检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Deng, Boyuan Sun, Jian Li, Zhuo Wang, Gang School of Automation Beijing Institute of Technology Beijing100081 China Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing100081 China

Continuous trajectory tracking control of quadrotors is complicated when considering noise from the environment. Due to the difficulty in modeling the environmental dynamics, tracking methodologies based on conventional control theory, such as model predictive control, have limitations on tracking accuracy and response time. We propose a Time-attenuating Twin Delayed DDPG, a model-free algorithm that is robust to noise, to better handle the trajectory tracking task. A deep reinforcement learning framework is constructed, where a time decay strategy is designed to avoid trapping into local optima. The experimental results show that the tracking error is significantly small, and the operation time is one-tenth of that of a traditional algorithm. The OpenAI Mujoco tool is used to verify the proposed algorithm, and the simulation results show that, the proposed method can significantly improve the training efficiency and effectively improve the accuracy and convergence stability. Copyright © 2023, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Ensemble of Extreme Learning Machines for Regression

Ensemble of Extreme Learning Machines for Regression

引用

Data Driven control and Learning systems (DDCLS)

作者： Atmane Khellal Hongbin Ma Qing Fei School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing Institute of Technology China

ISBN: (纸本)9781538626191

Regression, as a particular task of machine learning, performs a vital part in data-driven modeling, by finding the connections between the system state variables without any explicit knowledge about the system, using a collection of input-output data. To enhance the prediction performance and maximize the training speed, we propose a fully learnable ensemble of Extreme Learning Machines (ELMs) for regression. The developed approach learns the combination of different individual models, using the ELM algorithm, which is applied to minimize both the prediction error and the norm of the network parameters, which leads to higher generalization performance under Bartlett's theory. Moreover, the average based ELM ensemble may be viewed as a particular case of our model. Extensive experiments on many standard regression benchmark datasets have been carried out, and comparison with different models has been performed. The experimental findings confirm that the proposed ensemble can reach competitive results in term of the generalization performance, and the training speed. Furthermore, the influence of different hyper parameters on the performance, in term of the prediction error and the training speed, of the developed model has been investigated to provide a meaningful guideline to practical applications.

关键词： Computational modeling Training Mathematical model Predictive models Neural networks Prediction algorithms Approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

An Introduction to Domain Adaptive Object Detection from Synthesis to Reality

An Introduction to Domain Adaptive Object Detection from Syn...

引用

第40届中国控制会议

作者： Zhijun Xue Wenjie Chen Jing Li State Key Laboratory of Intelligent Control and Decision of Complex Systems School of AutomationBeijing Institute of Technology Beijing Institute of Technology Chongqing Innovation Center

In recent years,deep learning based object detection has achieved great *** methods typically assume that large amount of labeled training data is available,meanwhile,training and test data are independent and identically ***,the two assumptions are not always hold in *** many applications,it is prohibitively expensive and time-consuming to obtain large quantities of labeled *** computer graphics technology to generate a large number of labeled data provides a solution to this ***,direct transfer across domains from synthesis to reality often performs poorly due to the presence of domain *** adaptive object detection are concerned with accounting for these types of *** this paper,we present an introduction to these ***,we briefly introduce the object detection and domain ***,the synthetic object detection datasets and related software tools are ***,we present a categorization of approaches,divided into discrepancy-based methods,adversarial discriminative methods,reconstruction-based methods and ***,we also discuss some potential deficiencies of current methods and several open problems which can be explored in future work.

关键词： Deep learning Object detection Synthesis to reality Domain adaptation

来源：评论

学校读者我要写书评

暂无评论

Action Recognition Algorithm based on 2D Human Pose Estimation Method

Action Recognition Algorithm based on 2D Human Pose Estimati...

引用

第40届中国控制会议

作者： Chongkai Yu Wenjie Chen Ye Li Chen Chen State Key Laboratory of Intelligent Control and Decision of Complex Systems School of Automation Beijing Institute of Technology Beijing Institute of Technology Chongqing Innovation Center

This paper presents an action recognition method based on 2D human body node data in video. This method uses the pose estimation algorithm to detect the human body node data in each frame of video information. We get the two-dimensional coordinates and confidence data of the nodes, and optimize the arrangement of these data into a 3D array form similar to ***, we use the classical two-dimensional convolutional neural network to carry out classification training. The test on UCF-101 data set shows that this method can indeed improve the accuracy of action recognition based on RGB information to a certain extent, and reduce the training cost.

