Optimization Of COD Soft Measurement And Control Strategy Of Sewage Aeration Process

The biochemical reactions are complex and changeable during the sewage treatment process.COD(Chemical Oxygen Demand)as an important factor of the treatment,could be considered as a variable that hard to be measured accurately,economically,in addition,real-time.However,using soft sensing technology,these problems could be solved.Aeration process could be one of the core procedure in the whole sewage treatment,which could contribute a lot in water quality and energy saving,therefore,the optimization of aeration process may bring a significant lifting of the whole system.This thesis based on the SBR(Sequencing Batch Reactor Activated Sludge Process)method of the current sewage treatment plant,using COD as the main variable,and some auxiliary variables,which were field acquisition.After the pre-treatment of the original data,a LSSVM(Least Square Support Vector Machine)was used to model and forecast COD.In the same time,an optimized LSSVM algorithm by PSO(Particle Swarm Optimization)was proposed,also,COD variable was modeled and forecast on MATLAB.After the comparison of these two models,PSO-LSSVM was found to be more accurately and fleetly,which could be a better method of soft sensing COD between those two algorithms.Furthermore,in the sewage treatment processes,the aeration could be the most power-consuming procedure.According to the change of COD concentration,fuzzy PID controller is used to control the frequency converter and make the blower frequency adjustable running.In addition,the results of COD soft sensing procedure was considered as the feedback which could be a part of the closed loop control system.A simulation model was founded on SIMULINK platform,the results showed the fuzzy controller could maintain the COD concentration efficiency.When the optimum process curve was created,the proper values of COD and capacities of air-blower could be detected.When applied on field,a significant reduction of power consuming was found.The new control methods brought a 40-45% reduction of electricity using by 1 ton sewage water treatment.In summary,the results showed,those models,methods,as well as algorithms could be considered to be a more effective way topower-saving,and had high practical value.