Dengji Zhou

Dr. Dengji Zhou


Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

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

Representative Publications

View All Publications
1 January 2022 in Energy
Verified
Journal Article
Closed access
DZDengji Zhou
Dengji Zhou
SYSiyun Yan
Siyun Yan
DHDawen Huang
Dawen Huang
TSTiemin Shao
Tiemin Shao
WXWang Xiao
Wang Xiao
TYTianqi Yu
Tianqi Yu

1 November 2020 in Energy Reports
Verified
Journal Article
Open access
DZDengji Zhou
Dengji Zhou
SMShixi Ma
Shixi Ma
JHJiarui Hao
Jiarui Hao
DHDong Han
Dong Han
DHDawen Huang
Dawen Huang
TLTaotao Li
Taotao Li

28 April 2022 in Energy
Verified
Journal Article
Closed access
DZDengji Zhou
Dengji Zhou
XJXingyun Jia
Xingyun Jia
SMShixi Ma
Shixi Ma
TSTiemin Shao
Tiemin Shao
DHDawen Huang
Dawen Huang
TLTaotao Li
Taotao Li

1 October 2015 in Journal of Engineering for Gas Turbines and Power
Verified
Journal Article
Closed access
PDF file
DZDengji Zhou
Dengji Zhou
HZHuisheng Zhang
Huisheng Zhang
SWShilie Weng
Shilie Weng
As a crucial section of gas turbine maintenance decision-making process, to date, gas path fault diagnostic has gained a lot of attention. However, model-based diagnostic methods, like nonlinear gas path analysis (GPA) and genetic algorithms, need an accurate gas turbine model, and diagnostic methods without gas turbine model, like expert system, need a knowledge database. Both are difficult to gain. Thus, data-driven approach for gas path diagnosis, like artificial neural network, is increasingly attractive. Support vector machine (SVM), a novel computational learning method, seems to be a good choice for data-driven gas path fault diagnosis of gas turbine. In this paper, SVM is employed to diagnose a deteriorated gas turbine. The effect of sample number, kernel function, and monitoring parameters on diagnostic accuracy are studied, respectively. Additionally, the diagnostic result of SVM is compared to the result of artificial neural networks. The comparing result confirms that SVM has an obvious advantage over artificial neural networks method based on a small sample of data and can be employed to gas path fault diagnosis of gas turbine. In addition, SVM with radial basis kernel function is the best choice for gas turbine gas path fault diagnosis based on small sample.

3 March 2024 in Advanced Engineering Informatics
Verified
Journal Article
Closed access
DZDengji Zhou
Dengji Zhou
DHDawen Huang
Dawen Huang

15 March 2021 in Journal of Engineering for Gas Turbines and Power
Verified
Journal Article
Closed access
PDF file
DZDengji Zhou
Dengji Zhou
XJXingyun Jia
Xingyun Jia
JHJiarui Hao
Jiarui Hao
DWDuocai Wang
Duocai Wang
DHDawen Huang
Dawen Huang
TWTingting Wei
Tingting Wei
Frequent changes in operating conditions can result in the great loss of the service life of gas turbines that work at high speed, high pressure, and high temperature. To improve the control system of gas turbines is of great significance for extending the service life, boosting the dynamic performance, and reducing the maintenance cost. Due to good dynamic response characteristics, the present control methods are neither capable of tackling the nonlinearity of the system nor adaptive to the frequent variations of operating conditions. As a powerful learning paradigm, reinforcement learning (RL) can explore the operation environment and make decisions adaptively. A novel intelligent control framework for the gas turbine control system is constructed by coupling a RL agent with a dynamic simulation model and a damage estimation model. Compared with the proportion, integral, differential (PID) and fuzzy control, the proposed method achieves better performance in both extending life and depicting dynamic characteristics, and reduces the overall damage to as low as 0.01% in the loading process. Besides, the overshoot and the adjusting time of the novel approach are lower and shorter than those of PID and fuzzy control by more than 90% and about 14%, respectively, but it takes longer to accelerate. Finally, the effect of different types of RL agents and their hyperparameters are investigated. The results show that the deep deterministic policy gradient (DDPG) agent can achieve the best performance. Furthermore, in addition to extending the life and improving the dynamic performance, the controlling framework presented is recommended to construct an intelligent dynamic control system for achieving various purposes.