检索结果-南通市图书馆

Model updating for real time dynamic substructures based on UKF algorithm

Earthquake Engineering and Engineering Vibration 2020年第2期19卷 413-421页

作者： Su Tingli Tang Zhenyun Peng Lingyun Bai Yuting jin xuebo Kong Jianlei School of Computer Information and Engineering Beijing Technology and Business UniversityBeijing 100048China The Key Laboratory of Urban Security and Disaster Engineering Ministry of EducationBeijing University of TechnologyBeijing 100124China Beijing Key Laboratory of Big Data Technology for Food Safety Beijing Technology and Business UniversityBeijing 100048China China Light Industry Key Laboratory of Industrial Internet and Big Data Beijing Technology and Business UniversityBeijing 100048China

Combining the advantages of numerical simulation with experimental testing,real-time dynamic substructure(RTDS)testing provides a new experimental method for the investigation of engineered ***,not all unmodeled parts can be physically tested,as testing is often limited by the capacity of the test *** updating is a good option to improve the modeling accuracy for numerical substructures in *** this study,a model updating method is introduced,which has great performance in describing this *** order to determine the optimal parameters in this model,an Unscented Kalman Filter(UKF)-based algorithm was applied to extract the knowledge contained in the sensors *** the parameters that need to be identified are listed as the extended state variables,and the identification was achieved via the step-by-step state prediction and state update *** of the proposed method was verified through a group of experimental data,and results showed good ***,the proposed method was compared with the Extended Kalman Filter(EKF)-based method,and better accuracy was easily *** proposed parameter identification method has great applicability for structural objects with nonlinear behaviors and could be extended to research in other engineering fields.

关键词： dynamic substructure complex nonlinear model nonlinear estimation adaptive updating civil infrastructures

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Data Generative State Estimation Based on Transformer and LSTM

Data Generative State Estimation Based on Transformer and LS...

引用

第35届中国过程控制会议

作者： xuebo jin Tianxiao Sun Huijun Ma Beijing Technology and Business University School of Computer and Artificial Intelligence

In modern times, state estimation is widely used in fields such as autonomous driving and unmanned aerial vehicle navigation. However, in practice, it is a challenge to obtain accurate target motion model and noise covariance. This difficulty often leads to a decline in the accuracy of traditional Kalman filter estimation. In order to solve this problem, a data generating state estimation method based on attention parameter learning module is introduced. This method combines the encoder and long-short memory network(LSTM) capabilities of the transformer. With off-line learning of the generated data, the method captures the operational characteristics of the system without explicitly modeling the System dynamics and measurement characteristics. In addition, the output of the attention learning module can be estimated online using expectation maximization(EM) algorithm. Then, using the Kalman filter to complete the state estimation. The validity of the method is verified by GPS trajectory data. The experimental results show that the method has higher estimation precision than other models. Therefore, the proposed parameter adaptive, model-free state estimation method provides an attractive method for the use of deep learning networks in state estimation tasks.

关键词： State estimation Kalman filter Transformer LSTM

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

YOLOv7-Bw: Enhancing Remote Sensing Small Object Detection with Generative Model Data

YOLOv7-Bw: Enhancing Remote Sensing Small Object Detection w...

引用

第35届中国过程控制会议

作者： xuebo jin Anshuo Tong Huijun Ma Beijing Technology and Business university School of Computer and Artificial Intelligence

In recent years, deep learning technology has been increasingly applied in the detection of remote sensing ***, remote sensing images generally have large differences in object size and dense distribution, which requires high performance of detection algorithms. The current detection methods are generally inefficient, prone to missed detections and inaccurate detection boxes. Therefore, this article improves the YOLO algorithm and proposes a YOLOv7-based algorithm YOLOv7-bw, which achieves efficient detection of remote sensing images. Especially for object detection in small object dense and fuzzy remote sensing images, this algorithm significantly improves detection accuracy and robustness, promoting the application and development of object detection in the remote sensing industry. Meanwhile, compared with commonly used algorithms, our YOLOv7-bw performs better, proving that our proposed algorithm is feasible and can be better applied to remote sensing image detection.

关键词： remote sensing image YOLO object detection m AP

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

A deep network prediction model for heavy metal cadmium in the rice supply chain

引用

Journal of Future Foods 2021年第2期1卷 196-202页

作者： xuebo jin Jiashuai Zhang Xiaoyi Wang Xin Zhang Tianyang Guo Ce Shi Tingli Su Jianlei Kong Yuting Bai College of Artifi cial Intelligence Beijing Technology and Business UniversityBeijing 100048China Beijing Institute of Fashion Technology Beijing 100029China College of Food and Health Beijing Technology and Business UniversityBeijing 100048China Beijing Research Center for Information Technology in Agriculture Beijing 100097China

