Sihao Zhang

dblp:304/7311 · DBLP profile ↗
← Back
14ranked-venue papers
1as first author
14since 2021 · last 2026
0009-0006-7351-9550ORCID · corroborated

Domains — the database's venue-derived domains; a paper can count in several

Artificial intelligence and machine learning · 5 · 5 since 2021Systems, architecture and hardware · 4 · 1 first-author · 4 since 2021Databases, data management, data science and information retrieval · 4 · 4 since 2021Graphics, computer vision, multimedia, augmented reality and games · 1 · 1 since 2021
YearPublicationVenuePosition
2026 Gene recombination-guided convolution neural network for early fault diagnosis of aero-engines
Jinlei Wu, Lin Lin 0014, Song Fu, Lingyu Yue, Sihao Zhang
Adv. Eng. Informatics6
2026 GIRMSF-global information reconstruction and multi-scale feature sharpening framework for knowledge graph embedding
Lin Lin 0014, Shiwei Suo, Song Fu, Lizheng Zu, Sihao Zhang
Expert Syst. Appl.5
2026 AIR-YOLO: Enhanced small-object detection for transmission line inspection in complex backgrounds
Lihui Lu, Sihao Zhang, Junlu Jiang, Yuke Gu
Multim. Syst.3
2025 A 49.7-fs, 10.29-to-12.75 GHz Fractional-N ADPLL with a New Quad-Core Class-F3 Oscillator
abstract
This paper presents a fractional-N all digital phase locked loop (ADPLL) based on a new quad-core class-F3oscillator, which reduces the out-of-band phase noise of ADPLL. The LC tank of the proposed oscillator incorporates both parallel and series resonant cavities compactly. It increases an additional impedance peak around the third-harmonic of the fundamental oscillation voltage, thus enforcing a pseudo-square voltage waveform around the LC tank. Consequently, the effective impulse sensitivity function (ISF) decreases, thereby reducing the phase noise. Furthermore, the proposed topology demonstrates low phase noise penalty during multi-core resonance. The ADPLL is implemented in a 40nm CMOS process. The simulated phase noise of the quad-core class-F3oscillator at 1MHz offset varies from -124.5 to -126.1dBc/Hz depending on the output frequency, which ranges from 10.29 to 12.75GHz. The oscillator consumes 24.36mW from the supply and exhibits a figure-of-merit (FoM) of 192.5dBc/Hz. Ultimately, the ADPLL utilizing the proposed oscillator exhibits an excellent RMS jitter of 49.7fs and achieves a FoM of -250.5dB.
Sihao Zhang, Ningyuan Zhang, Chuancheng Wu, Ling Hao, Haoyu Bai, Huailin Liao
ISCAS1
2025 Reconstruction method of aircraft wing stress field under limited measurement points via multi-source heterogeneous information fusion
abstract
Due to the aircraft wing’s topological structure and lightweight design requirements, strain sensors installed on the wing are very limited. Traditional methods, relying on limited sensor data as a single information source, are insufficient for full-stress field monitoring, leading to a high prediction error. To address this issue, a novel wing stress field reconstruction method with limited measurement points is developed via multi-source heterogeneous information fusion. To be specific, two information fusion modules are designed to jointly overcome the challenges of limited measurement data and high non-linearity during full-stress field reconstruction. On one hand, the finite element mechanism-based information fusion module (FEMIFM) is proposed to derive and establish a mechanical model that relates the wing stress to positional parameter, in order to introduce physical information and reduce the non-linearity of the reconstruction mapping. On the other hand, the simulation stress expectation-based information fusion module (SSEIFM) leverages stress expectations derived from simulated stress fields under various operating conditions to incorporate statistical information, thereby enhancing the robustness and reasonableness of reconstruction results. Moreover, a soft-threshold loss function is proposed, which ignores zero-drift errors of strain sensors, improving the reconstruction accuracy of critical stress points. Finally, the developed method can be seamlessly integrated with popular neural networks (i.e., Transformer, convolutional neural networks, multilayer perceptron, etc.). Extensive experiments are conducted to validate the effectiveness of the developed method on an actual aircraft wing stress dataset.
Lin Lin 0014, Lingyu Yue, Jinlei Wu, Sihao Zhang, Shiwei Suo
Adv. Eng. Informatics5
2025 FD-LLM: Large language model for fault diagnosis of complex equipment
Sihao Zhang, Song Fu
Adv. Eng. Informatics2
2025 PSTFormer: A novel parallel spatial-temporal transformer for remaining useful life prediction of aeroengine
Song Fu, Yiming Jia, Lin Lin 0014, Shiwei Suo, Sihao Zhang
Expert Syst. Appl.6
2025 Prototype matching-based meta-learning model for few-shot fault diagnosis of mechanical system
Lin Lin 0014, Sihao Zhang, Song Fu, Shiwei Suo, Guolei Hu
Neurocomputing2
2024 A 0.12mm2 K/Ka Band RX Front-end in 40-nm CMOS with Inductor-Less LO Generators
abstract
This paper presents a broadband 18-33GHz receiver (RX) front-end with a core area of only 0.12mm2. In the local oscillator (LO) generators, a compact edge-combining (EC)-based frequency multiplication scheme is proposed to generate wideband and precise in-phase and quadrature (I/Q) LO signals at millimeter-wave frequencies. In the output stage of the LO chain, the switched-capacitor (SC) XOR modules triple the output clock frequency through edge-combining, which provides a 3× frequency reduction for the global clock distribution. The clock tripling is independently performed in I/Q differential LO signals, and the orthogonality of LO signals is not influenced by frequency multiplication. Implemented in TSMC 40nm GP CMOS, the proposed receiver core occupies only 0.46mm × 0.26mm. The RX front-end achieves a conversion gain of 42dB, a Noise Figure of 5.6dB, a baseband bandwidth of 100MHz, and an IQ phase variation < 1.3° (1σ interval, without calibration), while operating in the K/Ka band at 18-33GHz.
Haoyu Bai, Sihao Zhang, Huailin Liao
ISCAS3
2024 A 136μW Over 800m Range Backscatter-Like UHF Band Transceiver
abstract
Backscatter transceivers are often used in ultra-low-power Internet of Things (IoT) applications. However traditional backscatter transceiver (TRX) cannot actively control the transmitted power, which greatly limits the communication distance. This paper introduces a novel backscatter TRX operating in the ultra-high frequency (UHF) band, with a low power consumption of only 136μW and an impressive communication range surpassing 800 meters. Unlike the backscatter, the received continuous wave (CW) is not reflected via impedance modulating at the antenna but is directed into the TRX to be modulated and amplified. This technique significantly extends the communication range at a low power consumption. A coupler serves to prevent the direct entry of transmitted signals into the receiving path, ensuring the received CW remains unaffected to be modulated. Implemented in TSMC 40nm CMOS, the proposed TRX core occupies 0.7 mm2and works under a 0.7V supply voltage with the -54.6dBm input power and -20dBm output power.
Ling Hao, Keer Gao, Haoyu Bai, Chuancheng Wu, Sihao Zhang, Jiazheng Zhou, Huailin Liao
ISCAS6
2024 A low in-band phase noise Fractional-N ADPLL based on Switched-Capacitor-DPI
abstract
This paper presents a fractional-N all digital phase locked loop (ADPLL) based on a low in-band phase noise switched-capacitor digital phase interpolator (SC-DPI), which reduced the in-band phase noise of ADPLL. The SC-DPI utilized a capacitance vector-sum structure with a short signal chain to improve the in-band phase noise performance. Designed in 40-nm CMOS technology, from simulation results, the proposed SC-DPI worked at 1GHz and achieved -126.2 dBc/Hz in-band phase noise at 1kHz and -142.6 dBc/Hz at 1MHz. The proposed 8GHz ADPLL based on SC-DPI achieved -110.5 dBc/Hz and -117.2 dBc/Hz in-band phase noise at 1kHz and 10kHz respectively.
Ningyuan Zhang, Sihao Zhang, Huailin Liao
ISCAS2
2024 Channel attention & temporal attention based temporal convolutional network: A dual attention framework for remaining useful life prediction of the aircraft engines
Lin Lin 0014, Jinlei Wu, Song Fu, Sihao Zhang, Changsheng Tong, Lizheng Zu
Adv. Eng. Informatics4
2024 Integrating adversarial training strategies into deep autoencoders: A novel aeroengine anomaly detection framework
Lin Lin 0014, Lizheng Zu, Song Fu, Sihao Zhang, Shiwei Suo, Changsheng Tong
Eng. Appl. Artif. Intell.5
2024 An adaptive hybrid surrogate model for FEA of telescopic boom of rock drilling jumbo
Yancheng Lv, Changsheng Tong, Sihao Zhang, Shiwei Suo
Eng. Appl. Artif. Intell.6