Genetic algorithms for the application of Activated Sludge Model No. 1
Genetic algorithms for the application of Activated Sludge Model No. 1
复制标题
DOI：
10.2166/wst.2002.0636
复制
发表时间：
2002-01-01
影响因子：
2.7
通讯作者：
Kim, S
中科院分区：
文献类型：
--
作者：
Kim, S;Lee, H;Kim, S
遗传算法（GA）已集成到IWAASM No. 1中，用于校准重要的化学计量学和动力学参数。利用遗传算法的进化特征配置多焦点最优和全局最优。设计优化目标函数，使活性污泥系统出水浓度预测值与实测值之间的差异最小。采用反硝化布置，利用模拟基准的稳态和动态数据进行标定。该方法根据置信区间和目标函数给出了参数空间的分布。利用野外数据验证了遗传算法的定标能力，提出了动态定标方法来捕捉入流浓度的周期性变化。此外，为了在实际污水处理厂验证该方法，从海云台污水处理厂获得了底物浓度的测量数据集，并用于估计动态系统中的参数。校正参数后的模拟结果与实测出水COD浓度吻合较好。
The genetic algorithm (GA) has been Integrated into the IWAASM No. 1 to calibrate important stoichiometric and kinetic parameters. The evolutionary feature of GA was used to configure the multiple focal optima as well as the global optimum. The objective function of optimization was designed to minimize the difference between estimated and measured effluent concentrations at the activated sludge system, Both steady state and dynamic data of the simulation benchmark were used for calibration using denitrification layout. Depending upon the confidence intervals and objective functions, the proposed method provided distributions of parameter space. Field data have been collected and applied to validate calibration capacity of GA, Dynamic calibration was suggested to capture periodic variations of inflow concentrations. Also, in order to verify this proposed method in real wastewater treatment plant, measured data sets for substrate concentrations were obtained from Haeundae wastewater treatment plant and used to estimate parameters in the dynamic system. The simulation results with calibrated parameters matched well with the observed concentrations of effluent COD.