This letter focuses on the design and control of a novel seven-degree-of-freedom (7-DOF) robotic manipulator (D-Arm) to address the issue of backlash nonlinearity coupling unknown disturbance through the dual-motor anti-backlash control technology. Specifically, the first three axes of the D-Arm near the base are implemented as dual-motor joints (DMJs), while the remaining four axes are single-motor joints (SMJs), which achieve a more comprehensive performance. For DMJs, as over-actuated systems, we first discover an internal disturbance phenomenon named servo-conflict and consider it in the controller design. To mitigate the adverse effects of backlash coupling disturbance, an admittance control-based position compensator is proposed. Then, after the backlash elimination, a dual-motor linear active disturbance rejection controller is effectively developed for load tracking task of the DMJ. In the presence of unknown backlash and disturbance, the proposed strategy can improve both transient and steady-state position tracking response, reduce energy consumption without requiring any backlash model information. The effectiveness and simplicity of the developed control strategy are verified through comparative experiments on the D-Arm.
Considering system uncertainties (e.g., gear backlash, unmodeled dynamics, nonlinear friction and parameters perturbation) coupling disturbances weaken the motion performance of robotic systems, an observer-based finite-time prescribed performance sliding mode control with faster reaching law is proposed for robotic manipulators equipped with dual-motor joints (DMJs). In the case where the backlash information is completely unknown, the backlash is maximally eliminated using a simple but efficient dual-motor adaptive anti-backlash strategy. Thus, the design of position tracking controllers for DMJs can be simplified. Then, to deal with the influence of disturbances and residual uncertainties (excluding backlash), a novel finite-time adaptive sliding mode disturbance observer (ASMDO) is proposed to practically estimate the lumped uncertainties where their upper bounds are assumed to be unknown. Finally, a finite-time composite fast non-singular terminal sliding mode (TSM) controller, integrated with the prescribed performance principle, is proposed in this paper. To enhance the convergence rate, a novel TSM-type reaching law has been developed. The controller ensures that the tracking error is not only stabilized within a finite-time convergence rate but also adheres to a predefined maximum transient-steady-state error. The proposed scheme is implemented through simulation and experimental results, demonstrating its superior performance.
This paper presents a new deep visual-inertial odometry and depth estimation framework for improving the accuracy of depth estimation and ego-motion from image sequences and inertial measurement unit (IMU) raw data. The proposed framework predicts ego-motion and depth with absolute scale in a self-supervised manner. We first capture dense features and solve the pose by deep visual odometry (DVO), and then combine the pose estimation pipeline with deep inertial odometry (DIO) by the extended Kalman filter (EKF) method to produce the sparse depth and pose with absolute scale. We then join deep visual-inertial odometry (DeepVIO) with depth estimation by using sparse depth and the pose from DeepVIO pipeline to align the scale of the depth prediction with the triangulated point cloud and reduce image reconstruction error. Specifically, we use the strengths of learning-based visual-inertial odometry (VIO) and depth estimation to build an end-to-end self-supervised learning architecture. We evaluated the new framework on the KITTI datasets and compared it to the previous techniques. We show that our approach improves results for ego-motion estimation and achieves comparable results for depth estimation, especially in the detail area.
In the context of autonomous driving, occlusion significantly impacts the accuracy of monocular 3D object detection. To address this issue, a monocular 3D detection occlusion compensation algorithm is proposed to optimize the detection results affected by occlusion. The approach involves integrating an occlusion processing module into the 3D object detection pipeline. This module analyzes and handles occluded vehicles by leveraging 2D object detection results to infer occlusion relation-ships. Subsequently, the occluded local features associated with these relationships are captured by the local feature processing module and then input to the occlusion processing network. The resulting occlusion compensation offset is fed into the 3D object detection branch. Experimental results demonstrate that the method outperforms existing approaches on the KITTI 3D object detection benchmark.
Depth estimation problem is one of the most critical issues for robot SLAM.In this paper, the current unsupervised depth estimation method is mainly focused on single-scale features in CNN convolutional network architecture, and does not make full use of multi-scale features along the network.In addition, the design of the loss function for depth estimation is also adapted to the single scale.In this paper, we also propose a multi-scale unsupervised depth estimation for the above problems, and introduce a matching loss function to adapt to the training of multi-scale networks. Experiments were carried out on the KITTI dataset. The experimental results show that the proposed method improves the accuracy and efficiency of depth estimation.