Search

Your search keyword '"Zheng, Jinchuan"' showing total 400 results
Start Over Author "Zheng, Jinchuan"
400 results on '"Zheng, Jinchuan"'

1. VE: Modeling Multivariate Time Series Correlation with Variate Embedding

Author

Wang, Shangjiong, Man, Zhihong, Cao, Zhenwei, Zheng, Jinchuan, and Ge, Zhikang

Subjects
Computer Science - Machine Learning
Abstract

Multivariate time series forecasting relies on accurately capturing the correlations among variates. Current channel-independent (CI) models and models with a CI final projection layer are unable to capture these dependencies. In this paper, we present the variate embedding (VE) pipeline, which learns a unique and consistent embedding for each variate and combines it with Mixture of Experts (MoE) and Low-Rank Adaptation (LoRA) techniques to enhance forecasting performance while controlling parameter size. The VE pipeline can be integrated into any model with a CI final projection layer to improve multivariate forecasting. The learned VE effectively groups variates with similar temporal patterns and separates those with low correlations. The effectiveness of the VE pipeline is demonstrated through experiments on four widely-used datasets. The code is available at: https://github.com/swang-song/VE.

Published
2024

13. Optimizing Natural Image Quality Evaluators for Quality Measurement in CT Scan Denoising.

Author

Gunawan, Rudy, Tran, Yvonne, Zheng, Jinchuan, Nguyen, Hung, and Chai, Rifai

Abstract

Evaluating the results of image denoising algorithms in Computed Tomography (CT) scans typically involves several key metrics to assess noise reduction while preserving essential details. Full Reference (FR) quality evaluators are popular for evaluating image quality in denoising CT scans. There is limited information about using Blind/No Reference (NR) quality evaluators in the medical image area. This paper shows the previously utilized Natural Image Quality Evaluator (NIQE) in CT scans; this NIQE is commonly used as a photolike image evaluator and provides an extensive assessment of the optimum NIQE setting. The result was obtained using the library of good images. Most are also part of the Convolutional Neural Network (CNN) training dataset against the testing dataset, and a new dataset shows an optimum patch size and contrast levels suitable for the task. This evidence indicates a possibility of using the NIQE as a new option in evaluating denoised quality to find improvement or compare the quality between CNN models. [ABSTRACT FROM AUTHOR]

Published
2025
Full Text
View/download PDF

21. Fixed-Time Adaptive Sliding Mode Control for Vehicular Electronic Throttle With Actuator Saturation Using Extreme Learning Machine

Author

Hu, Youhao, Wang, Hai, Yazdani, Amirmehdi, Chen, Liqing, Yu, Ming, Zheng, Jinchuan, and Man, Zhihong

Abstract

This paper proposes an extreme learning machine (ELM)-based adaptive sliding mode control strategy for vehicular electronic throttle (VET) systems with parametric uncertainties, lumped uncertainty and actuator constraint. The proposed control strategy adopts a fixed-time sliding mode (FTSM) dynamical structure which ensures a fixed time convergence property for both reaching motion and sliding motion. For relaxing the upper bound information constraint of sliding mode control design, an ELM-based mechanism is utilized for learning the lumped uncertainty bound online, while the ELM output weights are updated adaptively in the sense of Lyapunov. Also, considering that the input voltage is normally limited due to the vehicular battery constraint for practical applications, to handle the input saturation scenario, an auxiliary system (AS) is further designed to guarantee a fixed time convergence of auxiliary state and the state is correspondingly fed into the controller design. The global stability analysis of the closed-loop system is rigidly given. Comparative experimental studies are conducted to illustrate the excellent performance of the proposed control.

