Efficiency And Model Establishment Of AC Bioelectrochemical Treatment Azo Wastewater

Aiming at the disadvantage of traditional bioelectrochemical process that is difficult to have the ability of reduction and oxidation at the same time,this study took the typical bi-azo dye Alizarin yellow(AYR)as the target pollutant,established an alternating current bioelectrochemical system,so that the dye was completely degraded.In addition,the influence of AC potential factor on the system is studied,and the optimal operating conditions are explored.At the same time,put the data of dye degradation into different prediction models,establish the relationship between the data and the model,and explore the prediction accuracy of each model for the same prediction sample.Specific research conclusions are as follows:(1)In the working electrode area of the reactor,the AC electrochemical reduction at negative current plays a major role in the decolorization and degradation of alizarin yellow,while the intermediate products(PPD and 5-ASA)are mineralized at positive current.The removal rate of alizarin yellow can reach 83.5%,and the removal rate of PPD and 5-ASA can reach more than 90% under the operating conditions of influent dye concentration of 100 mg/L and applied AC voltage of 0.5 V.Through the statistics of current and potential,calculate the electron flux in the total process and the electron transfer flux used in the pollutant conversion process,and calculate that the total electron utilization rate in the alternating current is 10%,which cannot be achieved by other processes.(2)The performance of the reactor is optimized from two aspects of frequency and waveform.In this experiment,the decolorization and mineralization effects of 10 min and 20 min cycles are relatively good.Both decolorization rates can be maintained at more than 60%,and the mineralization rate can also be maintained at more than 50%.However,due to the problem of equipment accuracy,the realization of 20 min cycle AC power needs a more accurate power source,while the limitation of short cycle can be reduced,so the 10 min cycle AC power is the most suitable.In the waveform experiment,sine alternating current and square wave alternating current have good removal efficiency for azo decolorization and mineralization,but the efficiency of sine alternating current is better than that of square wave alternating current,which also verifies the reliability of gradient current.(3)The prediction of BP neural network optimized by genetic algorithm will reflect the characteristics of local optimization,which leads to the deviation of the accuracy of each prediction sample under this model;The prediction of BP neural network optimized by particle swarm optimization algorithm will reflect the characteristics of global optimization,which directly leads to the maximum error of some points,and the prediction curve of the remaining points is almost coincident with the curve of the real value.Finally,the combination of the two algorithms is used to organically combine the local optimization of genetic algorithm and the global optimization of particle swarm optimization algorithm,that is,the BP neural network model of PSO-GA optimization has good accuracy in data prediction.At the same time,under the comparison of error analysis,compared with the independent genetic algorithm or particle swarm optimization algorithm,the particle swarm optimization-genetic algorithm organically combined with the two has a better optimization effect on the improvement of BP neural network.