem. In this study, we aimed to improve recognition performance by adding subject-specific prior knowledge to provide guidance for multi-subject gesture recognition. We proposed a time–frequency feature transform suite (TFFT) that takes the maps generated by continuous wavelet transform (CWT) as input. The TFFT can be connected to a neural network to obtain an end-to-end architecture. Thus, we integrated the suite into traditional neural networks, such as convolutional neural networks and long short-term memory, to adjust the intermediate features. The results of comparative experiments showed that the deep learning models with the TFFT suite based on CWT improved the recognition performance of the original architectures without the TFFT suite in gesture recognition tasks. Our proposed TFFT suite has promising applications in multi-subject gesture recognition and prosthetic control.

A method is presented for configuration selection to obtain the best tip-over stability of a modular reconfigurable mobile manipulator (MRMM) under various application situations. The said MRMM consists of a modular reconfigurable robot (MRR) mounted on a mobile platform. The MRR in different configurations creates different wrenches onto the mobile platform, leading to different tip-over moments of the MRMM, even though the joint speeds or tip speeds remain the same. The underlying problem pertains to selecting one configuration of MRR for reconfiguration that would obtain the best tip-over stability under a given application. First, all the permissible configurations are identified through an enumeration method. Then, the feasible configurations are determined based on application-oriented workspace classifications. At last, two workspace indices, vertical reach and horizontal reach, are used to select an optimal configuration. The tip-over stability analysis and evaluation of MRMM are carried out for verification for three cases including vertical, horizontal, and general 3D space applications. The results demonstrate the effectiveness of the proposed method.

The given article emphasizes the development and modeling of a hybrid navigational controller to optimize the path length and time taken. The proposed navigational controller is developed by hybridizing the metaheuristic moth–flame optimization (MFO) approach and the reinforcement learning (RL) approach. Input parameters like obstacle and target locations are fed to the MFO controller that implements a proper navigational direction selection. It forwards to the RL controller, which exercises further refinement of the output turning angle around obstacles. The collaboration of the global MFO approach with the local-based RL approach helps to optimize the path traversed by the humanoid robot in an unknown environment. The major breakthrough in this article is the utilization of humanoid robots for navigation purposes between various checkpoints. The humanoid robots are placed in a cluttered environment and assigned specific target positions to complete the assigned tasks. In the case of a multi-humanoid robot system, to avoid self-collision, it requires a Petri-Net controller to be configured in the navigation system to prevent deadlock situations and enhance the smooth completion of tasks without inter-collision among the humanoid robots. Simulations and real-time experiments are undertaken using different controllers involving single- and multi-humanoid robot systems. The robustness of the proposed controller is also validated in dynamic environment. Comparisons are carried with an established navigational controller in a similar environmental setup, which proves the proposed hybrid controller to be robust and efficient.

The axial offset joint has two rotating axes that do not intersect but have a specific offset in space. It is used widely in parallel manipulators (PMs). The offset-joint workspace can directly affect the PM workspace. This study performed a theoretical derivation and workspace analysis of a class of axial offset joints. First, a theoretical parametric model describing the rotation range of the offset joint is established that considers the interference of the offset joint because of the contact between the upper- and lower-joint brackets during movement. Second, the analytical expressions of the offset-joint workspace are formulated based on the coordinate system transformation. The offset-joint workspace is theoretically calculated in this study using formulations. Then, through a comparative analysis, the superiority of the offset joint compared with the universal joint is verified. The theoretical formulations in this paper can be used to calculate the workspace of a class of axial offset joints. Finally, based on a workspace analysis of three types of PMs using offset, universal, and spherical joints, the offset-joint PM workspace is much larger than those of the other two types.

Biped robots with dynamic motion control have shown strong robustness in complex environments. However, many motion planning methods rely on models, which have difficulty dynamically modifying the walking cycle, height, and other gait parameters to cope with environmental changes. In this study, a heuristic model-free gait template planning method with dynamic motion control is proposed. The gait trajectory can be generated by inputting the desired speed, walking cycle, and support height without a model. Then, the stable walking of the biped robot can be realized by foothold adjustment and whole-body dynamics model control. The gait template can be changed in real time to achieve gait flexibility of the biped robot. Finally, the effectiveness of the method is verified by simulations and experiments of the biped robot BHR-B2. The research presented here helps improve the gait transition ability of biped robots in dynamic locomotion.

In the control of space robots, flexible vibrations exist in the base, links and joints. When building a motion control scheme, the following three aspects should be considered: (1) the complexity in dynamic modeling; (2) the low accuracy of motion control and (3) the simultaneous suppression of multiple flexible vibrations. In this paper, we propose a motion vibration integrated saturation control scheme. First, the dynamic model of space robot with flexible-base, flexible-link and flexible-joint is established according to the assumed modes method and Lagrange equation. Second, singular perturbation theory is used to decompose the model into two subsystems: a slow subsystem containing the rigid motions of base and joints as well as the vibration of links, and a fast subsystem containing vibrations of base and joints. Third, an integrated sliding mode control with input restriction, output feedback and repetitive learning (ISMC-IOR) is designed, which can track the desired trajectories of base and joints with −3 orders of magnitude accuracy, while suppressing the multiple flexible vibrations of base, links and joints 50%–80% and 37% performance improvement over ISMC-IOR-NV were achieved. Finally, the algorithm is verified by simulations.

The repetitive motion planning movements of the redundant manipulator will cause oscillations and unintended swings of joints, which increase the risk of collisions between the manipulator and its surroundings. Motivated by this phenomenon, this paper presents an inverse kinematics algorithm for the spherical-revolute-spherical manipulator to solve the paradox raised by joint-drift and control the pose with no swing of the elbow. This algorithm takes the joint Cartesian positions set as the intermediary and divides the inverse solution process into two mapping processes within joint limits. Simulations are executed to evaluate this algorithm, and the results show this algorithm is applicable to repetitive motion planning and is capable of producing superior configurations based on its real-time ability and stable solve rate. Experiments using the 7-degree-of-freedom spherical-revolute-spherical manipulator demonstrate the effectiveness of this algorithm to remedy the joint-drift and elbow swing compared to Kinematics and Dynamics Library and TRAC-IK.