Published
2024
Full Text
View/download PDF

33. Combining Multistaged Filters and Modified Segmentation Network for Improving Lung Nodules Classification

Author

Gunawan, Rudy, Tran, Yvonne, Zheng, Jinchuan, Nguyen, Hung, Carrigan, Ann, Mills, Megan K., and Chai, Rifai

Abstract

Advancements in computational technology have led to a shift towards automated detection processes in lung cancer screening, particularly through nodule segmentation techniques. These techniques employ thresholding to distinguish between soft and firm tissues, including cancerous nodules. The challenge of accurately detecting nodules close to critical lung structures such as blood vessels, bronchi, and the pleura highlights the necessity for more sophisticated methods to enhance diagnostic accuracy. This paper proposed combined processing filters for data preparation before using one of the modified Convolutional Neural Networks (CNNs) as the classifier. With refined filters, the nodule targets are solid, semi-solid, and ground glass, ranging from low-stage cancer (cancer screening data) to high-stage cancer. Furthermore, two additional works were added to address juxta-pleural nodules while the pre-processing end and classification are done in a 3-dimensional domain in opposition to the usual image classification. The accuracy output indicates that even using a simple Segmentation Network if modified correctly, can improve the classification result compared to the other eight models. The proposed sequence total accuracy reached 99.7%, with 99.71% cancer class accuracy and 99.82% non-cancer accuracy, much higher than any previous research, which can improve the detection efforts of the radiologist.

Published
2024
Full Text
View/download PDF

35. Cooperative game-oriented optimal robust control for an uncertain wheeled mobile robot with position constraints.

Author

Zheng, Yunjun, Zhao, Han, Zheng, Jinchuan, and Shao, Ke

Abstract

The trajectory tracking control performance of the wheeled mobile robot (WMR) is subject to position constraints, system uncertainties, and external disturbances. This paper explores the mechanism of the position constraints (equality and inequality) and considers the system uncertainties to be time-varying but bounded with fuzzy prescribed bounds. To solve the position constraints and uncertainties problems, a cooperative game-oriented optimal robust control method is proposed to provide high-precision, fast-response performance and robustness for WMR. First, a creative diffeomorphism transformation is proposed to convert the bounded state variables (subject to inequality constraints) into unbounded state variables. Accordingly, a restructured WMR system with the new state variables is established, which is free from inequality constraints. The uncertainties' bound is described as a fuzzy number by employing fuzzy set theory, which builds the fuzzy dynamic system of the WMR. Second, the control goal for this study is to drive the WMR to follow the position constraints despite uncertainties. An optimal robust control is developed to guarantee that the tracking errors of the WMR converge to a predefined neighborhood of origin. Using Lyapunov stability analysis, a fuzzy performance index is constructed, including steady-state performance, time performance, and control effort. The fuzzy performance index is employed as the cost function based on the cooperative game theory. The optimal control design is formulated to seek the Pareto-optimal solution of control parameters. Finally, the stability analysis and comparative experiments demonstrate that the proposed control algorithm can ensure the WMR's superior trajectory tracking convergence performance and strong robustness against uncertainties/disturbances. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

49. On the efficient particle dispersion and transfer in the fabrication of SiC-particle-reinforced aluminum matrix composite

Author

Du, Andong, Lattanzi, Lucia, Jarfors, Anders E.W., Zheng, Jinchuan, Wang, Kaikun, Yu, Gegang, Du, Andong, Lattanzi, Lucia, Jarfors, Anders E.W., Zheng, Jinchuan, Wang, Kaikun, and Yu, Gegang

Abstract

Lightweight SiC-particle-reinforced aluminum composites have the potential to replace cast iron in brake discs, especially for electric vehicles. This study investigates the effect of SiC particle size and matrix alloy composition on the resulting transfer efficiency and particle distribution. The performance of a specially designed stirring head was studied using a water model, and the stirring head conditions were assessed to understand the particle transfer and dispersion mechanisms in the molten aluminum. The standard practice of thermal pre-treatment promotes the wetting of the reinforcing particles and commonly causes clustering before the addition to the melt. This early clustering affects the transfer efficiency and particle dispersion, where their interaction with the melt top-surface oxide skin plays an important role. In addition, the transfer efficiency was linked to the particle size and the chemical composition of the matrix alloy. Smaller particles aggravated the degree of clustering, and the addition of rare earth elements as alloying elements in the matrix alloy affected the particle dispersion. The stirring parameters should be selected to ensure cluster disruption when the carbides are added to the melt.

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results