检索结果-内蒙古大学图书馆

STLRF-Stack: A fault prediction model for pure electric vehicles based on a high dimensional imbalanced dataset

IET INTELLIGENT TRANSPORT SYSTEMS 2023年第2期17卷 400-417页

作者： Tian, Sheng Li, Jia Zhang, Jinming Li, Chengwei South China Univ Technol Sch Civil Engn & Transportat Guangzhou 510000 Peoples R China

Fault prediction is an important part of the safe operation and maintenance of pure electric vehicles, and it is crucial to predict faults efficiently and accurately. To address the problems that the pure electric vehicle fault prediction model has less than ideal prediction effect and classification bias to most classes due to unbalanced data set categories and high dimensionality of features, this paper firstly obtains a low dimensional balanced dataset based on hybrid sampling and joint feature selection. Then the classification ability of the model is improved through ensemble learning. Finally, a new hybrid model STLRF-Stack is proposed to realize efficient fault prediction of pure electric vehicles. This paper is validated on a real operating dataset of pure electric vehicles. The experimental results show that the prediction effect of the hybrid model proposed in this paper is better than that of the single model and related literature. In addition, this model can successfully predict the fault 5 min before the actual fault of electric vehicles occurs, which provides the drivers with the opportunity of accident prevention and early treatment. Therefore, the model has specific feasibility and practical application value.

关键词： actual fault fault diagnosis Civil and mechanical engineering computing pure electric vehicle fault prediction model learning (artificial intelligence) hybrid model STLRF-Stack feature selection pattern classification Transportation efficient fault prediction Statistics mechanical components Data handling techniques pure electric vehicles mechanical engineering computing Inspection and quality control low dimensional balanced dataset mechanical engineering applications of IT high dimensional imbalanced dataset fault 5 min accident prevention ideal prediction effect Other topics in statistics Maintenance and reliability

来源：评论

学校读者我要写书评

暂无评论

Health diagnosis of bus operation based on multi-source data

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2022年第6期16卷 754-768页

作者： Zhou, Xuemei Guan, Zhen Pang, Yusong Ji, Xiangfeng Lin, Xiaodan Tongji Univ Coll Transportat Engn Key Lab Rd & Traff Engn State Minist Educ Shanghai Peoples R China Delft Univ Technol Sect Transport Engn & Logist Mekelweg 2 NL-2628 CD Delft Netherlands Qingdao Univ Sch Business Dept Management Sci & Engn 62 Keda Branch Rd Qingdao 266000 Shandong Peoples R China

Based on the multi-source data available for bus operations, this paper proposes a health diagnosis system for single-line bus operation systems from two aspects: The operation efficiency and stability. Firstly, the index weight has been defined and calculated based on the Entropy Method. The composite index of bus operation has been obtained, and the health classification standards that pertain to efficient and effective bus operations have also been constructed (Very Healthy, Healthy, Sub-Healthy, and Unhealthy). Secondly, the more efficient machine learning method has been used in order to establish the classification algorithm training set. The effect of the k-Nearest Neighbour and Decision Tree Classification Model has also been compared and analysed in this particular study. Finally, a bus line in Foshan is taken as a case study to verify the effectiveness of the method. This paper can effectively improve the diagnosis efficiency and accuracy by introducing the artificial intelligence algorithm into bus operation diagnosis. It provides a foundation for the development of bus operation health diagnosis decision support system with the function of "bus disease" prevention and treatment.

关键词： health classification standards pattern classification decision trees bus operation health diagnosis decision support system multisource data mechanical engineering computing k-nearest neighbour bus disease prevention efficient machine learning method fault diagnosis single-line bus operation systems decision tree classification model decision support systems nearest neighbour methods entropy artificial intelligence algorithm

来源：评论

学校读者我要写书评

暂无评论

Method for optimising the performance of PML in anchor-loss limited model via COMSOL

引用

IET SCIENCE MEASUREMENT & TECHNOLOGY 2022年第6期16卷 327-336页

作者： Li, Peng Ou, Jun-Yu Yan, Jize Univ Southampton Elect & Comp Sci Southampton Hants England Univ Southampton Optoelect Res Ctr Southampton Hants England

