发明人：翟弟华，刘浩男，林俊丞，熊雨涵，张思华，王永康，夏元清，詹玉峰

申请日期：2023.09.22

摘要：本发明公开了一种基于控制障碍函数的机器人接触力约束控制方法，基于建立的机器人的运动学模型、动力学模型与接触力模型构建包含接触力的仿射非线性系统模型，相比于考虑单一的接触力模型具有更好的动态性能，通过将控制障碍函数应用于机器人接触力安全约束控制，与现有的位置安全约束控制方法相比拓宽了控制障碍函数在机器 ...

发明人：翟弟华，王永康，夏元清，李晓鸽，詹玉峰

申请日期：2023.09.22

摘要：本发明公开了一种可防御拜占庭攻击的联邦学习图像分类方法，利用各本地图像分类模型在虚拟数据集上的特征输出，并基于Median和MAD处理Gram矩阵以剔除被拜占庭攻击的恶意模型，再选择正常Worker节点提交的节点训练梯度参与聚合，完全避免了拜占庭攻击对全局模型的影响，并使其在正常图像数据中具有较高的 ...

作者：Xiaohui Hao¹;Yuanqing Xia²;CA;Hongjiu Yang³;Zhiqiang Zuo³; （¹a School of Automation, Beijing Institute of Technology, Beijing, People\'s Republic of China;b Yangtze Delta Region Academy of Beijing Institute of Technology, Jiaxing, People\'s Republic of China;²a School of Automation, Beijing Institute of Technology, Beijing, People\'s Republic of China;³c School of Electrical and Information Engineering, Tianjin University, Tianjin, People\'s Republic of China）

出处：International Journal of Systems Science

关键词：Route-based vehicle tracking;typical motion modelling;variable structure multiple models (VSMMs)

摘要：In this paper, a typical motion-based modelling and tracking issue is investigated for vehicle targets in linear road segment by a variable structure ...

作者：Dongyu Han¹;Kun Liu¹;Yeming Lin¹;Yuanqing Xia¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China）

出处：IEEE Transactions on Automatic Control 2024

关键词：Cost function;Vectors;Upper bound;Radio frequency;Convex functions;Sparse matrices;Robots

摘要：This paper considers a distributed stochastic non-convex optimization problem, where the nodes in a network cooperatively minimize a sum of $L$ -smoot ...

作者：Chen Yang¹;Wanze Lu¹;Yuanqing Xia¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China）

出处：IEEE Transactions on Reliability 2024

关键词：Reliability;Robots;Uncertainty;Kinematics;Industrial robots;Service robots;Reliability theory

摘要：A novel method for positioning accuracy analysis based on non-probabilistic time-dependent reliability is proposed for industrial robots with multisou ...

作者：Han Gao¹;Jiale Wang¹;Xuelin Liu¹;Yuanqing Xia¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China）

出处：IEEE Transactions on Fuzzy Systems 2024

关键词：Fuzzy logic;Observers;Vectors;Uncertainty;Convergence;Consensus control;Nonlinear dynamical systems

摘要：This paper aims to address the issue of fixed-time consensus control for uncertain nonlinear multiagent systems (MASs), in which only a group of follo ...

作者：Gaochang Zhang¹;Senchun Chai¹;Runqi Chai¹;Marcos Garcia¹;Yuanqing Xia¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China）

出处：IEEE Transactions on Intelligent Vehicles 2024

关键词：Autonomous vehicles;Programming;Optimization;Trajectory planning;Trajectory optimization;Wheels;Planning;Autonomous Vehicles;Goal Programming;Fuzzy Goal;Fuzzy Goal Programming;Simulation Results;Optimization Problem;Path Length;Multi-objective Optimization;Trajectory Optimization;Fuzzy Algorithm;Single-objective Optimization;Compromise Solution;Weighted Sum Method;Passenger Comfort;Objective Function;Expected Value;Additive Model;Less Than Or Equal;Greater Than Or Equal;Completion Time;Single-objective Optimization Problem;Trajectory Planning;Tolerable Upper Limit;Objective Function Value;Fuzzy Set Theory;Physical Experiments;Multi-objective Programming;Tolerance Factor;Multiple Objects;Multi-objective Optimization Problem

摘要：With the development of autonomous vehicle technology, planning an efficient trajectory for automatic parking while considering multiple factors such ...

作者：Sheng Yu¹;Di-Hua Zhai^1,2;Yuanqing Xia¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China ²Yangtze Delta Region Academy of Beijing Institute of Technology, Jiaxing, China）

出处：IEEE/ASME Transactions on Mechatronics 2024

关键词：Pose estimation;Point cloud compression;Robots;Solid modeling;Deformation;Feature extraction;Task analysis;Pose Estimation;Self-supervised Learning;Human Pose Estimation;Object Pose;Robotic Grasping;Category-level Object;Category-level Object Pose;Large Datasets;Point Cloud;RGB Images;Target Object;Series Of Methods;Objects In The Scene;Unknown Objects;Pose Estimation Methods;Object Point Cloud;Local Features;Network Performance;Global Features;Bounding Box;Pose Tracking;Point Cloud Features;Feature Fusion;Latent Vector;Self-supervised Training;RGB-D Images;Depth Images;Final Pose;Object Size;Synthetic Images

摘要：6-D object pose estimation is widely used in the robotic grasp, and a series of object pose estimation methods have been proposed. Among them, categor ...

作者：Liwen Yang¹;Yuanqing Xia¹;Xiaopu Zhang¹;Lingjuan Ye²;Yufeng Zhan¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China ²School of Mathematics and Statistics, Beijing Institute of Technology, Beijing, China）

出处：IEEE Transactions on Automation Science and Engineering 2024

关键词：Task analysis;Costs;Distribution strategy;Cloud computing;Scheduling algorithms;Dynamic scheduling;Schedules;Neural Network;Optimization Problem;Resource Management;Economic Costs;Neural Network Classifier;Transmission Time;Virtual Machines;Graph Neural Networks;Scheduling Scheme;Scheduling Algorithm;Execution Cost;Scheduling Framework;Nonlinear Function;Resource Constraints;Epigenome;Types Of Tasks;First Category;Time Task;Task Scheduling;Makespan;Task Execution Time;Message Passing;Scheduling Results;Descending Order;Heuristic Algorithm;Graph Topology;Task Prioritization;Rental Prices

摘要：Cloud workflow scheduling is a typical combinatorial optimization problem and becomes more challenging due to the increasing diversity of workflows. H ...

作者：Min Lin¹;Zhongqi Sun^1,2;Yuanqing Xia¹;Jinhui Zhang¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China ²Yangtze Delta Region Academy of Beijing Institute of Technology, Jiaxing, China）

出处：IEEE Transactions on Neural Networks and Learning Systems 2024

关键词：Costs;Predictive control;Discrete-time systems;Stability criteria;Convergence;Task analysis;Sun;Predictive Control;Model Predictive Control;Value Function;Linear System;Computational Burden;Prediction Horizon;Model Predictive Control Strategy;Terminal Cost;Terminal Constraint;Training Data;Optimal Control;State Space;Stability Analysis;Nonlinear Systems;Control Input;Control Performance;Weight Vector;Optimal Policy;Termination Condition;Running Costs;Temporal Difference Learning;Optimal Value Function;Reinforcement Learning Agent;Input Constraints;Theoretical Guarantees;Iterative Sampling;Value Function Approximation;Update Law;Nonholonomic;Policy Evaluation

摘要：This article proposes a novel reinforcement learning-based model predictive control (RLMPC) scheme for discrete-time systems. The scheme integrates mo ...