Yangyang Li

Dr. Yangyang Li


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

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

Representative Publications

View All Publications
13 May 2022 in Energies
Verified
Journal Article
Open access
PDF file
AZAo Zhou
Ao Zhou
HJHui Jin
Hui Jin
WCWenhan Cao
Wenhan Cao
MPMing Pang
Ming Pang
YLYangyang Li
Yangyang Li
CZChao Zhu
Chao Zhu
An experimental study regarding methanol–diesel dual-fuel (DF) engines was conducted on a modified engine to explore the effects of pilot injection timing and period on the two-stage combustion process caused by the pilot injection strategy. In this study, the two-stage combustion process was determined according to the first two peaks of the second derivative of an in-cylinder pressure (d2p/dφ2) curve. The results show that the peak pressure rise rate (PRR) tended to decrease with advancing pilot injection timing at a high co-combustion ratio (CCR), which reduced combustion noise. The start of the combustion of the main injection diesel (SOC2) could be advanced by increasing the pilot injection period or advancing pilot injection timing at a 42% CCR. At an 18% CCR, the pilot injection timing and period had no significant effect on SOC2. With the advancement of pilot injection timing, the start of the combustion of pilot injection diesel (SOC1) advanced, and generally, the coefficient of variation of the PRR (COVPRR) of the two-stage combustion process increased first and then decreased. However, with the increase in the pilot injection period, SOC1 almost always remained constant and the COVPRR of the two-stage combustion process generally increased.

1 January 2023 in Energy Engineering
Verified
Journal Article
Open access
PDF file
CZChao Zhu
Chao Zhu
ZLZhuopei Liu
Zhuopei Liu
HCHao Chen
Hao Chen
YLYangyang Li
Yangyang Li
Experiments were conducted on a diesel-methanol dual-fuel (DMDF) engine modified by a six-cylinder, turbo-charged, inter-cooled diesel engine. According to the number of diesel injection, the experiments are divided to two parts: the single injection mode and double injection mode. The results show that, at the double injection mode, the maximum of pressure rise rate is small and the engine runs smoothly, however, knock still occurs when the co-combustion ratio (CCR) is big enough. Under knock status, the power density of the block vibration concentrating at some special frequencies rises dramatically, and the special frequency of single injection mode (about 4.1 kHz) is lower than that of double injection mode (7–9 kHz). The cylinder pressure oscillations of knock status are very different from the non-knock status. Under knock status, cylinder pressure oscillations become more concentrated and fiercer at some special frequencies, and the same as the block vibration. The special frequency of single injection mode (3–6 kHz) is lower than that of double injection mode (above 9 kHz).

20 May 2026 in ACS Omega
Verified
Journal Article
Open access
HJHui Jin
Hui Jin
AZAo Zhou
Ao Zhou
HZHui Zhu
Hui Zhu
YLYangyang Li
Yangyang Li
YWYujia Wang
Yujia Wang
SLSongfeng Li
Songfeng Li
The development of green and low-carbon new energy is one of the effective ways to deal with environmental pollution and climate change. Methanol, as an electrocatalytic product, is an excellent energy source and fuel. However, liquid methanol poses issues such as volatility, flammability, and explosiveness during use and long-term storage. This study presented a new methanol–sodium stearate–water-mixed gel. The performance of this mixed gel as a methanol energy storage carrier was evaluated. The results indicated that the gel points of the mixed gel increased with the increased contents of water and gelling agent. The maximum gel point reach approximately 35 °C, which is beneficial for reducing the requirements for the storage temperature. However, an excessive content of the gelling agent would lead to a decrease in the calorific value. The calorific value of the mixed gel containing 10% water is comparable to that of the mixed gel containing 5% gelling agent. Therefore, the scheme of adding water to increase the gel point has a better effect. Additionally, as the content of the gelling agent increased, the three-dimensional network structure formed by the interwoven fibers within the mixed gel became more compact. Meanwhile, the precipitation phenomenon of the gelling agent occurred during the gelation process, resulting in the lower structure of the mixed gel being denser than the upper structure. Therefore, in the large-scale production process of the mixed gel, the gel preparation container should be selected to be flat in shape to improve the uniformity.

8 June 2026 in Journal of Vibroengineering
Verified
Journal Article
Open access
PDF file
ECE Cai
E Cai
YLYangyang Li
Yangyang Li
Domain shift caused by variations in rotational speed and load under cross-condition scenarios may distort the statistical distribution and feature representation of vibration signals, posing a challenge to the reliable deployment of intelligent fault diagnosis systems for rotating machinery. Existing multi-scale CNN–Transformer hybrid architectures generally do not explicitly consider the influence of such condition-induced domain shifts. To address this issue, a multi-scale CNN-Transformer hybrid method integrating a Parallel Attention Mechanism (PAM) and a Local Linear Unit (LLU) is proposed to enhance the stability of feature representations under varying operating conditions. In the proposed method, raw vibration signals are directly used as inputs, and a multi-scale CNN module is employed to capture transient impact features across multiple temporal scales. The PAM then adjusts feature responses in both channel and temporal dimensions through parallel attention branches, enabling adaptive feature reweighting to alleviate condition-induced statistical variations. Furthermore, a Transformer encoder embedded with LLU is adopted as the backbone to incorporate local structural modeling through depthwise separable convolution while preserving the global dependency modeling capability of self-attention. Experiments on three benchmark datasets (PU, PHM09, and CWRU) show that the proposed method achieves average accuracies of 80.15 %, 93.96 %, and 98.17 %, respectively, under unseen operating conditions, consistently outperforming several representative comparative methods. Additional ablation studies further verify the contribution of PAM and LLU to robust feature representation learning. Moreover, noise robustness experiments under multiple signal-to-noise ratio (SNR) levels demonstrate that the proposed method maintains stable performance under moderate noise conditions, highlighting its practical reliability in realistic industrial environments. Our code is available at https://github.com/caie8201/Multi-Scale-CNN-Transformer-Hybrid-Network.