关键词： Action recognition Pose estimation Two-dimensional convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

On Finite-Horizon Optimal control of First-Order Plus Time Delay systems

On Finite-Horizon Optimal Control of First-Order Plus Time D...

引用

第27届中国控制与决策会议

作者： Dawei Shi Tongwen Chen State Key Laboratory of Intelligent Control and Decision of Complex Systems School of AutomationBeijing Institute of Technology Department of Electrical and Computer Engineering University of Alberta

In this work,a finite-horizon optimal control problem for first-order plus time delay(FOPTD) processes is *** show that if the control horizon is greater than three and the prediction horizon is great than the control horizon plus the time delay in discrete time,the optimal controller is not affected by either of the two ***,under these conditions,the controller parameters are explicitly calculated,the closed-loop system is shown to be stable,and the controller is *** problem considered is related to the results on linear quadratic regulation of linear systems with time delays;however,the detailed parameterization of the state-space model introduced by the FOPTD process provides an additional opportunity to investigate the exact controller structure and properties(e.g.,the locations of the closed-loop poles),which are also the major difficulties encountered and overcome in this *** problem is motivated from phenomena experienced in designing industrial model predictive control(MPC) tuning algorithms,and extensive numerical examples indicate that the proposed results speed up the MPC autotuning algorithms by 70%.

关键词： Optimal control First-order plus time delay process Performance visualization MPC tuning

来源：评论

学校读者我要写书评

暂无评论

A Unified Framework Integrating decision Making and Trajectory Planning Based on Spatio-Temporal Voxels for Highway Autonomous Driving

引用

IEEE Transactions on intelligent Transportation systems 2022年第8期23卷 10365-10379页

作者： Zhang, Ting Song, Wenjie Fu, Mengyin Yang, Yi Tian, Xiaohui Wang, Meiling Beijing Institute of Technology State Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing100081 China Nanjing University of Science and Technology School of Automation Nanjing210014 China

intelligent decision making and efficient trajectory planning are closely related in autonomous driving technology, especially in highway environment full of dynamic interactive traffic participants. This work integrates them into a unified hierarchical framework with long-term behavior planning (LTBP) and short-term dynamic planning (STDP) running in two parallel threads with different horizon, consequently forming a closed-loop maneuver and trajectory planning system that can react to the dynamic environment effectively and efficiently. In LTBP, a novel voxel structure and the 'voxel expansion' algorithm are proposed for the generation of driving corridors in 3D configuration, which involves the prediction states of surrounding vehicles. By using Dijkstra search, the maneuver with minimal cost is determined in form of voxel sequences, then a quadratic programming (QP) problem is constructed for solving the optimal trajectory. And in STDP, another small-scaled QP problem is performed to track or adjust the reference trajectory from LTBP in response to the dynamic obstacles. Meanwhile, a Responsibility-Sensitive Safety (RSS) Checker keeps running at high frequency for real-time feedback to ensure security. Experiments on real data collected in different highway scenarios demonstrate the effectiveness and efficiency of our work. © 2000-2011 IEEE.

关键词： decision making

来源：评论

学校读者我要写书评

暂无评论

Adaptive object tracking based on vision sensor and milli-meter wave radar

引用

Journal of Computational Information systems 2014年第8期10卷 3283-3290页

作者： Jin, Lu Fu, Mengyin Yang, Yi Key Laboratory of Intelligent Control and Decision of Complex Systems Ministry of Education Beijing 100081 China School of Automation Beijing Institute of Technology Beijing 100081 China

The object tracking for driver assistant system by the multi-sensors fusion has drawn much attention recently. With the complexity of space alignment, a new space alignment algorithm is proposed to project the object by millimeter wave radar onto the image plane. The method needs not to estimate interior and exterior parameters between two sensors. Considering alignment result as the measurement, a new tracking algorithm based on the Mean Shift with Unscented Kalman Filter and Information Measure of Multi-Scale Image is proposed. The Unscented Kalman Filter firstly is used to predict the initial iteration position for Mean Shift algorithm. After the object position is located by Mean Shift algorithm, the size of tracking window is adjusted adaptively through the size change ratio of adjacent frame. The results show that the proposed algorithm can track accurately the object whose size is gradually increasing or decreasing, and the algorithm also reduces the iteration times, the location time and the tracking error. © 2014 Binary Information Press.

