Chuntao Rao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Chuntao Rao has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Minan Tang
Lanzhou Jiaotong University
7 shared publications - Zhongcheng Bai
School of Automation and Electrical Engineering, Lanzhou Jiaotong University Lanzhou China
5 shared publications - Yude Jiang
School of Automation and Electrical Engineering, Lanzhou Jiaotong University Lanzhou China
4 shared publications - Changyou Wang
College of Automation and Electrical Engineering, Lanzhou Jiaotong University , Lanzhou 730070, Gansu
3 shared publications - Wenxin Sheng
School of Automation and Electrical Engineering, Lanzhou Jiaotong University Lanzhou China
3 shared publications - Zhanglong Tao
Lanzhou Jiaotong University
3 shared publications
Latest Publications
- Open Access
- 24 August 2026
- Published by Elsevier in Energy Reports
- Vol. 16
Urban electric vehicle charging station siting method considering road impedance based on improved WOA
- 24 August 2026
- Published by Elsevier in Energy Reports
- Vol. 16
Open AccessUrban electric vehicle charging station siting method considering road impedance based on improved WOA
- 19 February 2026
- Published by Elsevier in Journal of the Franklin Institute
- Vol. 363 (6)
Dual closed-loop tracking control for wheeled mobile robot via HMPC and nonlinear ESO-TD integral sliding mode control
- 19 February 2026
- Published by Elsevier in Journal of the Franklin Institute
- Vol. 363 (6)
Dual closed-loop tracking control for wheeled mobile robot via HMPC and nonlinear ESO-TD integral sliding mode control
- 2 January 2026
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 104 (7) , 3527-3545
Model predictive control of coal‐fired power plant desulphurization system based on GJO ‐ TCN ‐ BiGRU ‐attention
The wet flue gas desulphurization (WFGD) system is characterized by nonlinearity, strong coupling, time‐varying, and input–output delays, which make it difficult for traditional control methods to accurately control the WFGD system. In this paper, a combination of neural network (NN) predictive model and model predictive control (MPC) is proposed to be used for the optimal control of the WFGD system. Firstly, the prediction model of the WFGD system is established by temporal convolutional network (TCN)‐bidirectional gated recurrent unit (BiGRU)‐attention. Secondly, the hyperparameters of the prediction model are optimized using the golden jackal optimization (GJO). Then, the WFGD system is controlled by MPC. Meanwhile, in this paper, the original dataset of the WFGD system is screened to remove outliers using a local outlier factor (LOF) detection algorithm. Finally, based on the Matlab R2023a platform, this method is compared with the PID algorithm, and the actual effect is improved by about 28%. The experimental results show that the GJO‐TCN‐BiGRU‐attention model can predict the desulphurization system well, and the control effect of the MPC method on the desulphurization system is better than PID algorithm.
- 2 January 2026
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 104 (7) , 3527-3545
Model predictive control of coal‐fired power plant desulphurization system based on GJO ‐ TCN ‐ BiGRU ‐attention
The wet flue gas desulphurization (WFGD) system is characterized by nonlinearity, strong coupling, time‐varying, and input–output delays, which make it difficult for traditional control methods to accurately control the WFGD system. In this paper, a combination of neural network (NN) predictive model and model predictive control (MPC) is proposed to be used for the optimal control of the WFGD system. Firstly, the prediction model of the WFGD system is established by temporal convolutional network (TCN)‐bidirectional gated recurrent unit (BiGRU)‐attention. Secondly, the hyperparameters of the prediction model are optimized using the golden jackal optimization (GJO). Then, the WFGD system is controlled by MPC. Meanwhile, in this paper, the original dataset of the WFGD system is screened to remove outliers using a local outlier factor (LOF) detection algorithm. Finally, based on the Matlab R2023a platform, this method is compared with the PID algorithm, and the actual effect is improved by about 28%. The experimental results show that the GJO‐TCN‐BiGRU‐attention model can predict the desulphurization system well, and the control effect of the MPC method on the desulphurization system is better than PID algorithm.
