Mingyuan Zhu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mingyuan Zhu has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Chongshi Gu
College of Water Conservancy and Hydropower Engineering, Hohai University, Nanjing 210098, China
2 shared publications Fuqiang Zhou
Xinjiang Institute of Water Resources and Hydropower Research, Ürümqi, China
2 shared publications- Hao Gu
College of Water Conservancy and Hydropower Engineering, Hohai University, Nanjing 210098, China
2 shared publications - Jianhe Peng
Anhui and Huaihe River Institute of Hydraulic Research (Anhui Provincial Water Conservancy Engineering Quality Testing Center Station), Hefei 230088, China
2 shared publications - Sen Zheng
College of Water Conservancy and Hydropower Engineering, Hohai University, Nanjing 210098, China
2 shared publications - Yan Wu
Xinjiang Institute of Water Resources and Hydropower Research, Urumqi 830049, China
2 shared publications
Latest Publications
- Open Access
- 8 September 2026
- Published by Springer Nature in Scientific Reports
A surrogate-driven optimization framework for thermal parameter inversion in hydraulic concrete structures
- 8 September 2026
- Published by Springer Nature in Scientific Reports
Open AccessA surrogate-driven optimization framework for thermal parameter inversion in hydraulic concrete structures
- Open Access
A Statistical Prediction Model for Sluice Seepage Based on MHHO-BiLSTM
The current seepage prediction model of the sluice gate is rarely used. To solve the problem, this paper selects the bidirectional long and short-term neural network (BiLSTM) with high information integration and accuracy, which can well understand and capture the temporal pattern and dependency relationship in the sequence and uses the multi-strategy improved Harris Hawks optimization algorithm (MHHO) to analyze its two hyperparameters: By optimizing the number of forward and backward neurons, the overfitting and long-term dependence problems of the neural network are solved, and the convergence rate is accelerated. Based on this, the MHHO-BiLSTM statistical prediction model of sluice seepage is established in this paper. To begin with, the prediction model uses water pressure, rainfall, and aging effects as input data. Afterward, the bidirectional long short-term memory neural network parameters are optimized using the multi-strategy improved Harris Hawks optimization algorithm. Then, the statistical prediction model based on the optimization algorithm proposed in this paper for sluice seepage is proposed. Finally, the seepage data of a sluice and its influencing factors are used for empirical analysis. The calculation and analysis results indicate that the optimization algorithm proposed in this paper can better search the optimal parameters of the bidirectional long short-term memory neural network compared with the original Harris Eagle optimization algorithm, optimizing the bidirectional long short-term memory neural network (HHO-BiLSTM) and the original bidirectional long short-term memory neural network (BiLSTM). Meanwhile, the bidirectional long and short-term neural network (BiLSTM) model shows higher prediction accuracy and robustness.
Open AccessA Statistical Prediction Model for Sluice Seepage Based on MHHO-BiLSTM
The current seepage prediction model of the sluice gate is rarely used. To solve the problem, this paper selects the bidirectional long and short-term neural network (BiLSTM) with high information integration and accuracy, which can well understand and capture the temporal pattern and dependency relationship in the sequence and uses the multi-strategy improved Harris Hawks optimization algorithm (MHHO) to analyze its two hyperparameters: By optimizing the number of forward and backward neurons, the overfitting and long-term dependence problems of the neural network are solved, and the convergence rate is accelerated. Based on this, the MHHO-BiLSTM statistical prediction model of sluice seepage is established in this paper. To begin with, the prediction model uses water pressure, rainfall, and aging effects as input data. Afterward, the bidirectional long short-term memory neural network parameters are optimized using the multi-strategy improved Harris Hawks optimization algorithm. Then, the statistical prediction model based on the optimization algorithm proposed in this paper for sluice seepage is proposed. Finally, the seepage data of a sluice and its influencing factors are used for empirical analysis. The calculation and analysis results indicate that the optimization algorithm proposed in this paper can better search the optimal parameters of the bidirectional long short-term memory neural network compared with the original Harris Eagle optimization algorithm, optimizing the bidirectional long short-term memory neural network (HHO-BiLSTM) and the original bidirectional long short-term memory neural network (BiLSTM). Meanwhile, the bidirectional long and short-term neural network (BiLSTM) model shows higher prediction accuracy and robustness.
- Open Access
Prediction for the Sluice Deformation Based on SOA-LSTM-Weighted Markov Model
Increasingly, deformation prediction has become an essential research topic in sluice safety control, which requires significant attention. However, there is still a lack of practical and efficient prediction modeling for sluice deformation. In order to address the limitations in mining the deep features of long-time data series of the traditional statistical model, in this paper, an improved long short-term memory (LSTM) model and weighted Markov model are introduced to predict sluice deformation. In the method, the seagull optimization algorithm (SOA) is utilized to optimize the hyper-parameters of the neural network structure in LSTM primarily to improve the model. Subsequently, the relevant error sequences of the fitting results of SOA-LSTM model are classified and the Markovity of the state sequence is examined. Then, the autocorrelation coefficients and weights of each order are calculated and the weighted and maximum probability values are applied to predict the future random state of the sluice deformation. Afterwards, the prediction model of sluice deformation on the SOA-LSTM-weighted Markov model is proposed. Ultimately, the presented model is used to predict the settlement characteristics of an actual sluice project in China. The analysis results demonstrate that the proposed model possesses the highest values of R2 and the smallest values of RMSE and absolute relative errors for the monitoring data of four monitoring points. Consequently, it concluded that the proposed method shows better prediction ability and accuracy than the SOA-LSTM model and the stepwise regression model.