Perfectly matched layer has been used for solving anchor-loss limited quality factor in the Micro electromechanical systems. However, setting up a well-behaved perfectly matched layer requires users to change the parameters of a perfectly matched layer to give correct results, while the current existing methods for choosing the right parameters are vague and lack theoretical support. Based on the mathematical theory of perfectly matched layer and simulation results of a beam structure's quality factor, this paper proposes a method for choosing the parameter to optimise the performance of perfectly matched layer in COMSOL. The accuracy of the proposed method is proved by matching the effect of substrate height on beam's quality factor with theory prediction. The author also studies the effect of beam height and beam width on the quality factor of the beam. The results demonstrate that simulated quality factors are in agreement with analytical values when the ratio of height over length is small but will show great divergence when height equals the length. This trend can be observed for the beam width as well. Especially for larger ratio of beam width over beam length, instead of decreasing monotonously as analytical equitation would expect, the simulated quality factor will converge into a stable value of 1700, which matches the result of two-dimensional model for the same beam structure. This means that a three-dimensional model has to be used for estimating the quality factor of a beam structure.

关键词： micromechanical devices PML beam width COMSOL beams (structures) mechanical engineering computing microelectromechanical systems solid modelling anchor-loss limited model simulated quality factor two-dimensional model perfectly matched layer mathematical theory three-dimensional model performance optimisation beam structure

来源：评论

学校读者我要写书评

暂无评论

An inferencing structure for MPO-based decentralized dynamic fault diagnosis

引用

IET CONTROL THEORY AND APPLICATIONS 2022年第2期16卷 182-191页

作者： Khaleghi, Milad Yazdi, Mojtaba Barkhordari Karimoddini, Ali Farsangi, Malihe Maghfoori Shahid Bahonar Univ Kerman Dept Elect Engn Kerman 76175133 Iran North Carolina Agr & Tech State Univ Dept Elect & Comp Engn Greensboro NC USA

This paper addresses the problem of fault diagnosis in discrete event systems through a decentralized structure. Within the proposed decentralized diagnosis structure, local most permissive bbservers (MPOs) are employed as diagnosis tools, and an effective protocol is proposed to coordinate local MPOs to infer diagnosis information. It is shown that the proposed technique can diagnose fault occurrences subjected to proper sensing decisions. A sufficient condition on sensing decision policies is derived for diagnosis of faults in a decentralized setting. Further, it is proven that the proposed decentralized structure has less complexity than a centralized structure. The details of the proposed technique are illustrated by applying the developed method to a manufacturing system.

关键词： diagnosis tools decentralized diagnosis structure diagnosis information mechanical engineering computing centralized structure decentralized structure fault diagnosis local most permissive bbservers MPO-based decentralized dynamic fault diagnosis sensing decision policies inferencing structure discrete event systems local MPOs decentralized setting fault occurrences observers

来源：评论

学校读者我要写书评

暂无评论

Intelligent fault diagnosis of rolling bearing using the ensemble self-taught learning convolutional auto-encoders

引用

IET SCIENCE MEASUREMENT & TECHNOLOGY 2022年第2期16卷 130-147页

作者： Zhang, Yilan Wang, Jinxi Zhang, Faye Lv, Shanshan Zhang, Lei Jiang, Mingshun Sui, Qingmei Shandong Univ Sch Control Sci & Engn Jinan 250061 Peoples R China

The lack of labelled data presents a common challenge in many fault diagnosis and machine learning tasks. It requires the model to be able to efficiently capture useful fault features from a smaller amount of labelled data. In this paper, a method to train multiple convolutional auto-encoders by self-learning method and integrate them using ensemble learning, called ensemble self-taught learning convolutional auto-encoders (STL-CAEs), is proposed, which can effectively extract features of bearing vibration signals. First, an ensemble learning strategy is proposed to obtain two auto-encoders that satisfy the strategy by optimizing the model parameters and structure. Then, a self-taught learning training method is proposed to solve the problem of little label data. Finally, ensemble learning and fault diagnosis is achieved by the SoftMax classifier. Applying the proposed method to the bearing data from Case Western Reserve University, the STL-CAEs have higher accuracy and generalization than common fault diagnosis methods such as CAE, CNN, SAE and EMD, and also have significant advantages in terms of diagnostic time and training time.