关键词： Alignment

来源：评论

学校读者我要写书评

暂无评论

Model Compression Based on YOLOv3 Object Detection Algorithm from the Perspective of UAV

Model Compression Based on YOLOv3 Object Detection Algorithm...

引用

第40届中国控制会议

作者： Bing Wu Zhijun Xue Wenjie Chen State Key Laboratory of Intelligent Control and Decision of Complex Systems School of AutomationBeijing Institute of Technology Beijing Institute of Technology Chongqing Innovation Center

Recently,unmanned aerial vehicles are widely used in surveillance,aerial photography,power grid line inspections and other *** order to deploy the YOLOv3[1] algorithm on drones,it is necessary to adopt the YOLOv3 algorithm with fewer parameters and a simpler *** paper implements the model compression of YOLOv3 based on methods such as sparseness,pruning,and knowledge *** paper implements the sparseness of the network by adding L1 regular expressions on the convolutional *** that,redundant channels and layers are removed through channel pruning and layer *** sparse and pruning,the mAP lost a *** using knowledge distillation after pruning,it attempts to recover mAP lost in sparseness and *** this method,the YOLOv3 algorithm can be deployed on embedded platforms such as RK3399 *** evaluate the model on the visdrone2019 *** experimental results show that after model compression,YOLOv3 is more suitable for deployment on embedded platforms.

关键词： Object Detection Model Compress RK3399pro YOLOv3

来源：评论

学校读者我要写书评

暂无评论

Study on Displacement control of the Pneumatic Muscle Actuator Basing on Model-Free Adaptive control with Nonlinear Feedback

Study on Displacement Control of the Pneumatic Muscle Actuat...

引用

International Conference on Measuring Technology and Mechatronics automation, ICMTMA

作者： Liu Yu Wang Tao Zhao Tong Fan Wei Liu Hao School of Automation Complicated System of Intelligent Control and Decision Key Laboratory of Education Ministry Beijing Institute of Technology Beijing China

ISBN: (纸本)9781424490103

Since the Pneumatic Muscle Actuator (PMA) has the characteristic of strong nonlinear and time lags, it is difficult to establish a precise mathematical mode. Model-Free Adaptive control (MFAC) is an advanced control algorithm that does not require building an off-line mathematical model. This paper is basing on the feature of the PMA and presents a model-free adaptive control algorithm with the nonlinear feedback. Finally, experimental results show the strong robustness, fast response, and high precision of this control algorithm on the displacement control of the PMA.

关键词： Adaptation model Adaptive control Mathematical model Muscles Robustness Simulation Displacement control

来源：评论

学校读者我要写书评

暂无评论

Two-stage ARMAX parameter identification based on bias-eliminated least squares estimation and Durban's method

引用

Zidonghua Xuebao/Acta Automatica Sinica 2012年第3期38卷 491-496页

作者： Xin, Bin Bai, Yong-Qiang Chen, Jie School of Automation Beijing Institute of Technology Beijing 100081 China Key Laboratory of Complex System Intelligent Control and Decision Ministry of Education Beijing 100081 China Decision and Cognitive Sciences Research Centre Manchester Business School The University of Manchester Manchester M15 6PB United Kingdom

This paper proposes a two-stage identification approach for the parameter identification of autoregressive moving average with exogenous variable (ARMAX) model. First, a bias-eliminated least squares method is employed to identify the autoregressive part with exogenous variable (ARX). Then, the Durbin's method is employed to transform the parameter identification of the moving average (MA) part into that of a long autoregressive (AR) model. The MA parameters are derived directly from the parameter relationship between the MA part and its equivalent long AR model. Finally, the noise variance can be computed by using the identified MA parameters. The performance comparison against the extended least-squares method in numerical simulations validates the effectiveness of the two-stage identification approach. Copyright © 2012 Acta Automatica Sinica. All rights reserved.

关键词： Parameter estimation

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：