Cadmium and its compounds are currently known as Class I carcinogens,and excessive intake can cause severe health damage to *** has a strong absorption effect on cadmium,and rice products with excessive cadmium content have caused several significant public health contamination *** is essential to predict the development trend of cadmium hazards in the rice supply chain so that countermeasures can be formulated to reduce the *** paper proposes a deep prediction model for cadmium hazards in the rice supply chain based on the regularization ***,a long and short-term memory network is used to build the depth prediction model by using the regularization method,and the noise penalty term is added to reduce the model fitting to the noise and prevent the over-fitting caused by the ***,the optimization of the model hyperparameters was carried out using a Bayesian optimization approach to develop the prediction ***,early warning system for prediction of cadmium hazards in the rice supply chain is built based on the deep prediction model proposed in this paper with SOA architecture,including data resource,business logic,and application service *** proposed model performs well on an actual data set of cadmium hazards in the rice supply chain and fits the data well.

关键词： Rice supply chain Cadmium Predictive models Long and short term memory Bayesian optimization

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Multi-Factor Remote Sensing Image Analysis and Cyanobacterial Bloom Prediction Based on 4D-Pix2Pix Model

Multi-Factor Remote Sensing Image Analysis and Cyanobacteria...

引用

第35届中国过程控制会议

作者： Li Wang Yafei Chen Xiaoyi Wang xuebo jin Zaiwen Liu Beijing Laboratory for Intelligent Environmental Protection School of Computer and Artificial Intelligence Beijing Technology &Business University

In the existing methods for pixel-level prediction of future remote sensing images based on remote sensing images, the impact of other factors on the outbreak of cyanobacterial blooms is often overlooked, with only the single factor of chlorophyll-a concentration considered. This paper takes into account the influence of chlorophyll-a, temperature, and phycocyanin concentration on the outbreak of cyanobacterial blooms. A 4D-Pix2Pix model is constructed based on the Pix2Pix model to achieve pixel-level prediction of multi-factor remote sensing images. By cascading the constructed Conv3D-LSTM with a 3D-Unet network that incorporates a residual structure as a generator, the Conv3D-LSTM can perform feature extraction on the two-dimensional space, one-dimensional time, and one-dimensional multi-factors of the multi-factor remote sensing image time series. The extracted features are then used by the 3D-Unet network with a residual structure to achieve pixel-level prediction of multi-factor remote sensing images at future times. Adversarial training is conducted with 3D-PatchGAN to optimize network performance. Finally, a cyanobacterial bloom outbreak level standard based on multi-factor remote sensing images is proposed to predict the spatiotemporal distribution of cyanobacterial blooms.

关键词： Remote sensing image prediction Pix2Pix model cyanobacteria bloom water eutrophication

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Inertial Pose Estimation Method Based on Multi-Genre Cascade Networks

Inertial Pose Estimation Method Based on Multi-Genre Cascade...

引用

第43届中国控制会议

作者： Chenguang Zhou Yuting Bai Tingli Su xuebo jin Jianlei Kong College of Computer and Artificial Intelligence Beijing Technology and Business University Beijing Laboratory for Intelligent Environmental Protection Beijing Technology and Business University

Attitude estimation using an inertial measurement unit is a critical challenge in aerospace, robotics control, unmanned systems, and related fields, with its precision directly impacting the performance and safety of navigation systems. Conventional methods for attitude estimation typically rely on intricate mathematical models and human expertise, and their accuracy is constrained by the intricate noise characteristics of sensors like accelerometers and gyroscopes. In response to this issue, the multi-genre cascade networks methodology is introduced, which integrates recurrent neural networks, multiple attention layers,and echo state networks to enable feature extraction and modeling of inertial measurement data. Experimental validation using publicly available datasets demonstrates that the fused results exhibit superior accuracy compared to conventional fusion methodologies and single neural network fusion strategies.

关键词： Attitude Estimation Composite Neural Network Attention Mechanism Echo State Network

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

A Generative Model-Based Network Framework for Ecological Data Reconstruction

引用

Computers, Materials & Continua 2025年第1期82卷 929-948页

作者： Shuqiao Liu Zhao Zhang Hongyan Zhou xuebo Chen School of Electronic and Information Engineering University of Science and Technology LiaoningAnshan114051China School of Computer Science and Software Engineering University of Science and Technology LiaoningAnshan114051China

This study examines the effectiveness of artificial intelligence techniques in generating high-quality environmental data for species introductory site selection *** Strengths,Weaknesses,Opportunities,Threats(SWOT)analysis data with Variation Autoencoder(VAE)and Generative AdversarialNetwork(GAN)the network framework model(SAE-GAN),is proposed for environmental data *** model combines two popular generative models,GAN and VAE,to generate features conditional on categorical data embedding after SWOT *** model is capable of generating features that resemble real feature distributions and adding sample factors to more accurately track individual sample *** data is used to retain more semantic information to generate *** model was applied to species in Southern California,USA,citing SWOT analysis data to train the *** show that the model is capable of integrating data from more comprehensive analyses than traditional methods and generating high-quality reconstructed data from them,effectively solving the problem of insufficient data collection in development *** model is further validated by the Technique for Order Preference by Similarity to an Ideal Solution(TOPSIS)classification assessment commonly used in the environmental data *** study provides a reliable and rich source of training data for species introduction site selection systems and makes a significant contribution to ecological and sustainable development.