关键词： STL-CAEs training method labelled data presents machine learning tasks vibrations learning (artificial intelligence) self-learning method useful fault features bearing vibration signals rolling bearings ensemble learning strategy convolutional neural nets common fault diagnosis methods multiple convolutional auto-encoders mechanical engineering computing feature extraction intelligent fault diagnosis label data vibrational signal processing bearing data fault diagnosis

来源：评论

学校读者我要写书评

暂无评论

Reliability analysis of boom of new lifting equipment based on improved wolf swarm algorithm

引用

JOURNAL OF engineering-JOE 2023年第1期2023卷

作者： Guan, Huashen Tang, Bo Zhou, Xiangman Tan, Haoxin Liang, Zhuwei Li, Ye Sun, Guofu Zhang, Junjie Jiangmen Power Supply Bur Guangdong Power Grid Co Jiangmen 529000 Peoples R China China Three Gorges Univ Coll Elect Engn & New Energy Yichang Peoples R China China Three Gorges Univ Coll Mech & Power Engn Yichang Peoples R China

Aiming at the low accuracy or inability to solve the problem when the traditional reliability optimization method deals with the practical engineering reliability analysis problem, this paper proposes a structural reliability analysis method using the improved wolf swarm algorithm (IWSA). Firstly, the reliability optimization mathematical model is established with the minimum reliability index as the objective function and the limit state equation as the equality constraint. Then the augmented multiplier method is used to transform the constrained optimization mathematical model into an unconstrained optimization mathematical model. Finally, IWSA is used to solve the problem. The proposed method is applied to a numerical example and an engineering example. The computation results of the proposed method in solving the numerical example are compared with the Monte Carlo simulation (MCS), which verifies that the proposed method in this paper has good calculation accuracy. The maximum deformation under the working condition is used as the target to establish a limit state function, and the reliability index is obtained by the proposed method combining with the finite element analysis software, which provides a certain theoretical guiding significance for the reliability analysis of engineering structures.

关键词： lifting equipment augmented multiplier method equality constraint finite element analysis improved wolf swarm algorithm reliability optimization mathematical model optimisation maximum deformation mechanical engineering computing unconstrained optimization mathematical model reliability finite element analysis software practical engineering reliability analysis problem IWSA deformation engineering structures constrained optimization mathematical model minimum reliability index structural reliability analysis method limit state function limit state equation

来源：评论

学校读者我要写书评

暂无评论

A reinforcement learning backstepping-based control design for a full vehicle active Macpherson suspension system

引用

IET CONTROL THEORY AND APPLICATIONS 2022年第14期16卷 1417-1430页

作者： Lin, Yu-Chen Nguyen, Ha Ly Thi Yang, Ji-Fan Chiou, Hung-Jui Feng Chia Univ Dept Automat Control Engn Taichung 40724 Taiwan Feng Chia Univ PhD Program Elect & Commun Engn Taichung Taiwan Automot Res & Testing Ctr Elect Control & Informat Commun Technol Team Res & Dev Div Changhua Taiwan

This paper presents a reinforcement learning backstepping-based control scheme for the design of a full vehicle active suspension system such that both the superior ride comfort and stabilization are achieved. It is well known that the traditional backstepping control scheme is suitable to solve higher order nonlinear system based on the strict-feedback form;however, the disadvantage of the procedure requires computing the derivatives of virtual control signals, which is a very complex task. Therefore, a reinforcement learning scheme using the deep deterministic policy gradient (DDPG) control strategy is developed to replace the process of finding virtual control force in backstepping method. It not only avoids the complexity of analytically calculating the derivatives of the virtual control signals, but also retains the systems robustness when there exist random disturbances from road irregularities. To verify the performance of the proposed active suspension control scheme, the random road unevenness profiles according to ISO 8608 is considered as vertical and lateral disturbance input excitation of a full-vehicle suspension system. Compared with conventional passive suspension system and backstepping control scheme, the proposed approach can be demonstrated to not only effectively improve ride comfort under random road excitation, but also improve the transient response and robustness.

关键词： suspensions (mechanical components) road vehicles vehicle dynamics deep deterministic policy gradient control strategy automotive engineering reinforcement learning deep learning (artificial intelligence) higher order nonlinear system system robustness nonlinear control systems ride comfort vehicle active Macpherson suspension system random road excitation stability virtual control signals strict-feedback form random road unevenness profiles control nonlinearities feedback mechanical engineering computing gradient methods control system synthesis reinforcement learning backstepping-based control design vehicle active suspension system control engineering computing active suspension control scheme virtual control force

来源：评论

学校读者我要写书评

暂无评论

Bearing fault diagnosis based on multi-band filtering

引用

IET SCIENCE MEASUREMENT & TECHNOLOGY 2022年第2期16卷 101-117页

作者： Lu, Yixiang Song, Zhihong Gao, Qingwei Zhu, De Sun, Dong Anhui Univ Sch Elect Engn & Automat 111JiulongRoad Hefei Peoples R China

