Hong Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Hong Zhang has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
- Jiawen Xu
Southeast University
2 shared publications - Liye Zhao
School of Instrument Science and Engineering, Southeast University, Nanjing, China
2 shared publications - Luxiang Xu
Institute for Advanced Study, University of Chinese Academy of Sciences, Hangzhou, China
2 shared publications - Ning Guo
Science and Technology on Vacuum Technology and Physics Laboratory, Lanzhou Institute of Physics, Lanzhou, China
2 shared publications - Xingyu Chen
Key Laboratory of Micro-Inertial Instruments and Advanced Navigation Technology, School of Instrument Science and Engineering, Southeast University, Nanjing, China
2 shared publications - Zhuoping Dai
Jiangsu Key Laboratory of Robot Sensor and Control Laboratory, School of Instrument Science and Engineering, Southeast University, Nanjing, China
2 shared publications
Latest Publications
- 5 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-16
Ultralow-Frequency Drifting Compensation for Microthrust Measurement Using Autoencoder Prediction-Based Auto-Zero Mechanism
Micro-thrust measurement technology plays a pivotal role in detecting space gravitational waves. However, stochastic drifting in the ultra-low frequency range (0.1 mHz) poses a significant challenge to the precision of micro-thrust measurements. To mitigate the low-frequency drifting during micro-thrust measurement, this paper introduces a multiple-scale drifting compensation method for ultra-low frequency micro-thrust measurement based on autoencoder prediction. The multiple-scale drifting compensating consists of a large-scale drifting detrending and a micro-scale prediction and compensation. A short-time micro-scale drift skeleton of the raw signal is extracted using a bidirectional Butterworth low-pass filter. An unsymmetrical autoencoder model is employed to predict the following time sequence of the low-frequency drift skeleton. Then, the predicted skeleton sequence is removed from the raw signal for drifting compensation. Concurrently, zero load data is utilized for experimental testing to select the optimal experimental method, followed by validation experiments using actual measurement data to demonstrate the method’s effectiveness. It is experimentally validated that the proposed method can successfully reduce the noise level of the measured micro-thrust data to 0.43 μN/Hz1/2 at a frequency of 7.94×10-5 Hz in the laboratory without vibration suspension and temperature control. The proposed method can reduce the magnitude of drifting by approximately 3 orders at 0.1 mHz, confirming its effectiveness in compensating low-frequency drifting.
- 5 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-16
Ultralow-Frequency Drifting Compensation for Microthrust Measurement Using Autoencoder Prediction-Based Auto-Zero Mechanism
Micro-thrust measurement technology plays a pivotal role in detecting space gravitational waves. However, stochastic drifting in the ultra-low frequency range (0.1 mHz) poses a significant challenge to the precision of micro-thrust measurements. To mitigate the low-frequency drifting during micro-thrust measurement, this paper introduces a multiple-scale drifting compensation method for ultra-low frequency micro-thrust measurement based on autoencoder prediction. The multiple-scale drifting compensating consists of a large-scale drifting detrending and a micro-scale prediction and compensation. A short-time micro-scale drift skeleton of the raw signal is extracted using a bidirectional Butterworth low-pass filter. An unsymmetrical autoencoder model is employed to predict the following time sequence of the low-frequency drift skeleton. Then, the predicted skeleton sequence is removed from the raw signal for drifting compensation. Concurrently, zero load data is utilized for experimental testing to select the optimal experimental method, followed by validation experiments using actual measurement data to demonstrate the method’s effectiveness. It is experimentally validated that the proposed method can successfully reduce the noise level of the measured micro-thrust data to 0.43 μN/Hz1/2 at a frequency of 7.94×10-5 Hz in the laboratory without vibration suspension and temperature control. The proposed method can reduce the magnitude of drifting by approximately 3 orders at 0.1 mHz, confirming its effectiveness in compensating low-frequency drifting.
- 24 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-15
A High-Generalization Variational Denoising Autoencoder for Micronewton Thrust Signal Noise Removal and Step Reconstruction
Removing noise and recovering the micronewton thrust signal are of great significance in high-precision static thrust measurements. Typically, the micronewton thrust signal is in the shape of a staircase signal. Existing methods have limitations in decoupling sharp step edges and flat regions from noisy signals while ensuring the accuracy of step amplitude reconstruction. In this study, we have developed a novel generative denoising method, named variational denoising autoencoder (VDAE), based on a unique deep learning-based Bayesian framework. Specifically, the encoder-parameterized approximate posterior maps the distribution of essential features (i.e., thrust step amplitudes) of limited training samples to a latent space with a Gaussian distribution. This distribution transformation gives the latent space the ability to describe complete continuous step amplitudes. VDAE inherits the excellent generalization ability of the generative model and greatly improves the amplitude accuracy of the denoised signals. In addition, considering the different scale features in the clean staircase signal, a trend feature disentangler (TFD) is introduced in the encoder. The TFD adaptively extracts ultra-high-frequency sharp step edge features and ultra-low-frequency flat region features. Furthermore, to address the issue of recovering sharp step edges, total variation (TV) sparse representation is introduced into the loss function, guiding the decoder to reconstruct the thrust step. Extensive simulations and experiments were carried out to demonstrate the effectiveness and superiority of the proposed method in micronewton thrust step reconstruction and measurement noise removal.
- 24 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-15
A High-Generalization Variational Denoising Autoencoder for Micronewton Thrust Signal Noise Removal and Step Reconstruction
Removing noise and recovering the micronewton thrust signal are of great significance in high-precision static thrust measurements. Typically, the micronewton thrust signal is in the shape of a staircase signal. Existing methods have limitations in decoupling sharp step edges and flat regions from noisy signals while ensuring the accuracy of step amplitude reconstruction. In this study, we have developed a novel generative denoising method, named variational denoising autoencoder (VDAE), based on a unique deep learning-based Bayesian framework. Specifically, the encoder-parameterized approximate posterior maps the distribution of essential features (i.e., thrust step amplitudes) of limited training samples to a latent space with a Gaussian distribution. This distribution transformation gives the latent space the ability to describe complete continuous step amplitudes. VDAE inherits the excellent generalization ability of the generative model and greatly improves the amplitude accuracy of the denoised signals. In addition, considering the different scale features in the clean staircase signal, a trend feature disentangler (TFD) is introduced in the encoder. The TFD adaptively extracts ultra-high-frequency sharp step edge features and ultra-low-frequency flat region features. Furthermore, to address the issue of recovering sharp step edges, total variation (TV) sparse representation is introduced into the loss function, guiding the decoder to reconstruct the thrust step. Extensive simulations and experiments were carried out to demonstrate the effectiveness and superiority of the proposed method in micronewton thrust step reconstruction and measurement noise removal.
- 24 September 2017
- Published by Springer Nature
- p. 177-186
A Prototype System for Three-Dimensional Cardiac Modeling and Printing with Clinical Computed Tomography Angiography
- 24 September 2017
- Published by Springer Nature
- p. 177-186
A Prototype System for Three-Dimensional Cardiac Modeling and Printing with Clinical Computed Tomography Angiography