| Abstrakt: |
Solid particle sedimentation is assumed to be a complex procedure in water and wastewater treatment plants. There is great interest in applying and developing different simulation and optimization methods to design primary sedimentation tanks (PSTs). In traditional techniques, mechanical and physical parameters are set by sequential error loops. This study proposes a hybrid method based on the response surface methodology, efficient metaheuristics, and scenario building methods using different experimental methods to eliminate the disadvantage of existing techniques. This novel framework creates a robust and sustainable design for the PSTs. First of all, the parameters of the considered PST, based on the economic, improve process and tank efficiency scenarios are tuned and optimized by the central composition design and response surface methodology. To forecast an efficient response value for these scenarios, different metaheuristic algorithms, including the genetic algorithm (GA), pattern search algorithm (PSA), and simulate annealing algorithm, are applied. Results demonstrated that PSA, GA, and PSA with 0.02, 0.032, and 0.063 compared to experimental practices have the best calibration for predicting response in economic, improved process and tank efficiency scenarios, correspondingly. Finally, experimental tests have proven that the optimum Retention Time is equal to 2 h based on the biological oxygen demand and the total suspended solids eliminations in the laboratory-scale setup. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink