YU Jun-qi

YU Jun-qi


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
31 March 2023 in Electronics
Verified
Journal Article
Open access
PDF file
XDXisheng Ding
Xisheng Ding
YJYU Jun-qi
YU Jun-qi
A convolutional neural network (CNN) was designed and built on an embedded building lighting control system to determine whether the application of CNN could increase the accuracy of image recognition and reduce energy consumption. Currently, lighting control systems rely mainly on information technology, with sensors to detect people’s existence or absence in an environment. However, due to the deviation of this perception, the accuracy of image detection is not high. In order to validate the effectiveness of the new system based on CNN, an experiment was designed and operated. The importance of the research lies in the fact that high image detection would bring in less energy consumption. The result of the experiment indicated that, when comparing the actual position with the positioning position, the difference was between 0.01 to 0.20 m, indicating that the image recognition accuracy of the CNN-based embedded control system was very high. Moreover, comparing the luminous flux of the designed system with natural light and the designed system without natural light with the system without intelligent control, the energy savings is about 40%.

5 June 2025 in Engineering Research Express
Verified
Journal Article
Closed access
HLHouxin Liu
Houxin Liu
YJYU Jun-qi
YU Jun-qi
MWMeng Wang
Meng Wang
WCWenqiang Cao
Wenqiang Cao
HLHaiyan Liu
Haiyan Liu
In integrated energy systems (IES), the deep coupling of heterogeneous energy flows results in complex interdependencies between energy sources and loads. However, current optimization strategies for IES scheduling often overlook the dynamic coupling relationships between sources and loads, which poses significant challenges to system scheduling and planning. To address this issue, this study proposes a scenario generation method based on a CNN-GRU-Attention architecture. Convolutional Neural Networks (CNN) are employed to extract local features, Gated Recurrent Units (GRU) are used to model the temporal dynamics of time series data, and an Attention mechanism is incorporated to enhance the weighting of critical information—thereby capturing the source-load correlations more effectively. Furthermore, a reward-penalty tiered carbon emission mechanism is introduced, which incentivizes substantial emission reductions and penalizes excessive emissions through segmented carbon pricing. Based on this, a low-carbon economic scheduling model for an integrated electricity-heat energy system is constructed with the objective of minimizing both operational and carbon emission costs. Simulation studies based on a representative IES in Northwest China demonstrate that the proposed CNN-GRU-Attention model effectively learns features and patterns from historical data and captures source-load correlations with high accuracy. The implementation of the reward-penalty tiered carbon trading mechanism results in an 18.32% reduction in carbon emissions, albeit with a 7.67% increase in carbon costs. When source-load correlation is comprehensively considered, the total system cost decreases by 7.34%, and carbon emissions are further reduced by 18.83%. Seasonal analysis reveals reductions in operational costs of 15.64%, 8.36%, and 11.72% in winter, summer, and transition seasons, respectively, indicating strong adaptability across seasons.

24 April 2020 in Advances in Structural Engineering
Verified
Journal Article
Closed access
PDF file
XLXinhao Lin
Xinhao Lin
YXYazhou Xu
Yazhou Xu
HZHui Zhang
Hui Zhang
QRQianqian Ren
Qianqian Ren
YJYU Jun-qi
YU Jun-qi
For fretting fatigue, micro-slipping often causes initial micro-cracks in the contact surfaces, which gradually propagate and eventually result in fracture failure. However, it is difficult to directly observe and measure the crack initiation and propagation processes of single-lap bolted joints due to the obstacle of testing technique. Therefore, this article presents an elastic analysis–based method, in which the total fretting fatigue lifetime is divided into initiation life predicted by continuum damage mechanics and propagation life calculated by SEAM Tool in combination with Paris’s law. Since the initiation life model implicitly reflects the result of damage process, one can directly calculate the initiation life based on the elastic stress analysis. The predicted fretting fatigue life, initial crack location, and propagation path are in reasonable agreement with the fretting fatigue test and scanning electron microscope observation results. In addition, it is found that among the main factors affecting fretting fatigue of steel single-lap bolted joints, the initial life is affected by cyclic stress, friction coefficient, and tightening torque, and the propagation path is mainly influenced by tightening torque.

3 April 2019
Verified
Book Chapter
Closed access
PDF file
YJYU Jun-qi
YU Jun-qi
XYXiong Yang
Xiong Yang
AZAnjun Zhao
Anjun Zhao
MZMeng Zhou
Meng Zhou
YRYanhuan Ren
Yanhuan Ren


1 May 2026 in Engineering Research Express
Verified
Journal Article
Closed access
HWHaifeng Wang
Haifeng Wang
JWJiangtao Weng
Jiangtao Weng
YJYU Jun-qi
YU Jun-qi
WCWenqiang Cao
Wenqiang Cao
TWTongtong Weng
Tongtong Weng
CLChaoyang Li
Chaoyang Li
The growing integraion of photovoltaic (PV) systems in buildings poses significant challenges for energy management due to the intrinsic volatility and intermittency of PV output. Accurate multi-step PV forecasting is critical for supporting building energy management systems in scheduling, load optimization, and renewable energy dispatch. To tackle the dual challenges of modeling complex temporal dynamics and improving multi-step forecasting accuracy, this study introduces a channel-attention-based hybrid prediction framework, TCN-iTimesNet. The framework first leverages a temporal convolutional network (TCN) to capture long-range dependencies and extract stable global representations, which are subsequently processed by an improved TimesNet module for multi-scale periodic modeling. A convolutional kernel decoupling mechanism disentangles intra- and inter-period convolutions, enabling precise extraction of fine-grained features and long-range coupling patterns. In parallel, integration of a squeeze-and-excitation channel attention mechanism adaptively amplifies critical features, boosting both forecasting accuracy and model robustness. Experiments on a publicly available photovoltaic dataset from China indicate that the proposed model consistently outperforms mainstream baselines—including Transformer, TimesNet, CNN-LSTM, Informer, and DLinear—in predictive accuracy and long-horizon forecasting performance. Cross-dataset validation further demonstrates that TCN-iTimesNet maintains strong robustness and generalization across different PV datasets. The results validate the effectiveness of the proposed framework in multi-step photovoltaic power forecasting, providing a robust data foundation for building energy management and renewable energy dispatch planning.