论文部分内容阅读
针对污水处理过程能耗过高问题,提出一种基于神经网络的动态多目标优化控制方法。该方法对污水处理过程中的曝气能耗和泵送能耗同时优化,通过NSGA-II进化算法实现溶解氧浓度和硝态氮浓度设定值的动态寻优,由PID控制实现底层跟踪。采用神经网络在线建模方法构造污水处理过程多目标优化模型,解决了优化变量与性能指标间没有精确数学描述的问题。基于国际基准仿真平台BSM1的实验表明,与PID控制、单目标优化控制方法相比,多目标优化控制在保证出水水质达标的前提下可以获得更优的节能效果。
Aiming at the problem of high energy consumption in sewage treatment, a dynamic multi-objective optimization control method based on neural network is proposed. The proposed method optimizes both the aeration energy consumption and the pumping energy consumption in the wastewater treatment process. The optimization of the dissolved oxygen concentration and the nitrate nitrogen concentration set value is achieved by the NSGA-II evolutionary algorithm, and the bottom tracking is realized by the PID control. The multi-objective optimization model of sewage treatment process is constructed by online modeling method of neural network, which solves the problem that there is no accurate mathematical description between optimization variables and performance indexes. Experiments based on the international benchmark simulation platform BSM1 show that compared with PID control and single-objective optimization control, the multi-objective optimization control can achieve better energy-saving effect under the premise of ensuring effluent quality compliance.