知识和数据驱动的污水处理反硝化脱氮过程协同优化控制

Knowledge-data-driven Cooperative Optimal Control for Wastewater Treatment Denitrification Process

为有效提升城市污水处理过程的脱氮效果,提出一种知识和数据驱动的反硝化脱氮过程协同优化控制(Knowledge-data-driven cooperative optimal control,KDDCOC).所提方法主要有以下两个方面:首先,建立一种基于自适应知识核函数的协同优化控制目标模型,动态描述出水水质(Effluent quality,EQ)以及泵送能耗(Pumping energy consumption,PE)、关键变量的协同关系;其次,提出一种知识引导的协同优化算法(Knowledge guide-based cooperative optimization algorithm,KGCO),快速准确求解硝态氮(Nitrate nitrogen,SNO)优化设定值,提高KDDCOC的响应速度.KDDCOC利用比例−积分−微分(Proportional-integral-derivative,PID)控制器对硝态氮优化设定值进行跟踪,将提出的KDDCOC应用于城市污水处理过程基准仿真模型1号(Benchmark simulation model No.1,BSM1),实验结果表明,该方法能够提高出水水质,降低运行能耗,有效改善脱氮效果.

In order to effectively improve the performance of wastewater treatment denitrification process,a knowledge-data-driven cooperative optimal control(KDDCOC)is proposed.The main work of this paper includes the following two points:First,a cooperative optimal control objective model,based on adaptive knowledge kernel function,is designed to dynamically describe the cooperative relationship among effluent quality(EQ),pumping energy consumption(PE),and key variables;Second,a knowledge guide-based cooperative optimization algorithm(KGCO)is proposed to quickly and accurately obtain the optimal set-points of nitrate nitrogen(SNO).Then,the response speed of KDDCOC is improved.A proportional-integral-derivative(PID)controller is used to track the optimal setpoints of nitrate nitrogen.The proposed KDDCOC is applied to the benchmark simulation model No.1(BSM1)of wastewater treatment process.The experimental results indicate that KDDCOC can improve the effluent quality and the efficiency of denitrification,reduce the energy consumption.