Bearing is an important part of rotary machinery, which usually operates under a variety of complicated and severe conditions, and is prone to break down. To reduce and prevent the loss caused by the fault of bearings, a method based on multi-band filtering (MBF) is proposed and applied to bearing fault diagnosis in this paper. Bearing signal is decomposed into multiple sub-band signals by an MBF constructed from a specific prototype filter (finite impulse response filter) through cosine modulation. Then, the required sub-band can be selected adaptively according to bearing fault frequency. To make the fault frequency more prominent, an adaptive filtering method is exploited to reduce the noise contained in the selected sub-band. Finally, bearing fault diagnosis is realized by envelope spectrum analysis. Experimental results show that better performance in both simulated and real data are achieved by the proposed algorithm, which indicates that the proposed method can realize bearing fault diagnosis efficiently when compared to other state of the art methods.

关键词： bearing fault frequency multiband filtering vibrational signal processing FIR filters bearing signal mechanical engineering computing fault diagnosis specific prototype filter bearing fault diagnosis adaptive filtering method adaptive filters finite impulse response filter multiple sub-band signals machine bearings

来源：评论

学校读者我要写书评

暂无评论

Feasibility of using low-sampled accelerometer measurements for bolt joint looseness detection

引用

IET RENEWABLE POWER GENERATION 2022年第13期16卷 2762-2777页

作者： Nicholas, Gary Mills, Robin Song, Wooyong Lee, Hyunjoo Dwyer-Joyce, Rob Univ Sheffield Leonardo Ctr Tribol Dept Mech Engn Sheffield S Yorkshire England Univ Sheffield Dept Mech Engn Dynam Res Grp Sheffield S Yorkshire England Offshote Renewables Energy Catapult Glasgow Lanark Scotland

In this work, the feasibility of using low-sampled vibration signals for bolt joint tightness detection was investigated. Testing was carried out on multiple bolt joint configurations using a bench top electrodynamic shaker rig. Two data-processing methods were successfully used to deduce bolt joint loosening from the accelerometer measurements, namely the resonant frequency and regression methods (ARX and AR-ARX). Both methods were able to detect loosening of bolt joints, however, the latter possesses higher sensitivity in detecting the position of the loosened bolt among an array of bolts. As the resonant frequency of wind turbines is low (0.35-2 Hz), the minimum sampling rate for bolt joint tightness detection is consequently also low (twice the resonant frequency). This facilitates potential use of existing accelerometer instrumentation on wind turbines, typically sampled at low rates.

关键词： minimum sampling rate data-processing methods accelerometers loosened bolt bolt joint looseness detection bench top electrodynamic shaker rig mechanical engineering computing bolt joint tightness detection resonant frequency accelerometer measurements joint loosening condition monitoring fasteners wind turbines vibrations bolt joints multiple bolt joint configurations

来源：评论

学校读者我要写书评

暂无评论

Using optical flow algorithm based on dynamic illumination mode to examine defects on highly reflective turbine blade surface

引用

IET IMAGE PROCESSING 2022年第14期16卷 3988-4010页

作者： Shang, Xuan Song, Wenyan Chen, Zhen Zhang, Congxuan Northwestern Polytech Univ Sch Power & Energy Xian Shaanxi Peoples R China Nanchang Hangkong Univ Sch Measuring & Opt Engn Nanchang 330063 Jiangxi Peoples R China

Notion of optical flow literally refers to the displacements of intensity patterns. In that sense, extracting interested information from 2D scene is analogy to modulation/demodulation in random signal processing. To address the limitations presented in computer vision based on static image, we propose a novel metal component defect detection method, specified as the instance of turbine blade surface detection, using optical flow *** start the specified pattern recognition in 2D presentation, we modulate the brightness constancy assumption equation as illumination varying model, by sampling the second image with function whose frequency was chosen according to the Nyquist sampling theorem, and a sinusoidal factor was introduced as an additive factor. This tunable channel based on 2D image transfers intensity features into optical modes. Then, we implement optical flow estimation on two sequential images. Experimental results reveal grayscale space shows completness in representing the optical modes of turbine blade with various kinds of surface characteristics. By modifying the index of information content, we propose quantitative index to evaluate the performance of our method. Evaluation reveals optical flow algorithm is qualified to examine defects on highly reflective turbine blade, and our method extends the application of optical flow.

关键词： blades brightness constancy assumption equation optical flow algorithm image sequences mechanical engineering computing static image illumination varying model feature extraction Nyquist sampling theorem sequential images pattern recognition turbines metal component defect detection method random signal processing image motion analysis computer vision optical flow estimation highly reflective turbine blade surface turbine blade surface detection inspection dynamic illumination mode optical modes

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：