IRFM-CMNN effluent BOD concentration prediction method based on DE algorithm

Abstract

The invention discloses an IRFM-CMNN effluent BOD concentration prediction method based on a DE algorithm, which aims to solve the problems that the effluent BOD has strong nonlinearity and other characteristics and is difficult to accurately measure in real time due to mutual influence between the effluent BOD and each sewage effluent index in the current sewage treatment process.Based on the biochemical reaction characteristics of urban sewage treatment, an interactive recursive fuzzy membership cerebellar model neural network based on a DE algorithm is designed and used for predicting the concentration of the BOD key water quality parameter of the effluent of sewage treatment, and the problem that the BOD of the effluent of sewage treatment is difficult to accurately measure is solved; the result shows that the interactive recursive fuzzy membership cerebellar model neural network can rapidly and accurately measure the BOD concentration of sewage treatment effluent, and the operation parameters of the sewage treatment system are adjusted in a targeted manner according to the accurate prediction result, so that the stable and effective operation of the sewage treatment system is ensured; the unqualified water treatment efficiency of a sewage treatment plant is improved.

Description

technical field [0001] According to the biochemical reaction characteristics of sewage treatment, the present invention utilizes an interactively recursive fuzzy-membership cerebellar model neural network (IRFM-CMNN) based on a differential evolution algorithm (differentialevolution, DE) to realize sewage treatment The key water quality parameter in the process is the prediction of the concentration of biochemical oxygen demand (BOD) in the effluent. The BOD concentration in the effluent is an important parameter that characterizes the degree of water pollution and has a very important impact on the improvement of the water treatment system and the human living environment; The online real-time prediction of effluent BOD concentration is the basic link to realize the improvement of water quality, which belongs to the field of intelligent system control and sewage treatment; Background technique [0002] Sewage treatment effluent BOD is one of the important indicators to indi...

AI Technical Summary

Problems solved by technology

[0003] The role of sewage treatment plants in protecting the water environment cannot be ignored. There are physical, chemical, biological and physical-chemical methods for sewage treatment at home and abroad; among them, biological methods are widely used, mainly including activated sludge method, anaerobic Oxygen biological treatment method, and biofilm method; sewage is treated by biological method and then discharged through the effluent tank. The indicators in the effluent tank mainly include PH, suspended solids (suspended solids, SS), chemical oxygen demand (chemical oxygen demand, COD), effluent ammonia nitrogen (NH 3 -N), effluent chromaticity, phosphate and biochemical oxygen demand, etc., the content of most effluent indicators can be easily and quickly measured and controlled by online instruments and equipment; however, there are differences between effluent BOD indicators and various parameters Interaction and complex mechanism relations make it difficult to establish a prediction model for effluent BOD; many sewage treatment plants rely on traditional methods such as manual testing and microbial sensor methods to obtain the detection results of daily effluent BOD concentration indicators; technical methods such as manual testing It has a strong lag, and it takes at least 5 days from the beginning of sample collection to the preliminary measurement results, which seriously affects the timeliness of treating unqualified water; the microbial sensor method can directly detect the effluent indicators, but the sewage treatment process There are many water quality indicators and their characteristics are different. It is obviously unrealistic to design corresponding sensors corresponding to the response characteristics of each indicator; secondly, the cost of sensor equipment is high, and the maintenance cost is huge, which seldom appears in actual detection scenarios; thus, Using traditional detection methods to improve effluent water quality not only takes a lot of time, manpower and equipment costs, but also has relatively large measurement errors;

Images