关键词： Convolutional Neural Network(CNN) VAE GAN TOPSIS data reconstruction

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Continuous Estimation of Motion State in GPS/INS Integration Based on NARX Neural Network

Continuous Estimation of Motion State in GPS/INS Integration...

引用

第37届中国控制会议

作者： Yuting Bai Baihai Zhang Senchun Chai xuebo jin Xiaoyi Wang Tingli Su School of Automation Beijing Institute of Technology School of Computer and Information Engineering Beijing Technology and Business University

The accurate estimation of motion state has been the key node in the control of motion systems. The traditional GPS/INS（Global Position System/Inertial Navigation System） integration may be invalid in the outage of GPS signals. The continuous estimation framework and method were proposed to estimate the location in different environments. Firstly, a continuous framework was designed combining the traditional GPS/INS integration and the intelligent neural network. The methods were switched according to the condition of sensors to realize the continuous estimation. Secondly, NARX（Nonlinear Autoregressive with Exogenous Inputs） neural network was built to model the nonlinear mapping relation between INS and GPS. The time series data were analyzed in NARX neural network to excavate the data features in the time dimension. Lastly, the experiment was conducted to verify the method proposed. And the results showed that the solution is valid in the motion state estimation with INS when GPS is in the outage.

关键词： State estimation NARX neural network GPS INS information fusion

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

基于多流高斯概率融合网络的病虫害细粒度识别

引用

农业工程学报 2020年第13期36卷 148-157页

作者：孔建磊金学波陶治王小艺林森北京工商大学人工智能学院北京100048 国家环境保护食品链污染防治重点实验室北京100048 中国轻工业工业互联网与大数据重点实验室北京100048 北京农业智能装备技术研究中心北京100097

为解决由于现有深度迁移学习无法有效匹配实际农业场景部署应用,而导致大规模、多类别、细粒度的病虫害辨识准确低、泛化鲁棒差等问题,该研究利用农业物联网中多种设备终端获取12.2万张181类病虫害图像,并提出了基于多流概率融合网络MPF... 详细信息

为解决由于现有深度迁移学习无法有效匹配实际农业场景部署应用,而导致大规模、多类别、细粒度的病虫害辨识准确低、泛化鲁棒差等问题,该研究利用农业物联网中多种设备终端获取12.2万张181类病虫害图像,并提出了基于多流概率融合网络MPFN(Multi-stream Gaussian Probability Fusion Network)的病虫害细粒度识别模型。该模型设计多流深度网络并行的细粒度特征提取层,挖掘可区分细微差异的不同级别局部特征表达,经过局部描述特征聚合层和高斯概率融合层的整合优化,发挥多模型融合信息互补及置信耦合的优势,既可以有效区分不同类病虫害的种间微小差异,又可容忍同类病虫害种内明显差异干扰。对比试验表明,该研究MPFN模型对各类病虫害的平均识别准确率达到93.18%,性能优于其他粗粒度及细粒度深度学习方法;而平均单张处理时间为61ms,能够满足农业生产实践中物联网各终端病虫害细粒度图像识别需求,可为智能化病虫害预警防控提供技术应用参考,进而为保障农作物产量和品质安全提供基础。

关键词：农业图像识别病虫害深度神经网络特征聚合高斯概率融合

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

温度和水分影响下的粮食真菌污染预测及优化控制

引用

中国粮油学报 2023年第8期38卷 9-20页

作者：王立郑浪金学波王小艺于家斌白玉廷北京工商大学人工智能学院北京100048 北京服装学院北京100105

构建以温度和水分为影响因素的真菌生长模型预测真菌生长趋势,并提出对粮食储藏过程温度和水分的优化控制方法。在现有基于温度的微生物生长动力学模型基础上,加入水分影响因素,并通过改进二维正态分布模型将菌落直径生长模型转换为真... 详细信息

构建以温度和水分为影响因素的真菌生长模型预测真菌生长趋势,并提出对粮食储藏过程温度和水分的优化控制方法。在现有基于温度的微生物生长动力学模型基础上,加入水分影响因素,并通过改进二维正态分布模型将菌落直径生长模型转换为真菌孢子数增长模型;通过对模型离散化、增量化实现对真菌生长趋势的预测;根据粮食真菌污染的特点,改进预测控制模型,建立以真菌生长为预测变量、以温度和水分为控制变量的优化控制目标函数,并采用不动点法迭代求解实现菌落生长速率和温度、水分控制成本的全局最小化。采用稻谷储藏实验数据验证,真菌生长模型拟合的确定系数达到0.7553,真菌生长速率降低到0.0263×10^(6)个/g·d。

关键词：粮食真菌污染预测控制

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

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：