Research On Intelligent Diagnosis Of Sludge Swelling Based On Fuzzy Neural Network

As a method often used in sewage treatment plants is the activated sludge method.The activated sludge in it can eliminate the susceptible substances contaminated in the water,viscous organic compounds and solid suspended solids.At the same time,activated sludge can also remove nitrogen and phosphorus contained in sewage,but due to sludge expansion in various activated sludge treatment processes,and sludge expansion with a high incidence,and It will cause harm to the sewage treatment process.Sludge expansion not only has a particularly bad influence on the sewage treatment process,but also causes inestimable impact and loss on municipal wastewater treatment.In this regard,the prediction of sludge expansion becomes more indispensable.The activated sludge method of sewage treatment is actually a measure of artificially cultivating a microorganism to degrade the biodegradable organic matter in oxidized sewage to complete sewage filtration.The activated sludge method has the advantages of convenient handling,relatively single structure,rationalization and other advantages.Even though it is often used in the process of sewage treatment in sewage plants,the occurrence of sludge expansion in high-frequency words will cause problems such as operation management and control,but it involves the development of urban sewage treatment.Sludge loss and lowering of sedimentation performance,followed by unqualified water quality in the effluent process are caused by sludge expansion.Seriously,it will cause the reduction of sewage treatment capacity,and eventually lead to the collapse of the treatment system.The article first analyzes the mechanism of sludge expansion.According to the requirements of sludge settling performance,the biochemical reaction mechanism of sewage treatment process,find out the variables related to SVI.Secondly,in order to improve the prediction accuracy of SVI and the adaptability of the prediction method to the input and output space,a self-organizing design of fuzzy neural network is carried out in this paper.First,combine the analysis of the influence of neuron input and output on the network performance,design a dynamic mechanism to achieve the growth and deletion of neurons,and secondly,use the least squares algorithm to modify the network weights;finally,use the LM algorithm to the center value and width of the network Perform training to improve the learning accuracy of the network.The network has a more compact structure and higher prediction accuracy.An SVI prediction model based on fuzzy neural network is designed.Obtain the process variable data related to SVI,and standardize the data,select the process variable with significant influence as the auxiliary variable,use the auxiliary variable as the input of the fuzzy neural network,and use the SVI as the network output,design the SVI prediction model based on the fuzzy neural network,Use the actual sewage treatment data to correct the model parameters,and realize the use of auxiliary variable data to predict SVI.Finally,a fault variable algorithm based on fuzzy neural network is proposed to intelligently diagnose the causes of sludge swelling.