- Robot Manipulation and Learning
- Topology Optimization in Engineering
- Robotics and Sensor-Based Localization
- Composite Structure Analysis and Optimization
- Image and Object Detection Techniques
- Robotic Mechanisms and Dynamics
- Advanced Numerical Analysis Techniques
- 3D Surveying and Cultural Heritage
- Advanced Multi-Objective Optimization Algorithms
- Advanced Measurement and Metrology Techniques
- Advanced machining processes and optimization
- 3D Shape Modeling and Analysis
- Robotic Path Planning Algorithms
- Industrial Vision Systems and Defect Detection
- Soft Robotics and Applications
- Advanced Vision and Imaging
- Manufacturing Process and Optimization
- Human Pose and Action Recognition
- Fault Detection and Control Systems
- Image Processing and 3D Reconstruction
- Structural Health Monitoring Techniques
- Dynamics and Control of Mechanical Systems
- Gear and Bearing Dynamics Analysis
- Advanced Machining and Optimization Techniques
- Hydraulic and Pneumatic Systems
Shanghai Jiao Tong University
2015-2024
3D object pose estimation for robotic grasping and manipulation is a crucial task in the manufacturing industry. In cluttered occluded scenes, 6D of low-textured or textureless industrial challenging problem due to lack color information. Thus, point cloud that hardly affected by lighting conditions gaining popularity as an alternative solution estimation. This article proposes deep learning-based using input, which consists instance segmentation The divides scene into multiple clouds, each...
3D object pose estimation for grasping and manipulation is a crucial task in robotic industrial applications. Robustness efficiency are desirable properties that still very challenging complex cluttered scenes, because objects have different appearances, illumination occlusion when seen from viewpoints. This article proposes Semantic Point Pair Feature (PPF) method estimation, which combines the semantic image segmentation using deep learning with voting-based estimation. The Part Mask RCNN...
This article presents a numerical approach of topology optimization with multiple materials for the heat conduction problem. The multiphase level set model is used to implicitly describe geometric boundaries material regions different conductivities. multi-material representation has no emergence intermediate density. objective construct optimal conductive paths which improve efficiency transfer. dissipation thermal transport potential capacity taken as function. sensitivity analysis...
A regional temperature measure model is constructed to obtain a small number of constraints for local transient control. The sensitivity derived using the adjoint variable method. multiple criteria and three-phase topology optimization are further investigated heat conduction design. material layout design replaced by static problem, which subsequently solved method moving asymptotes. Finally, several numerical examples provided demonstrate feasibility validity proposed problems.
Mura is a typical vision defect of LCD panel, appearing as local lightness variation with low contrast and blurry contour. This letter presents new machine inspection way for based on the level set method. First, real Gabor filters are applied to eliminating global textured backgrounds. Then, method employed image segmentation region-based active contours model, which an improvement Chan-Vese's model so that it more suitable Mura. Using some results from segmentation, defects quantified SEMU...
For topology optimization with transient loads, heat compliance varies analysis. The peak value of the should be minimized. Thus, this article proposes a global measure to handle kind for conduction problem. model is then constructed by measure. finite-element, equivalent static and continuum shape based sensitivity analyses are derived using adjoint variable method. Through case studies, effectiveness proposed validated.
Optimization design of the transient heat conduction problem is different from steady-state because varying temperature. This article proposes integral temperature gradient over a fixed time interval as an objective function. The shape sensitivity derived. derivative information function about element densities used strategy to create holes. A narrow band on triangular mesh proposed improve computational efficiency. Numerical examples are presented demonstrate effectiveness approach for...
In this paper, a scale factor-based interpolated discrete Fourier transform (IpDFT) algorithm is proposed to estimate the dominant chatter frequency. practical measurement systems of machining processes, number acquired sine-wave cycles (NASCs) small. Therefore, significant spectral interference from negative frequency leads degradation conventional IpDFTs that neglect such interference. The IpDFT overcomes problem by completely removing long-range leakage investigated component. We...
Abstract This paper presents topology optimization of nonlinear heat conduction problems with multiple domains and constraints, including regional temperature material volume for reducing temperature. Maximum approximation temperatures in the constraint regions are accurately dynamically calculated, though temperature-dependent thermal conductivity change update distribution. A measure constant error to approximate maximum is adaptive different ranges. strategy hole nucleation generation...
Isogeometric analysis (IGA) provides a feasible technique to seamlessly integrate computer-aided design (CAD) into the existing finite element analysis. In this article, Bézier elements-based IGA method is established address stress-related topology optimization of structures, which incorporates geometrical representation, structural and unified process as well can naturally handle curvilinear edge elements for stress problems. To enhance smoothness boundaries, multilevel isogeometric...
Point cloud classification and segmentation are crucial tasks for point processing have wide range of applications, such as autonomous driving robot grasping. Some pioneering methods, including PointNet, VoxNet, DGCNN, etc., made substantial advancements. However, most these methods overlook the geometric relationships between points at large distances from different perspectives within cloud. This oversight constrains feature extraction capabilities consequently limits any further...
The frequency of a real sinusoid is often estimated by interpolated discrete Fourier transform (IpDFT) algorithms because the high accuracy and low computational burden. However, conventional IpDFTs focus on estimation when number acquired sine wave cycles (NASC) more than two. Frequency with small NASC applicable in manufacturing processes, radar level measurement, other applications. spectral leakage negative (the image component) significantly distorts positive small, dramatically...
With the increasing application of high speed machining for pockets, conventional methods tool path generation show their limitations, such as existence a curvature region along path. This paper presents level set approach pockets. Pocket boundary is first embedded into higher dimensional function, namely, function. Then paths are realised by propagation pocket boundary. The topological changes can be naturally handled during evolution moving fronts. In order to avoid appearance corners...
Abstract This article presents the Bézier extraction based isogeometric approach to multi‐objective topology optimization of periodic microstructures. The utilizes B‐splines elements as finite element representation at both macro and micro levels. equivalent elastic properties representative volume (RVE) are computed by numerical homogenization method with boundary conditions on control points (CPs) B‐spline mesh. problems including material bulk modulus maximization, negative Poisson's...
External force estimation for industrial robots can be applied to the scenes such as human–robot interaction and robot machining. The model-based methods have gained attention of many researchers because they only need motor signals. However, performance above-mentioned depends on accuracy inverse dynamic model (IDM). Traditional focus improving IDM obtain a better performance. suppressing negative impact modelling error achieved without reducing error. Thus, this article proposes an...
The weighted wavelet packet entropy (WWPE) is presented for real-time detection of chatter before it has fully developed in turning processes. This method based on low dimensionality process dynamics shown by chatter. energy frequency band that contains information weighted. Compared with the WPE, WWPE more sensitive to variation caused onset Therefore, can be detected an earlier stage. From computational point view, matrix employed instead Mallat algorithm, which computationally expensive...