CN114354867A - Method for predicting volatile fatty acid in sewage - Google Patents
Method for predicting volatile fatty acid in sewage Download PDFInfo
- Publication number
- CN114354867A CN114354867A CN202111612107.8A CN202111612107A CN114354867A CN 114354867 A CN114354867 A CN 114354867A CN 202111612107 A CN202111612107 A CN 202111612107A CN 114354867 A CN114354867 A CN 114354867A
- Authority
- CN
- China
- Prior art keywords
- volatile fatty
- fatty acid
- acid
- sewage
- regression model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Activated Sludge Processes (AREA)
Abstract
The invention relates to sewage treatment, in particular to a method for predicting fatty acid in sewage. A method for predicting volatile fatty acid in sewage comprises the following steps: performing a static process experiment of biological sewage treatment, analyzing the Soluble Chemical Oxygen Demand (SCOD), ammonia nitrogen, phosphate, nitrate and Volatile Fatty Acid (VFA) of a sample, and detecting and measuring several indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid by the detection of the volatile fatty acid; constructing a regression model of the main components of the volatile fatty acid; constructing a partial least squares regression model of the volatile fatty acid; and (5) constructing a volatile fatty acid BP neural network regression model. The invention constructs a virtual monitoring model of volatile fatty acid which is difficult to monitor water quality variables on line, performs a static process experiment of sewage biological treatment to obtain a certain amount of data, performs model prediction based on principal component regression, partial least square regression and BP neural network, and solves the problem of obtaining real-time data of volatile fatty acid in sewage.
Description
Technical Field
The invention relates to sewage treatment, in particular to a method for predicting fatty acid in sewage.
Background
For sewage treatment plants with different inlet water qualities, corresponding sewage treatment processes need to be applied, but the automatic monitoring level of the water quality in the sewage treatment process needs to be improved, the space needing to be improved after the sewage treatment process is built is small, and the automation of the water treatment process has a great development space. The sewage treatment in-process is because the change of the inflow flow and the quality of water of intaking is undulant everyday, can make sewage play water quality of water be in unstable state, can lead to the mud inflation and lead to the processing failure when serious, and mechanical equipment trouble is frequent in sewage treatment in-process, and the transient nature and the uncertainty of technology operation have brought the serious challenge for sewage biological treatment system's steady operation and energy saving and consumption reduction, and consequently, it is indispensable to carry out real-time supervision control to sewage treatment process.
The concentration of key water quality parameters of a sewage treatment plant, such as Chemical Oxygen Demand (COD), Total Phosphorus (TP) and phosphate, and Volatile Fatty Acid (VFA) needs to be manually detected, the time for obtaining a result is far lagged behind the sewage treatment process, when the index concentration is found to exceed the standard, sewage is discharged into a water body, secondary pollution is caused, and if a detection result is waited, the treatment result is influenced.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides a method for predicting volatile fatty acid in sewage. The invention constructs a virtual monitoring model of volatile fatty acid which is difficult to monitor water quality variables on line, performs a static process experiment of sewage biological treatment to obtain a certain amount of data, performs model prediction based on principal component regression, partial least square regression and BP neural network, and performs regression on several nonlinear models, thereby solving the problem of obtaining real-time data of volatile fatty acid in sewage.
The invention is realized by the following steps: a method for predicting volatile fatty acid in sewage comprises the following steps: performing a static process experiment of biological sewage treatment, analyzing the Soluble Chemical Oxygen Demand (SCOD), ammonia nitrogen, phosphate, nitrate and Volatile Fatty Acid (VFA) of a sample, and detecting and measuring several indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid by the detection of the volatile fatty acid; constructing a regression model of the main components of the volatile fatty acid; constructing a partial least squares regression model of the volatile fatty acid; and (5) constructing a volatile fatty acid BP neural network regression model.
As a further limitation to the technical scheme, a static process experiment of the sewage biological treatment is carried out, and a sample is obtained on time to obtain data.
As a further limitation to the technical scheme, the static process experiment of the biological sewage treatment is carried out, and the experiment needs to be repeated for 5 times.
As a further limitation to the technical scheme, 1L of effluent of the primary sedimentation tank is added into the device, 1L of return sludge is added, 5mg/L of phosphate and 70mg/L of sodium acetate (converted into COD) are added, and the mixture is stirred for 1 hour under oxygen-poor condition and is aerobic for 3.5 hours. After one round of culture, standing and precipitating to take out supernatant, taking a sample after the first round, taking the sample once every 10min in an anoxic period, and obtaining the sample according to 1, 2, 3 and 3.5h in an aerobic period. A total of 5 experiments were performed to obtain 45 samples.
