| Municipal wastewater treatment is an important way to reduce environmental pollution,improve the efficiency of water resources recycling,and promote the development of harmonious society.It has become an important part of comprehensive utilization of water resources in China.In municipal wastewater treatment process(MWWTP),activated sludge process is mainly used to promote microbial adsorption,decomposition and oxidation of degradable organic pollutants in wastewater,and realize the wastewater purification.However,with the increasing scale of municipal wastewater treatment and the gradual complexity of the treatment technologies,the operation of municipal wastewater treatment plant is facing great challenges,especially the problems of high operating cost and substandard effluent quality.Therefore,it is an urgent problem to design the operation strategy of MWWTP,which can guarantee the effluent quality within the discharge limits and reduce the operation cost.In MWWTP,the influent quality and quantity are passively accepted,multiple interactive reaction units and coupling operation indices are contained,which makes the operation process dynamic,nonlinear and coupling.Therefore,the efficient operation of MWWTP cannot be separated from the effective optimal control strategy.Optimal control strategy can realize the optimization of the operation indices by designing the reasonable optimization strategy and control strategy.However,there still faces great challenges when implementing the optimal control strategy in MWWTP.The mechanism of comprehensive operation indices such as effluent quality and energy consumption of MWWTP is complex,and it is difficult to describe its characteristics through mechanism models.Moreover,the comprehensive operation indices are changing with the dynamic reaction process and operation time,which is challenging to balance the coupling comprehensive operation indices.In order to solve the above problems,based on the in-depth analysis of the operation mechanism of MWWTP,the comprehensive operation indices models based on adaptive kernel function,dynamic optimization algorithm for multiple conflicting operation objectives,intelligent optimal control strategy for MWWTP and cooperation optimal control strategy for multiple indices are designed in this paper.And the main innovative work and points are as follows:(1)Design of comprehensive operation indices models for MWWTPTo accurately obtain the dynamic characteristics of comprehensive operation indices including effluent quality and energy consumption,comprehensive operation indices models,based on the adaptive kernel functions,are established in the paper.Firstly,the key characteristic variables of effluent quality and energy consumption are mined based on the operation mechanism analysis and principal component analysis of MWWTP,which are selected as the inputs of the comprehensive operation indices models.Secondly,the comprehensive operation indices models,based on adaptive kernel functions,are designed to describe the nonlinear relationship of effluent quality,energy consumption and the input variables.Finally,the convergence of the designed comprehensive operation indices models is discussed to ensure its applicability.The practical application results show that the proposed models can achieve high model accuracy and realize the effective description of the dynamic characteristics of the effluent quality and energy consumption.(2)Research on dynamic optimization algorithm for multi-conflicting operation objectivesIn order to dynamically balance the conflicting operation objectives,a self-learning dynamic multi-objective particle swarm optimization algorithm is proposed in this paper.Firstly,a set of performance indices,by using the Chebyshev distance,are designed to obtain the state of the non-dominated solutions in the evolutionary process.Then the states of the particles can be detected.Secondly,a self-learning population size adjusting mechanism and an external archive updating mechanism,based on the designed performance indices are designed to realize the adaptive adjustment of the population size and the non-dominated solution.Then the performance of the particles and the evolutionary states can be matched.Finally,the feasibility analysis of the optimal solutions is discussed in the stages with fixed population size and dynamic population size.Then the effectiveness of the self-learning dynamic multi-objective optimization algorithm can be guaranteed.The experimental results of benchmark functions and the real-world application demonstrate that the proposed optimization algorithm can obtain feasible optimization solutions and realize the effective balance of the conflicting objectives.(3)Design of intelligent optimal control for WWTPTo guarantee the effluent quality within the limits and reduce the energy consumption,an intelligent optimal control strategy is proposed for MWWTP in this paper.Firstly,the optimization objectives are established by analyzing the operation characteristics of MWWTP.Then the dynamics of the operation process can be accurately described.Secondly,a dynamic multi-objective particle swarm optimization algorithm is designed to optimize the established optimization objectives.Then the optimal set-points of the control variables dissolved oxygen and nitrate nitrogen can be obtained.Thirdly,an adaptive fuzzy neural network controller is designed to trace the optimal set-points of dissolved oxygen and nitrate nitrogen.Finally,the proposed intelligent optimal control strategy is evaluated on a benchmark simulation platform,and the results illustrate that the proposed intelligent optimal control strategy can improve the operation effects,realize the balance of the effluent quality and energy consumption.(4)Research on cooperative optimal control for MWWTPIn order to coordinate the multiple time-scale operation indices,a hierarchical optimal control strategy is proposed in this paper.Firstly,the characteristics of the multiple time-scale operation indices are analyzed,and then the hierarchical optimization objectives are established based on the different operation time scales.Then the dynamic characteristics of the operation indices can be accurately described.Secondly,a cooperative optimization algorithm,based on the dynamic multi-objective particle swarm optimization,is designed to optimize the hierarchical optimization objectives.Then the optimal solutions of the control variables can be obtained.Thirdly,a model predictive control strategy is proposed to trace the time-varying optimal solutions of control variables.The optimal solutions of the control variables can be fatly traced.Finally,the stability and convergence of the proposed hierarchical optimal control strategy are analyzed to ensure its effectiveness.The application results demonstrate that the proposed hierarchical optimal control strategy can promote the operation efficiency,and realize the cooperative operation of the multiple operation indices.(5)Development of multi-objective optimal control system for WWTPIn order to verify the effectiveness and applicability of the proposed multi-objective optimal control strategy for MWWTP,a multi-objective optimal control system is developed.Firstly,different functional modules of the multi-objective optimal control system are developed,where data acquisition and treatment module,operation indices establishment module,dynamic optimization module,optimal control module,etc are contained.Secondly,the configuration software of the multi-objective optimal control system is developed to monitor the operation status of MWWTP.Finally,the proposed multi-objective optimal control system is applied to a small-scale municipal wastewater treatment plant to verify its effectiveness.The results demonstrated that the developed multi-objective optimal control system can improve the opearion efficiency and realize the optimal operation. |
PDF Full Text Request