- Open Access
- 1 January 2026
- Published by SAGE Publications in Science Progress
- Vol. 109 (1)
Wind turbine bearing fault classification identification based on optimized variational mode decomposition and convolutional neural network–bidirectional gated recurrent unit–Attention
As a key core component of wind turbine generators, the rolling bearings in the gearbox directly affect the overall performance and reliability of the wind turbine generators. Accurate prediction and timely diagnosis can effectively improve the efficiency of the wind turbine generators. This paper takes the rolling bearing operation data as the research object and proposes a bearing fault classification research method based on the combination of variational mode decomposition (VMD) optimization and convolutional neural network-bidirectional gated recurrent unit (CNN-BiGRU)-Attention model. Firstly, to address the sensitivity of intrinsic mode function (IMF) components in the VMD decomposition process, an improved RIEM algorithm is adopted to optimize the hyperparameters of the VMD algorithm. This process aims to adaptively adjust the penalty factor and decomposition layers of the VMD algorithm and find the optimal IMF component to determine the most suitable IMF component in the signal data. Secondly, to fully explore the complex characteristics of fault signals, composite multi-scale slope entropy is used to extract features from the optimized input data. By conducting multidimensional analysis on the local and global characteristics of the signal at different time scales, efficient representation of fault features is achieved. Finally, based on MATLAB, a simulation experiment platform is established. This paper conducts research on the classification of rolling bearing faults through the CNN-BiGRU-Attention model. The results show that the model established in this paper has significant effects and stable performance. The research in this paper provides new technical ideas for fault diagnosis of rolling bearings in wind turbine generator gear.
- 1 January 2026
- Published by SAGE Publications in Science Progress
- Vol. 109 (1)
Open AccessWind turbine bearing fault classification identification based on optimized variational mode decomposition and convolutional neural network–bidirectional gated recurrent unit–Attention
As a key core component of wind turbine generators, the rolling bearings in the gearbox directly affect the overall performance and reliability of the wind turbine generators. Accurate prediction and timely diagnosis can effectively improve the efficiency of the wind turbine generators. This paper takes the rolling bearing operation data as the research object and proposes a bearing fault classification research method based on the combination of variational mode decomposition (VMD) optimization and convolutional neural network-bidirectional gated recurrent unit (CNN-BiGRU)-Attention model. Firstly, to address the sensitivity of intrinsic mode function (IMF) components in the VMD decomposition process, an improved RIEM algorithm is adopted to optimize the hyperparameters of the VMD algorithm. This process aims to adaptively adjust the penalty factor and decomposition layers of the VMD algorithm and find the optimal IMF component to determine the most suitable IMF component in the signal data. Secondly, to fully explore the complex characteristics of fault signals, composite multi-scale slope entropy is used to extract features from the optimized input data. By conducting multidimensional analysis on the local and global characteristics of the signal at different time scales, efficient representation of fault features is achieved. Finally, based on MATLAB, a simulation experiment platform is established. This paper conducts research on the classification of rolling bearing faults through the CNN-BiGRU-Attention model. The results show that the model established in this paper has significant effects and stable performance. The research in this paper provides new technical ideas for fault diagnosis of rolling bearings in wind turbine generator gear.
- 26 August 2025
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 104 (3) , 1314-1333
Optimal control of carbon capture system in coal‐fired power plant based on fusion of CNN ‐ LSTM and EMPC for MOGWO optimization
Coal‐fired power plants remain a major source of global CO 2 emissions, posing urgent challenges for climate mitigation. This study proposes an AI‐enabled optimal control strategy for solvent‐based post‐combustion CO 2 capture (PCC), aiming to reduce energy consumption and enhance environmental performance. A deep learning model combining convolutional neural networks and long short‐term memory (CNN‐LSTM) is developed to predict multi‐step CO 2 capture rates. This model is integrated with a multi‐objective grey wolf optimizer (MOGWO)‐based economic model predictive controller (EMPC), which dynamically adjusts key operating variables. Simulation results demonstrate that the proposed control scheme achieves a 12.4% reduction in capture cost and a 10.6% decrease in energy consumption, while maintaining a high CO 2 capture efficiency of 95.3%. These improvements not only optimize process economics while significantly lowering the environmental impact of carbon capture systems. These improvements contribute directly to carbon mitigation, offering a promising pathway for cleaner and more sustainable operation of coal‐based power generation under dynamic industrial conditions.
