检索结果-南通市图书馆

iterative Semi-Supervised learning Using Softmax Probability

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

Computers, Materials & Continua 2022年第9期72卷 5607-5628页

作者： Heewon Chung Jinseok Lee Department of Biomedical Engineering College of Electronics and InformationKyung Hee UniversityYongin-siGyeonggi-do17104Korea

For the classification problem in practice,one of the challenging issues is to obtain enough labeled data for training.Moreover,even if such labeled data has been sufficiently accumulated,most datasets often exhibit long-tailed distribution with heavy class imbalance,which results in a biased model towards a majority class.To alleviate such class imbalance,semisupervised learning methods using additional unlabeled data have been considered.However,as a matter of course,the accuracy is much lower than that from supervised learning.In this study,under the assumption that additional unlabeled data is available,we propose the iterative semi-supervised learning algorithms,which iteratively correct the labeling of the extra unlabeled data based on softmax probabilities.The results show that the proposed algorithms provide the accuracy as high as that from the supervised learning.To validate the proposed algorithms,we tested on the two scenarios:with the balanced unlabeled dataset and with the imbalanced unlabeled dataset.Under both scenarios,our proposed semi-supervised learning algorithms provided higher accuracy than previous state-of-the-arts.Code is available at https://***/HeewonChung92/iterative-semi-learning.

关键词： Semi-supervised learning class imbalance iterative learning unlabeled data

iterative learning Disturbance Observer Based Attitude Stabilization of Flexible Spacecraft Subject to Complex Disturbances and Measurement Noises

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

IEEE/CAA Journal of Automatica Sinica 2021年第9期8卷 1576-1587页

作者： Tongfu He Zhong Wu School of Instrumentation and Optoelectronics Engineering Beihang UniversityBeijing 100191China

To realize high-precision attitude stabilization of a flexible spacecraft in the presence of complex disturbances and measurement noises,an iterative learning disturbance observer(ILDO)is presented in this paper.Firstly,a dynamic model of disturbance is built by augmenting the integral of the lumped disturbance as a state.Based on it,ILDO is designed by introducing iterative learning structures.Then,comparative analyses of ILDO and traditional disturbance observers are carried out in frequency domain.It demonstrates that ILDO combines the advantages of high accuracy in disturbance estimation and favorable robustness to measurement noise.After that,an ILDO based composite controller is designed to stabilize the spacecraft attitude.Finally,the effectiveness of the proposed control scheme is verified by simulations.

关键词： Disturbance observer iterative learning measure-ment noise spacecraft attitude control

iterative learning based fault diagnosis for discrete linear uncertain systems

在线全文

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2014年第3期25卷 496-501页

作者： Wei Cao Ming Sun College of Computer and Control Engineering Qiqihar University

In order to detect and estimate faults in discrete lin-ear time-varying uncertain systems, the discrete iterative learning strategy is applied in fault diagnosis, and a novel fault detection and estimation algorithm is proposed. And the threshold limited technology is adopted in the proposed algorithm. Within the chosen optimal time region, residual signals are used in the proposed algorithm to correct the introduced virtual faults with iterative learning rules, making the virtual faults close to these occurred in practical systems. And the same method is repeated in the rest optimal time regions, thereby reaching the aim of fault diagnosis. The proposed algorithm not only completes fault detection and estimation for discrete linear time-varying uncertain systems, but also improves the reliability of fault detection and decreases the false alarm rate. The final simulation results verify the validity of the proposed algorithm.

关键词： discrete linear uncertain system threshold limited technology iterative learning fault estimation.

iterative learning based fault detection and estimation in nonlinear systems

在线全文

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2012年第3期23卷 419-424页

作者： Wei Cao Wang Cong Ming Sun College of Automation Harbin Engineering University Harbin 150001 E R. China College of Computer and Control Engineering Qiqihar University Qiqihar 161006 E R. China

Aiming at a class of nonlinear systems that contains faults, a novel iterative learning scheme is applied to fault detec- tion, and a novel algorithm of fault detection and estimation is proposed. This algorithm first constructs residual signals by the output of the practical system and the output of the designed fault tracking estimator, and then uses the residuals and the difference- value signal of the adjacent two residuals to gradually revise the introduced virtual faults, which can cause the virtual faults to close to the practical faults in systems, thereby achieving the goal of fault detection for systems. This algorithm not only makes full use of the existing valid information of systems and has a faster tracking con- vergent speed than the proportional-type （P-type） algorithm, but also calculates more simply than the proportional-derivative-type （PD-type） algorithm and avoids the unstable effects of differential operations in the system. The final simulation results prove the validity of the proposed algorithm.

