Research On Modeling And Optimization Control Of Printing And Dyeing Wastewater Treatment Combined Process System Based On BP Neural Network

With the Water Pollution Prevention Ordinance and Textile Dyeing and Finishing Industry Water Pollutant Discharge Standard promulgated and implemented step by step in recent years,industrial wastewater treatment condition becomes increasingly serious.In order to cope with the new standards,existing wastewater facilities and processes need to be carried out on a large scale and up-to-date.At the same time,it is also imperative for the simulation and optimization control of newly developed combination processes.Iron carbon microelectrolysis,biological contact oxidation(BCO)and advanced oxidation process(AOP)were selected in this paper to do the degradation tests of printing and dyeing wastewater.BP neural network models based on each process and the overall combined treatment system were constructed.Through the combination of model prediction and multi-objective optimization,the optimal control strategy was obtained with cost function as the goal and the discharge standard as the constraints.On the basis of the strategy,this paper also developed the corresponding PLC programme and the touch control interface based on MCGS configuration software.The specific research contents and conclusions are as follows:(1)Microelectrolysis,BCO and AOP were tested to investigate the optimum experimental conditions.The results of microelectrolysis tests showed that the removal rate of chemical oxygen demand(COD)and chroma was 70.19% and 91.32%,respectively,when the initial pH was 2.7,gas-water ratio was 14: 1 and the reaction time was 44 min.BCO results showed that the optimum hydraulic retention time(HRT)was 12 h,the optimum influent pH was 7.5 and the dissolved oxygen(DO)was 2~3 mg/L.Continuous operation showed that COD and NH3-N removal rates were stable at 85% and 83%.The optimum oxidant dosage of the AOP based on CSE-HG as oxidant was 0.8 ml/L,the optimum reaction time was 2 h,the optimum reaction pH was 2.5,and the average removal of COD and NH3-N observed by the continuous test were 62.2% and 25.6%,respectively.A combined treatment system test results showed that the average removal rate of COD and NH3-N was 96.3% and 84.2% under the optimum process parameters.(2)Based on the tests' data of each unit process and system combination,the BP neural network models of microelectrolysis,BCO and AOP process unit and system were established based on the data supplemented by the dynamic model.The ‘input-implict-output' structures of the models were 8-15-2,8-16-3,7-12-1 and 9-28-10,respectively.The results showed that the prediction error of each model was within ± 10%,± 8%,± 8% and ± 5%,respectively,which indicates that the models had good prediction abilities and the models were constructed successfully.(3)The control parameters of the equipments were selected by the process test parameters,and the cost function under the theoretical condition is obtained by calculation.Corresponding to the influent flow of 100 ~ 300 L/h,the theoretical calculation of tons of water running costs is 5.53 ~ 8.68 yuan.Tons of water cost function:Taking the cost function as the objective function,the effluent quality as the constraint condition,and optimization results show that when the system effluent meets 3 of the GB4287-2012 effluent standards,the corresponding control strategies are: microelectrolysis and AOP reaction pH=4,influent flow = 300 L/h,CSE-HG dosage a = 0.6 mg/L,corresponding to ORP value of 465 mV,biochemical DO control value of 3.6 mg/L,tons of water cost of 4.933 yuan.Micro-electrolysis and AOP reaction pH=2.5,influent flow = 200 L/h,CSE-HG dosage a = 0.6 mg/L,corresponding to ORP value of 485 mV,biochemical DO control value of 3.2 mg/L,cost 5.955 yuan.Micro-electrolysis and AOP reaction pH=2.5,inlet flow rate = 120 L/h,CSE-HG dosage a = 0.8 mg/L,corresponding to ORP value of 525 mV,biochemical DO control value of 2.4 mg/L,Tons of water costs 6.741 yuan.(4)With different plans for control objectives,control requirements were offered.According to the requirements,(a)system electrical circuit was designed including the control cabinet layout,control system distribution circuit,the main equipment external circuit and PLC wiring design.(b)PLC programs were designed including PID control of dosing pump,dead zone setting,range conversion program,manual/automatic switching procedures and error procedures.(c)Touch screen interfaces including process window,data display window,control screen windows.The control system was tested in the pilot test to show a continuous and stable effluent control results.