- 26 August 2025
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 104 (3) , 1314-1333
Optimal control of carbon capture system in coal‐fired power plant based on fusion of CNN ‐ LSTM and EMPC for MOGWO optimization
Coal‐fired power plants remain a major source of global CO 2 emissions, posing urgent challenges for climate mitigation. This study proposes an AI‐enabled optimal control strategy for solvent‐based post‐combustion CO 2 capture (PCC), aiming to reduce energy consumption and enhance environmental performance. A deep learning model combining convolutional neural networks and long short‐term memory (CNN‐LSTM) is developed to predict multi‐step CO 2 capture rates. This model is integrated with a multi‐objective grey wolf optimizer (MOGWO)‐based economic model predictive controller (EMPC), which dynamically adjusts key operating variables. Simulation results demonstrate that the proposed control scheme achieves a 12.4% reduction in capture cost and a 10.6% decrease in energy consumption, while maintaining a high CO 2 capture efficiency of 95.3%. These improvements not only optimize process economics while significantly lowering the environmental impact of carbon capture systems. These improvements contribute directly to carbon mitigation, offering a promising pathway for cleaner and more sustainable operation of coal‐based power generation under dynamic industrial conditions.
- 19 March 2025
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 103 (10) , 4904-4924
Prediction control of CO2 capture in coal‐fired power plants based on ERIME‐optimized CNN‐LSTM‐multi‐head‐attention
Predicting CO2 concentration in post‐combustion carbon capture (PCC) systems is challenging due to complex operating conditions and multivariate interactions. This study proposes an enhanced RIME algorithm (ERIME) optimization‐based convolutional neural network (CNN)‐long short‐term memory (LSTM)‐multi‐head‐attention (ECLMA) model to improve prediction accuracy. The local outlier factor (LOF) algorithm was used to remove noise from the data, while mutual information (MI) determined time lags, and the smoothed clipped absolute deviation (SCAD) method optimized feature selection. The CNN‐LSTM‐multi‐head‐attention model extracts meaningful features from time series data, and parameters are optimized using the ERIME algorithm. Using a simulated dataset from a 600 MW supercritical coal‐fired power plant, the results showed that after LOF outlier removal, root mean square error (RMSE) and mean absolute error (MAE) improved by 10%–13%. Post‐MI delay reconstruction reduced RMSE to 0.00999 and MAE to 11.6937, with R2 rising to 0.9929. After variable selection, RMSE and MAE further reduced to 0.00907 and 9.9697, with R2 increasing to 0.9983. After ERIME optimization, the ECLMA model outperformed traditional models, reducing RMSE and MAE by up to 91.55% and 84.94%, respectively, compared to CNN, and by 85.91% and 69.47%, respectively, compared to LSTM. These results confirm the model's superior accuracy and stability.
