Font Size: a A A

Design And Application Of Real-time Sludge Settlement Ratio Soft Sensor Based On Artificial Neural Network

Posted on:2020-09-02Degree:MasterType:Thesis
Country:ChinaCandidate:H J PanFull Text:PDF
GTID:2381330596963869Subject:Control engineering
Abstract/Summary:PDF Full Text Request
Activated sludge is one of the most widely used treatment methods for municipal wastewater,and sludge bulking is a common problem,which has a negative impact on sewage treatment and even causes huge economic losses.Therefore,the prevention of sludge bulking is of great practical significance.Settling Velocity(SV%)is one of the main indicators of sludge settling performance,which is useful for sludge bulking prediction.However,the sewage treatment system has the characteristics of high non-linearity,complex operation mechanism,uncertainty and real-time change.Traditional methods have exposed many shortcomings,so the solution of neural network has practical significance.As the title suggests,this thesis is intended to design a real-time sludge settling ratio soft sensor,which includes the following aspects: soft sensor model,Back Propagation(BP)network working principle,network optimization,soft sensor architecture,testing and application.The main research contents and results are as follows:(1)The mechanism of sludge bulking was analyzed,and the soft sensing model of BP network was constructed by comparing various modeling methods such as statistical regression and fuzzy rules.(2)To optimize the BP network.Particle Swarm Optimization(PSO)is used to optimize the weight threshold of the network,and golden section method is used to optimize the hidden layer structure of the network.Secondly,aiming at overcoming the shortcomings of PSO itself,logarithmic inertia weight,adaptive mutation operator and dynamic learning factor are used to optimize the network more deeply.(3)A real-time soft sensor is designed.The sensor includes field data collection of auxiliary variables,BP network calculation,real-time data communication and user interface development.The test results verify the correctness of the model and the practicability of the soft sensor.The soft sensor based on improved PSO-BP network designed in this paper has the advantages of convenient construction,stable operation,high precision and good timeliness.It can be used as a reference scheme for measuring SV%.
Keywords/Search Tags:sludge bulking, sludge settlement ratio, particle swarm optimization, back propagation neural network, soft measuremen
PDF Full Text Request
Related items