关键词： iterative learning nonlinear system fault detection fault estimation.

iterative learning-based many-objective history matching using deep neural network with stacked autoencoder

在线全文

学校读者我要写书评

暂无评论

Petroleum Science 2021年第5期18卷 1465-1482页

作者： Jaejun Kim Changhyup Park Seongin Ahn Byeongcheol Kang Hyungsik Jung Ilsik Jang Department of Energy Systems Engineering Seoul National UniversitySeoul08826Republic of Korea Department of Energy and Resources Engineering Kangwon National UniversityChuncheonKangwon24341Republic of Korea Geo-ICT Convergence Research Team Korea Institute of Geoscience and Mineral ResourcesDaejeon34132Republic of Korea Department of Energy and Resources Engineering Chosun UniversityGwangju61452Republic of Korea

This paper presents an innovative data-integration that uses an iterative-learning method,a deep neural network(DNN)coupled with a stacked autoencoder(SAE)to solve issues encountered with many-objective history matching.The proposed method consists of a DNN-based inverse model with SAE-encoded static data and iterative updates of supervised-learning data are based on distance-based clustering schemes.DNN functions as an inverse model and results in encoded flattened data,while SAE,as a pre-trained neural network,successfully reduces dimensionality and reliably reconstructs geomodels.The iterative-learning method can improve the training data for DNN by showing the error reduction achieved with each iteration step.The proposed workflow shows the small mean absolute percentage error below 4%for all objective functions,while a typical multi-objective evolutionary algorithm fails to significantly reduce the initial population uncertainty.iterative learning-based manyobjective history matching estimates the trends in water cuts that are not reliably included in dynamicdata matching.This confirms the proposed workflow constructs more plausible geo-models.The workflow would be a reliable alternative to overcome the less-convergent Pareto-based multi-objective evolutionary algorithm in the presence of geological uncertainty and varying objective functions.

关键词： Deep neural network Stacked autoencoder History matching iterative learning Clustering Many-objective

Trajectory tracking control of a bionic robotic fish based on iterative learning

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2020年第7期63卷 79-87页

作者： Ming WANG Yanlu ZHANG Huifang DONG Junzhi YU School of Information and Electrical Engineering Shandong Jianzhu University State Key Laboratory for Turbulence and Complex System Department of Mechanics and Engineering ScienceBIC-ESAT College of Engineering Peking University

A bionic robotic fish has great potential application prospect. High maneuverability swimming control of a bionic robotic fish has been one of the research hotspots in the robotic fish field. In this paper,an iterative learning method has been proposed to solve the trajectory tracking control problem of robotic fish swimming. First, a dynamic model of the multi-joint bionic robotic fish is established. By considering a three-joint robotic fish as an example, the unified expression of the dynamic equation of the three-joint bionic robotic fish is obtained by Lagrange method. Second, the iterative learning controller for controlling the bionic robotic fish is designed. Then the convergence of the iterative learning controller is proved.Finally, the trajectory tracking control simulation experiment based on iterative learning is conducted. The simulation results show that the trajectory tracking control method based on iterative learning for a bionic robotic fish is effective.

关键词： bionic robot fish iterative learning trajectory tracking motion control

Robust iterative learning Control of Single-phase Grid-connected Inverter

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

International Journal of Automation and computing 2014年第4期11卷 404-411页

作者： Zhong-Qiang Wu Chun-Hua Xu Yang Yang Key Lab of Industrial Computer Control Engineering of Hebei Province Yanshan University

For the single phase inductance-capacitance-inductance(LCL) grid-connected inverter in micro-grid, a kind of robust iterative learning controller is designed. Based on the output power droop characteristics of inverter, the current sharing among the inverters is achieved. iterative learning strategy is suitable for repeated tracking control and inhibiting periodic disturbance, and is designed using robust performance index, so that it has the ability to overcome the uncertainty of system parameters. Compared with the repetitive control, the robust iterative learning control can get high precision output waveform, and enhance the tracking ability for waveform, and the distortion problem of the output signal can be solved effectively.

关键词： Inductance-capacitance-inductance (LCL) filter iterative learning robust control power droop-characteristics repetitive control

Arranging transient process used in iterative learning control system

在线全文

学校读者我要写书评

暂无评论

Journal of Pharmaceutical Analysis 2009年第1期21卷 42-45页

作者： Li-chuan Hui, Hui Lin School of Automation, Northwestern Polytechnical University, Xi’an 710129, China. School of Automation Northwestern Polytechnical University Xi'an 710129 China

Considering the same initial state error in each repetitive operation in the iterative learning system, a method of arranging the transient process is given. During the current iteration, the system will track the transient function firstly, and then the expected trajectory. After several iterations, the learning system output will trend to the arranged curve, which has avoided the effect of the initial error on the controller. Also the transient time can be changed as you need, which makes the designing simple and the operation easy. Then the detailed designing steps are given via the robot system. At last the simulation of the robot system is given, which shows the validity of the method.

关键词： iterative learning initial state transient process robot system

Variable Stiffness iterative learning Control for FlexibleJoint Robot

在线全文

学校读者我要写书评

暂无评论

Variable Stiffness Iterative Learning Control for FlexibleJo...