Research On The Application Of CGWO-GRNN And CFD To MBR Sludge Production

Membrane Bioreactor(MBR)is a new type of wastewater treatment system that combines membrane separation technology with biological treatment technology.The utility model has the advantages of small occupied area,high effluent quality,less waste generation,organic matter which is difficult to degrade such as ammonia nitrogen,and convenient management and operation.However,membrane fouling restricts the performance of the MBR system and causes a significant increase in the energy consumption of the MBR system.The apparent yield of the sludge and the system water production are important parameters for measuring the pollution of the MBR membrane.Therefore,the membrane fouling mechanism is analyzed and the membrane is studied.Pollution control technology is critical to the application of MBR.In this paper,the de-correlation and dimensionality reduction of sludge yield influencing factors are first performed.Based on this,this paper applies the GRNN network to build a sludge yield prediction model.The model has fast convergence speed and small calculation amount,which is suitable for predicting sludge yield of MBR system with less sample data.And only need to determine a hyperparameter(smooth parameter ?)can reduce the influence of human factors to the greatest extent.Therefore,this paper introduces the Gray Wolf Optimization Algorithm(GWO)to optimize the smooth parameter ? of the GRNN network.The prediction results show that under the same conditions,the average relative error of the GWO-GRNN model is reduced by 15.6%compared to the GRNN model,which improves the prediction accuracy of the model to a certain extent,but it still fails to achieve the purpose of accurate prediction.Further research on GWO finds that the algorithm is prone to fall into a local optimum in the later stages of calculation.In order to improve the global optimization capability of GWO and the prediction accuracy of the model,a non-linear algorithm based on the cosine law is used to control the parameters and adopt Dynamically adjusting the weight improves the Gray Wolf Optimization Algorithm(CGWO).The improved algorithm not only prevents the algorithm from falling into a local optimum,but also improves the prediction accuracy of the model.By comparing the prediction results of sludge production,it can be found that the predicted results using the CGWO-GRNN model have the smallest deviation from the actual MBR system operation results,and the accuracy of the CGWO-GRNN prediction model is improved by 9.5%over the GWO-GRNN prediction model.The prediction model achieved the expected effect of accurately predicting sludge output,solved the problem of measuring membrane pollution by predicting sludge output,and had certain reference significance for MBR process design.In addition,in order to reduce the energy consumption of MBR system and membrane pollution,this article will improve the membrane module.The geometric model of the membrane module before and after the improvement is meshed.By observing the liquid streamline diagram and the shear force distribution cloud diagram of the membrane surface,it can be found that the improved membrane module can not only improve the disorder of the material-liquid flow and the existence of a large number of shear force blind spots,And can also improve the average surface shear force.Simulation results show that the improved membrane module can reduce membrane fouling and improve aeration efficiency.