Open AccessPrediction for the Sluice Deformation Based on SOA-LSTM-Weighted Markov Model
Increasingly, deformation prediction has become an essential research topic in sluice safety control, which requires significant attention. However, there is still a lack of practical and efficient prediction modeling for sluice deformation. In order to address the limitations in mining the deep features of long-time data series of the traditional statistical model, in this paper, an improved long short-term memory (LSTM) model and weighted Markov model are introduced to predict sluice deformation. In the method, the seagull optimization algorithm (SOA) is utilized to optimize the hyper-parameters of the neural network structure in LSTM primarily to improve the model. Subsequently, the relevant error sequences of the fitting results of SOA-LSTM model are classified and the Markovity of the state sequence is examined. Then, the autocorrelation coefficients and weights of each order are calculated and the weighted and maximum probability values are applied to predict the future random state of the sluice deformation. Afterwards, the prediction model of sluice deformation on the SOA-LSTM-weighted Markov model is proposed. Ultimately, the presented model is used to predict the settlement characteristics of an actual sluice project in China. The analysis results demonstrate that the proposed model possesses the highest values of R2 and the smallest values of RMSE and absolute relative errors for the monitoring data of four monitoring points. Consequently, it concluded that the proposed method shows better prediction ability and accuracy than the SOA-LSTM model and the stepwise regression model.
- 6 January 2023
- Published by SAGE Publications in Structural Health Monitoring
- Vol. 22 (5) , 3416-3436
Multi-output displacement health monitoring model for concrete gravity dam in severely cold region based on clustering of measured dam temperature field
Constructing an accurate dam displacement health monitoring (DHM) model is crucial to ensure the safety of the dam. However, previous studies on DHM focused on the analysis and prediction of a single measurement point, with little work on multiple measurement points, which leads to low efficiency in evaluating the overall status of dams. Furthermore, the majority of these models are based on hydraulic engineering in moderate climatic areas, which results in low accuracy when applied to severely cold regions. To address these issues, the HT cT model is proposed based on full consideration of extreme climate and engineering measures in cold regions, which replaces the commonly used harmonic function or air temperature with the features extracted from the measured temperature field after clustering as a temperature factor. The method effectively overcomes multicollinearity while ensuring accuracy. Moreover, multi-output least-square support vector regression (MOLSSVR), a multi-output model that can forecast multiple measurement points simultaneously, is proposed. By combining it with the HT cT, the efficiency of the model is significantly improved. In addition, grey wolf optimization (GWO) is introduced to search for the optimal hyperparameters of the coupling model. The feasibility, accuracy, robustness, and long-term predictive capability of the proposed model are validated with measured displacements of a concrete gravity dam in a severely cold region. The results show that the proposed model outperforms the four popular machine learning (ML) models, including two support vector regression (SVR)-based models and two neural network-based models, which illustrate that the proposed model has outstanding accuracy and excellent long-term predictive capability. It provides an accurate and efficient novel approach for dam displacement safety monitoring in severely cold regions.
- 6 January 2023
- Published by SAGE Publications in Structural Health Monitoring
- Vol. 22 (5) , 3416-3436
Multi-output displacement health monitoring model for concrete gravity dam in severely cold region based on clustering of measured dam temperature field
Constructing an accurate dam displacement health monitoring (DHM) model is crucial to ensure the safety of the dam. However, previous studies on DHM focused on the analysis and prediction of a single measurement point, with little work on multiple measurement points, which leads to low efficiency in evaluating the overall status of dams. Furthermore, the majority of these models are based on hydraulic engineering in moderate climatic areas, which results in low accuracy when applied to severely cold regions. To address these issues, the HT cT model is proposed based on full consideration of extreme climate and engineering measures in cold regions, which replaces the commonly used harmonic function or air temperature with the features extracted from the measured temperature field after clustering as a temperature factor. The method effectively overcomes multicollinearity while ensuring accuracy. Moreover, multi-output least-square support vector regression (MOLSSVR), a multi-output model that can forecast multiple measurement points simultaneously, is proposed. By combining it with the HT cT, the efficiency of the model is significantly improved. In addition, grey wolf optimization (GWO) is introduced to search for the optimal hyperparameters of the coupling model. The feasibility, accuracy, robustness, and long-term predictive capability of the proposed model are validated with measured displacements of a concrete gravity dam in a severely cold region. The results show that the proposed model outperforms the four popular machine learning (ML) models, including two support vector regression (SVR)-based models and two neural network-based models, which illustrate that the proposed model has outstanding accuracy and excellent long-term predictive capability. It provides an accurate and efficient novel approach for dam displacement safety monitoring in severely cold regions.