Yu Lin

The robotic riveting system requires a rivet robotic positioning process for rivet-in-hole insertions, which can be divided into two stages: rivet path-following and rivet spot-positioning. For the first stage, varying parameter-linear sliding surfaces are proposed to achieve robust rivet path-following against robot errors and external disturbances of the robotic riveting system. For the second stage, a second-order sliding surface is applied to attain accurate rivet spot-positioning within a finite time required by the riveting process. In order to improve the dynamic performance of the robot riveting system, the motion of robot end-effector between the two adjacent riveting spots has been properly designed. Overall, the proposed control scheme can guarantee not only the stability of the robot control system but also the robust rivet path-following and quick rivet spot-positioning in the presence of the robot errors and external disturbances of the robotic riveting system. The simulation and experimental results demonstrate the effectiveness of the proposed control scheme.

This paper addresses the problem of path tracking of industrial robots. The main idea is to correct a preplanned path through an iterative learning control (ILC) method. Instead of seeking the conventional ILC strategy, an iterative learning identification method, which is called calibration-based ILC, is developed to identify the robot kinematic parameters along the path in a local working zone. To facilitate calibration-based ILC, we propose two objectives. The first objective is to find the exact values of robot kinematic parameters based on the ILC scheme. The second objective is to search the fastest learning convergence speed and robustness in the iterative domain. Based on the identification of robot kinematic parameters, we then propose an algorithm for the accurate path tracking of industrial robots. The simulation and experimental results demonstrate that the performance of path tracking can be improved significantly via the proposed method.

Industrial robot calibration packages, such as ABB CalibWare, are merely developed only for robot calibration. As a result, the robotic tooling systems designed and fabricated by the user are often calibrated in an ad hoc fashion. In this paper, a systematic approach for robotic tooling calibration is presented in order to overcome this problem. The idea is to include the tooling system as an extended body in the robot kinematic model, from which two error models are established. The first model is associated with the robot, while the second model is associated with the tooling. Once the robot is fully calibrated, the first error will be eliminated. Thus, our method is focused on the second error model. For this calibration, a self-calibration method is developed by using a calibration plate with multiple holes. Then, the tooling calibration model is formulated against the distance between the two holes. For measurements of the distances, a camera is mounted on the tooling system. To present the error mapping, a virtual kinematic link is proposed, which directly connects the camera to the tooling system. For error identification, a linear computational method is utilized to determine the tooling calibration parameters. The linear computational method is based on the Taylor expansion of the robotic tooling model. Once the error is identified, the next step is to implement the error in the tooling kinematic model. A case study is provided to demonstrate the effectiveness of the proposed method.

In this paper, we propose a novel outlier diagnosis method for robust pose estimation of rigid body motions from outlier contaminated 3D point measurements. Due to incorrect correspondences in a cluttered measuring environment, observed point data are contaminated by outliers, which are unusual gross errors that lie out of an overall error distribution. Standard least-squares methods for pose estimation are highly sensitive to outliers. For this reason, an outlier diagnosis method is developed to preprocess measured point data prior to pose estimation. This diagnosis method detects and removes outliers based on a relaxation method with rigid body constraints of a rigid body. Simulations and experiments prove the effectiveness and advantages of high breakdown point and ease of implementation.

Developed in this paper is a hybrid method for calibration of modular reconfigurable robots (MRRs). The underlying problem under study is unique to MRRs, that is, how to calibrate a set of MRR’s geometric parameters that are applicable to all feasible configurations. For this reason, a hybrid search method is developed to ensure a global search over the MRRs’ workspace for each feasible configuration. By combining a genetic algorithm method with a Monte Carlo method, this method includes three levels of search, namely, pose, workspace, and configuration-space. The final set of global solutions is generated progressively from the results of these three levels of search. The effectiveness of this method is demonstrated through a case study.

In this study, a data decorrelation method is developed for processing 3D position measurements obtained by a line sensors based photogrammetric measuring system. The measured 3D data suffer from correlations due to the principle of the triangulation. This nature causes complexity when a calibration procedure is carried out. Additionally, the correlated data may affect its further exploration and applications, such as robot visual servoing. For this reason, a data decorrelation method based on multivariate statistical analysis is proposed to preprocess the 3D coordinate measurements, in such a way that each component of the coordinate can be treated individually. Consequently, the calibration procedure can be significantly simplified in a less time and memory consuming manner. The proposed method includes the local and global data decorrelation. Evaluation and comparisons of these two approaches have been analyzed with a conclusion on the better performance of the global data decorrelation. The optimal decorrelated configuration based on the results has been identified for the measuring system.