Volume 24, Issue 2 (2024)
MCEJ 2024, 24(2): 111-123 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Optimization of location and diversion angle of lateral intakes in U shaped channel. MCEJ 2024; 24 (2) :111-123
URL: http://mcej.modares.ac.ir/article-16-68041-en.html
URL: http://mcej.modares.ac.ir/article-16-68041-en.html
Abstract: (1421 Views)
Lateral intake and stream diversion facilities, which are widely used in irrigation, land drainage and municipal sewage systems, deal with sediment transport. In this research, using a numerical discrete phase model, simulation of the sediment transport phenomenon is carried out in a 180-degree channel bend with a lateral intake that is located at a position of 115 degrees outside the arc and with a 45-degree diversion angle. The numerical model was calibrated with laboratory data.
Then, in 31 positions of the channel bend from the 10 to 140 degrees with 5 degrees' intervals and with 5 diversion angles of 10, 30, 50, 70 and 90 degrees, and for three diversion discharge ratios of 20%, 30% and 40%, the calibrated numerical model has been implemented and the percentage of sediment entered to the lateral intake were determined for each model. The results were used as the data necessary for training and validation of Artificial neural network (ANN) model.
The inputs of ANN model were location of lateral intake in outer bank of channel bend (φ) and diversion angle of lateral intake (θ), diverted water discharge ratio (Qr). The output is diverted sediment ratio into the lateral intake (Gr).
The data were used for training and validation of the ANN model and the best structure was obtained for the neural network models using R2, SSE and MARE criteria and compared to regression models. The best structure of ANN model was obtained as a 5-layer model with 3 membership functions for each input variable and 27 conditional rules. The accuracy of the results obtained from the models in terms of the mean absolute relative error (MARE) indicates the ability of the neural network models to predict the amount of sediment diverted to the lateral intake with respect to regression models. Then, by developing a new simulation-optimization method based on the Multi-Objective Genetic Algorithm Optimization model (NSGAII), the FLUENT numerical model and neural network models, the optimal position and diversion angle of the lateral intake in the U-shaped channel is determined to minimize the amount of sediment inflow into the lateral basin and maximize the rate of dewatering discharge.
By connecting the best model obtained from the ANN model, to NSGA-II model, the Pareto Front contains a list of optimal positions and optimal diversion angle of lateral intake are obtained with respect to the objective function and its limitations in the U-shaped channel.
Finally, to determine the optimal position and angle of the lateral basin in the U-shaped channel by connecting the Pareto front obtained from the Multipurpose Genetic Algorithm (NSGA-II) to the TOPSIS multi-criteria decision-making method, the best option is found among the Pareto front options. The results include the optimum position and optimum angle of lateral intake dewatering along with diversion discharge and sediment rates corresponding to these values. The results show that the optimum diversion angle obtained is 38-degree and the optimum position obtained is 127-degree
Then, in 31 positions of the channel bend from the 10 to 140 degrees with 5 degrees' intervals and with 5 diversion angles of 10, 30, 50, 70 and 90 degrees, and for three diversion discharge ratios of 20%, 30% and 40%, the calibrated numerical model has been implemented and the percentage of sediment entered to the lateral intake were determined for each model. The results were used as the data necessary for training and validation of Artificial neural network (ANN) model.
The inputs of ANN model were location of lateral intake in outer bank of channel bend (φ) and diversion angle of lateral intake (θ), diverted water discharge ratio (Qr). The output is diverted sediment ratio into the lateral intake (Gr).
The data were used for training and validation of the ANN model and the best structure was obtained for the neural network models using R2, SSE and MARE criteria and compared to regression models. The best structure of ANN model was obtained as a 5-layer model with 3 membership functions for each input variable and 27 conditional rules. The accuracy of the results obtained from the models in terms of the mean absolute relative error (MARE) indicates the ability of the neural network models to predict the amount of sediment diverted to the lateral intake with respect to regression models. Then, by developing a new simulation-optimization method based on the Multi-Objective Genetic Algorithm Optimization model (NSGAII), the FLUENT numerical model and neural network models, the optimal position and diversion angle of the lateral intake in the U-shaped channel is determined to minimize the amount of sediment inflow into the lateral basin and maximize the rate of dewatering discharge.
By connecting the best model obtained from the ANN model, to NSGA-II model, the Pareto Front contains a list of optimal positions and optimal diversion angle of lateral intake are obtained with respect to the objective function and its limitations in the U-shaped channel.
Finally, to determine the optimal position and angle of the lateral basin in the U-shaped channel by connecting the Pareto front obtained from the Multipurpose Genetic Algorithm (NSGA-II) to the TOPSIS multi-criteria decision-making method, the best option is found among the Pareto front options. The results include the optimum position and optimum angle of lateral intake dewatering along with diversion discharge and sediment rates corresponding to these values. The results show that the optimum diversion angle obtained is 38-degree and the optimum position obtained is 127-degree
Article Type: Original Research |
Subject:
Water
Received: 2023/03/13 | Accepted: 2023/10/19 | Published: 2024/04/29
Received: 2023/03/13 | Accepted: 2023/10/19 | Published: 2024/04/29
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |