| The wastewater control in China has changed from simply improving water environment quality to the organic combination of water quality improvement,water resource protection,water ecological protection,energy conservation and emission reduction.The key to realizing the organic combination of water quality improvement,water resource protection,water ecological protection,energy conservation and emission reduction is that improve key features monitoring and energy consumption control capabilities of wastewater treatment plants(WWTP).However,due to the problems such as long time-consuming traditional monitoring method and low degree of automation,the key features monitoring and energy consumption control capabilities of WWTP in China are relatively limited,can not to adjust working conditions dynamically according to the variation of wastewater key features in time,resulting in the widespread problem of high energy consumption and low efficiency in China’s WWTP.In view of the above phenomenon,in order to improve the monitoring ability of the key features of wastewater,and ensure that the effluent quality is up to the standard,realize the energy consumption control of wastewater treatment process.According to the timing characteristics of wastewater data,a monitoring method of wastewater key features based on a hybrid neural network is designed.Through the full analysis of the mechanical characteristics in the wastewater treatment process,the hybrid neural network prediction model and multi-objective ant lion optimization(MOALO)algorithm are combined to design the energy consumption optimization control method,which is verified on the activated sludge water Benchmark Simulation Model no.1platform(BSM1).The main research contents of this thesis are as follows:(1)An effluent COD prediction model based on adaptive hybrid mutation particle swarm optimization(AHMPSO)and attention mechanism(AM)optimization of long short-term memory neural network(LSTM)is designed.Aiming at the nonlinear and uncertain characteristics of wastewater data,a neural network model based on AHMPSOLSTM-AM was proposed to predict the effluent COD.As the hyperparameters of the LSTM are difficult to select,AHMPSO is used to optimize the LSTM.A nonlinear variable inertia weight with random factors is introduced to balance the global search ability and the local search ability and to improve the convergence speed of the PSO algorithm.In addition,the hybrid mutation strategy is added in the search process to reduce the risk of particles falling into local optimal solutions.Finally,to improve the accuracy of prediction and the firmness ability of the model,the self-attention mechanism is introduced,and the dynamic data feature information is obtained by the method of variable weighting calculation,so as to improve the ability of the LSTM neural network to learn the importance of the local features of the data.The experimental results show that compared with the LSTM model,LSTM-AM model,and PSO-LSTM-AM model,the AHMPSO-LSTM-AM model achieves better prediction accuracy and stability in terms of predicting effluent COD.(2)Energy consumption optimization control method.Aiming at the situation of high energy consumption and high operating cost of the wastewater treatment plant,combined the AHMPSO-LSTM-AM model with multi-objective Ant-lion optimization algorithm(MOALO)and proposed an energy consumption optimization control method based on the BSM1 benchmark simulation platform.Firstly,the effluent COD prediction model and energy consumption prediction model are established by AHMPSO-LSTMAM neural network.Then,the dissolved oxygen and nitrate nitrogen were taken as the parameters to be optimized,the effluent COD prediction model and the energy consumption prediction model were taken as the objective function of the MOALO optimization algorithm,and the optimal values of the parameters to be optimized were solved as the set values of PID controller.Finally,the control method is validated on the BSM1 benchmark simulation platform.The simulation results show that compared with the open-loop control and the fixed set value PID method,the energy consumption optimization control method proposed in this thesis can effectively reduce energy consumption under the premise of ensuring wastewater quality. |
