Search results

The purpose of this paper is to study the controller design of flexible manipulator. Flexible manipulator system is a nonlinear, strong coupling, time-varying system, which is introduced elastodynamics in the study and complicated to control. However, due to the flexible manipulator, system has a significant advantage in response speed, control accuracy and load weight ratio to attract a lot of researchers.

Since the order of flexible manipulator system is high, designing controller process will be complex, and have a large amount of calculation, but this paper will use the dynamic surface control method to solve this problem.

Dynamic surface control method as a controller design method which can effectively solve the problem with the system contains nonlinear and reduced design complexity.

The authors assume that the dynamic parameters of flexible manipulator system are unknown, and use Radial Basis Function neural network to approach the unknown system, combined with the dynamic surface control method to design the controller.

Teaching by demonstration (TbD) is a promising way for robot learning skills in human and robot collaborative hybrid manufacturing lines. Traditionally, TbD systems have only concentrated on how to enable robots to learn movement skills from humans. This paper aims to develop an extended TbD system which can also enable learning stiffness regulation strategies from humans.

Here, the authors propose an extended dynamical motor primitives (DMP) framework to achieve this goal. In addition to the advantages of the traditional ones, the authors’ framework can enable robots to simultaneously learn stiffness and the movement from human demonstrations. Additionally, Gaussian mixture model (GMM) is used to capture the features of movement and of stiffness from multiple demonstrations of the same skill. Human limb surface electromyography (sEMG) signals are estimated to obtain the reference stiffness profiles.

The authors have experimentally demonstrated the effectiveness of the proposed framework. It shows that the authors approach could allow the robot to execute tasks in a variable impedance control mode with the learned movement trajectories and stiffness profiles.

In robot skill acquisition, DMP is widely used to encode robotic behaviors. So far, however, these DMP modes do not provide the ability to properly represent and generalize stiffness profiles. The authors argue that both movement trajectories and stiffness profiles should be considered equally in robot skill learning. The authors’ approach has great potential of applications in the future hybrid manufacturing lines.

Visual simultaneous localization and mapping (SLAM) has limitations such as sensitivity to lighting changes and lower measurement accuracy. The effective fusion of information from multiple modalities to address these limitations has emerged as a key research focus. This study aims to provide a comprehensive review of the development of vision-based SLAM (including visual SLAM) for navigation and pose estimation, with a specific focus on techniques for integrating multiple modalities.

This paper initially introduces the mathematical models and framework development of visual SLAM. Subsequently, this paper presents various methods for improving accuracy in visual SLAM by fusing different spatial and semantic features. This paper also examines the research advancements in vision-based SLAM with respect to multi-sensor fusion in both loosely coupled and tightly coupled approaches. Finally, this paper analyzes the limitations of current vision-based SLAM and provides predictions for future advancements.

The combination of vision-based SLAM and deep learning has significant potential for development. There are advantages and disadvantages to both loosely coupled and tightly coupled approaches in multi-sensor fusion, and the most suitable algorithm should be chosen based on the specific application scenario. In the future, vision-based SLAM is evolving toward better addressing challenges such as resource-limited platforms and long-term mapping.

This review introduces the development of vision-based SLAM and focuses on the advancements in multimodal fusion. It allows readers to quickly understand the progress and current status of research in this field.

This paper aims to provide a comprehensive analysis of the state of the art in energy efficiency for autonomous mobile robots (AMRs), focusing on energy sources, consumption models, energy-efficient locomotion, hardware energy consumption, optimization in path planning and scheduling methods, and to suggest future research directions.

The systematic literature review (SLR) identified 244 papers for analysis. Research articles published from 2010 onwards were searched in databases including Google Scholar, ScienceDirect and Scopus using keywords and search criteria related to energy and power management in various robotic systems.

The review highlights the following key findings: batteries are the primary energy source for AMRs, with advances in battery management systems enhancing efficiency; hybrid models offer superior accuracy and robustness; locomotion contributes over 50% of a mobile robot’s total energy consumption, emphasizing the need for optimized control methods; factors such as the center of mass impact AMR energy consumption; path planning algorithms and scheduling methods are essential for energy optimization, with algorithm choice depending on specific requirements and constraints.

The review concentrates on wheeled robots, excluding walking ones. Future work should improve consumption models, explore optimization methods, examine artificial intelligence/machine learning roles and assess energy efficiency trade-offs.

This paper provides a comprehensive analysis of energy efficiency in AMRs, highlighting the key findings from the SLR and suggests future research directions for further advancements in this field.

The purpose of this paper is to enhance the robot’s ability to complete multi-step contact tasks in unknown or dynamic environments, as well as the generalization ability of the same task in different environments.