- 19 March 2025
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 103 (10) , 4904-4924
Prediction control of CO2 capture in coal‐fired power plants based on ERIME‐optimized CNN‐LSTM‐multi‐head‐attention
Predicting CO2 concentration in post‐combustion carbon capture (PCC) systems is challenging due to complex operating conditions and multivariate interactions. This study proposes an enhanced RIME algorithm (ERIME) optimization‐based convolutional neural network (CNN)‐long short‐term memory (LSTM)‐multi‐head‐attention (ECLMA) model to improve prediction accuracy. The local outlier factor (LOF) algorithm was used to remove noise from the data, while mutual information (MI) determined time lags, and the smoothed clipped absolute deviation (SCAD) method optimized feature selection. The CNN‐LSTM‐multi‐head‐attention model extracts meaningful features from time series data, and parameters are optimized using the ERIME algorithm. Using a simulated dataset from a 600 MW supercritical coal‐fired power plant, the results showed that after LOF outlier removal, root mean square error (RMSE) and mean absolute error (MAE) improved by 10%–13%. Post‐MI delay reconstruction reduced RMSE to 0.00999 and MAE to 11.6937, with R2 rising to 0.9929. After variable selection, RMSE and MAE further reduced to 0.00907 and 9.9697, with R2 increasing to 0.9983. After ERIME optimization, the ECLMA model outperformed traditional models, reducing RMSE and MAE by up to 91.55% and 84.94%, respectively, compared to CNN, and by 85.91% and 69.47%, respectively, compared to LSTM. These results confirm the model's superior accuracy and stability.
- 18 November 2024
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 103 (7) , 3153-3170
Prediction of SO2 concentration in WFGD system based on GWO optimized CNN‐BiLSTM‐attention
Due to external disturbances, the parameters of the desulphurization system are uncertain, and their relationships are complex, which makes it difficult to predict the concentration of SO2 at the desulphurization system outlet. In this paper, grey wolf optimization (GWO) optimized convolutional neural network (CNN)‐bi‐directional long short‐term memory (BiLSTM)‐Attention algorithm was used for prediction, and the problem of low SO2 concentration prediction accuracy at outlet has been resolved. First, the outliers of the thermal power plant desulphurization data were processed using the local outlier factor (LOF) algorithm. Secondly, CNN‐BiLSTM model was constructed using CNN and BiLSTM, and attention module was added to realize feature extraction and better capture the regularity of input data. Then, the CNN‐BiLSTM‐Attention model was optimized using GWO and its hyperparameters were improved. Finally, based on the Matlab R2023a platform, the prediction comparison as well as the error analysis of the desulphurization data were carried out. In the prediction of SO2 concentration in low‐flow continuous slurry supply mode, the error of the combined model decreased by 23.2% on average compared to the CNN‐BiLSTM‐Attention model. In the prediction of SO2 concentration in the high‐flow intermittent slurry supply mode, the error of the combined model decreased by 20.8% on average. According to the results, the combined model surpasses both the single model and several other algorithmic combination models in terms of performance metrics, and the predictions are more accurate.
- 18 November 2024
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 103 (7) , 3153-3170
Prediction of SO2 concentration in WFGD system based on GWO optimized CNN‐BiLSTM‐attention
Due to external disturbances, the parameters of the desulphurization system are uncertain, and their relationships are complex, which makes it difficult to predict the concentration of SO2 at the desulphurization system outlet. In this paper, grey wolf optimization (GWO) optimized convolutional neural network (CNN)‐bi‐directional long short‐term memory (BiLSTM)‐Attention algorithm was used for prediction, and the problem of low SO2 concentration prediction accuracy at outlet has been resolved. First, the outliers of the thermal power plant desulphurization data were processed using the local outlier factor (LOF) algorithm. Secondly, CNN‐BiLSTM model was constructed using CNN and BiLSTM, and attention module was added to realize feature extraction and better capture the regularity of input data. Then, the CNN‐BiLSTM‐Attention model was optimized using GWO and its hyperparameters were improved. Finally, based on the Matlab R2023a platform, the prediction comparison as well as the error analysis of the desulphurization data were carried out. In the prediction of SO2 concentration in low‐flow continuous slurry supply mode, the error of the combined model decreased by 23.2% on average compared to the CNN‐BiLSTM‐Attention model. In the prediction of SO2 concentration in the high‐flow intermittent slurry supply mode, the error of the combined model decreased by 20.8% on average. According to the results, the combined model surpasses both the single model and several other algorithmic combination models in terms of performance metrics, and the predictions are more accurate.