This paper proposes a framework that combines learning from demonstration (LfD), behavior tree (BT) and broad learning system (BLS). First, the original dynamic motion primitive is modified to have a better generalization ability for representing motion primitives. Then, a BT based on tasks is constructed, which will select appropriate motion primitives according to the environment state and robot ontology state, and then the BLS will generate specific parameters of the motion primitives based on the state. The weights of the BLS can also be optimized after each successful execution.

The authors carried out the tasks of cleaning the desktop and assembling the shaft hole on Baxter and Elite robots, respectively, and both tasks were successfully completed, which proved the effectiveness of the framework.

This paper proposes a framework that combines LfD, BT and BLS. To the best of the authors’ knowledge, no similar methods were found in other people’s work. Therefore, the authors believe that this work is original.

This paper aims to deal with the force and position tracking problem when a robot performs a task in interaction with an unknown environment and presents a hybrid control strategy based on variable admittance control and fixed-time control.

A hybrid control strategy based on variable admittance control and fixed-time control is presented. Firstly, a variable stiffness admittance model control based on proportional integral and differential (PID) is adopted to maintain the expected force value during the task execution. Secondly, a fixed-time controller based on radial basis function neural network (RBFNN) is introduced to handle the model uncertainties and ensure the fast position tracking convergence of the robot system, while the singularity problem is also avoided by designing the virtual control variable with piecewise function.

Simulation studies conducted on the robot manipulator with two degrees of freedom have verified the superior performance of the proposed control strategy comparing with other methods.

A hybrid control scheme for robot–environment interaction is presented, in which the variable stiffness admittance method is adopted to adjust the interaction force to the desired value, and the RBFNN-based fixed-time position controller without singularity problem is designed to ensure the fast convergence of the robot system with model uncertainty.

The purpose of this paper is to enhance the performance of robots in peg-in-hole assembly tasks, enabling them to swiftly and robustly accomplish the task. It also focuses on the robot’s ability to generalize across assemblies with different hole sizes.

Human behavior in peg-in-hole assembly serves as inspiration, where individuals visually locate the hole firstly and then continuously adjust the peg pose based on force/torque feedback during the insertion process. This paper proposes a novel framework that integrate visual servo and adjustment based on force/torque feedback, the authors use deep neural network (DNN) and image processing techniques to determine the pose of hole, then an incremental learning approach based on a broad learning system (BLS) is used to simulate human learning ability, the number of adjustments required for insertion process is continuously reduced.

The author conducted experiments on visual servo, adjustment based on force/torque feedback, and the proposed framework. Visual servo inferred the pixel position and orientation of the target hole in only about 0.12 s, and the robot achieved peg insertion with 1–3 adjustments based on force/torque feedback. The success rate for peg-in-hole assembly using the proposed framework was 100%. These results proved the effectiveness of the proposed framework.

This paper proposes a framework for peg-in-hole assembly that combines visual servo and adjustment based on force/torque feedback. The assembly tasks are accomplished using DNN, image processing and BLS. To the best of the authors’ knowledge, no similar methods were found in other people’s work. Therefore, the authors believe that this work is original.

This paper aims to provide a precise tracking control scheme for multi-input multi-output “MIMO” nonlinear systems with unknown input time-delay in industrial process.

The predictive control scheme based on multi-dimensional Taylor network (MTN) model is proposed. First, for the unknown input time-delay, the cross-correlation function is used to identify the input time-delay through just the input and output data. And then, the scheme of predictive control is designed based on the MTN model. It goes as follows: a recursive d-step-ahead MTN predictive model is developed to compensate the influence of time-delay, and the extended Kalman filter (EKF) algorithm is applied for its learning; the multistep predictive objective function is designed, and the optimal controlled output is determined by iterative refinement; and the convergence of MTN predictive model and the stability of closed-loop system are proved.

Simulation results show that the proposed scheme is of desirable generality and capable of performing the tracking control for MIMO nonlinear systems with unknown input time-delay in industrial process effectively, such as the continuous stirred tank reactor (CSTR) process, which provides a considerably improved performance and effectiveness. The proposed scheme promises strong robustness, low complexity and easy implementation.

For the limitations of proposed scheme, the time-invariant time-delay is only considered in time-delay identification and control schemes. And the CSTR process is only introduced to prove that the proposed scheme can adapt to practical industrial scenario.

The originality of the paper is that the proposed MTN control scheme has good tracking performance, which solves the influence of time-delay, coupling and nonlinearity and the real-time performance for MIMO nonlinear systems with unknown input time-delay.