As a further limitation to the technical scheme, the Soluble Chemical Oxygen Demand (SCOD), ammonia nitrogen, phosphate, nitrate and Volatile Fatty Acid (VFA) of each sample need to be measured, and the detection of the volatile fatty acid is used for measuring several indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid.
As a further limitation to the technical scheme, the experimental article is specifically a 2.5L beaker.
As a further limitation to the above technical solution, the agitator is required to constantly agitate in the anoxic region and the aerobic region.
As a further limitation to the above technical solution, a regression model needs to be constructed, but a training set and a test set do not need to be distinguished.
As a further limitation to the above technical solution, a volatile fatty acid principal component regression model is constructed by using Soluble Chemical Oxygen Demand (SCOD), ammonia nitrogen, phosphate, nitrate and volatile fatty acid (several indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid are measured by detection of volatile fatty acid).
As a further limitation to the above technical scheme, a partial least squares regression model of volatile fatty acids is constructed by using Soluble Chemical Oxygen Demand (SCOD), ammonia nitrogen, phosphate, nitrate and volatile fatty acids (several indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid are measured by detection of volatile fatty acids).
As a further limitation to the above technical scheme, a volatile fatty acid BP neural network regression model is constructed by using Soluble Chemical Oxygen Demand (SCOD), ammonia nitrogen, phosphate, nitrate and volatile fatty acid (several indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid are measured by detection of volatile fatty acid).
The invention constructs a regression model for monitoring variable volatile fatty acid of water quality on line, performs a static process experiment of sewage biological treatment to obtain a certain amount of data, performs model prediction based on principal component regression, partial least square regression and BP neural network, and performs regression on several nonlinear models, solves the problem of obtaining real-time data of volatile fatty acid in sewage, and has great practical significance for monitoring, obtaining and timely regulating the volatile fatty acid of sewage treatment plants in China.
Drawings
FIG. 1 is a simulation diagram of a static process experiment of biological sewage treatment.
FIG. 2 is SCOD and ammonia nitrogen concentration time data obtained from a static experiment of biological sewage treatment.
FIG. 3 is the time data of phosphate and nitrate concentration obtained from static experiment of biological sewage treatment.
FIG. 4 shows the regression model results of the main components of volatile fatty acids.
FIG. 5 shows the results of partial least squares regression model of volatile fatty acids.
FIG. 6 shows the result of neural network regression of the volatile fatty acid BP.
Detailed Description
The invention is further described with reference to the following figures and examples:
as shown in figure 1, three beakers 1, 2 and 3 of 2.5L are taken, a device is built, the static process experiment of the sewage biological treatment is repeated for 5 times totally, as shown in figure 1 by the mark 4, 1L of primary sedimentation tank effluent is added into the device, 1L of return sludge is added, 5mg/L of phosphate and 70mg/L of sodium acetate (converted into COD) are added, and the mixture is aerobically stirred (marked 5) for 1h and then is aerobically stirred for 3.5h (marked 6). After one round of culture, standing and precipitating to take out supernatant, taking a sample after the first round, taking the sample once every 10min during the anoxic period, and taking the sample according to 1, 2, 3 and 3.5h during the aerobic period. 1L of primary sedimentation tank effluent is added in the second round, 5mg/L of phosphate and 70mg/L of sodium acetate (converted into COD) are added, and as a result, as shown in figures 2 and 3, the anaerobic stirring is carried out, the SCOD value is 82mg/L immediately after the mixing, and the SCOD can be consumed to 0 at the end of the second round of experiment. The concentration of the ammonia nitrogen after mixing is about 22mg/l, the ammonia nitrogen is not changed during the anoxic period, and the ammonia nitrogen is gradually reduced to 7mg/l in the aerobic process. After mixing, the phosphate reaches about 5mg/L, which indicates that the phosphate of the inlet water is about 1mg/L, because 5mg/L of phosphate is added to a laboratory, the phosphate obviously rises to about 8mg/L in an anoxic stage, gradually decreases to 3mg/L in an aerobic period, the concentration of nitrate after mixing is 5mg/L, the phosphate decreases to 0mg/L in the anoxic period, and the phosphate increases to about 11 in the aerobic period again.
And R language software is used as a model building tool. 33 groups of data are left after the experimental data are preprocessed, and a model is constructed by applying the 33 groups of data without distinguishing a training set and a testing set. To evaluate the fit and predictive effect of the virtual sensor model, R2 was used as a criterion for evaluating the virtual sensor model. The results of volatile fatty acid predictions based on partial least squares regression, principal component regression, and BP neural network regression methods are shown in fig. 4-6, where R2 is 0.1521, 0.1241, and 0.1302, and RMES is 0.6040, 0.6139, and 0.6689, respectively.
Claims (14)
1. A method for predicting volatile fatty acid in sewage is characterized by comprising the following steps:
performing a sewage biological treatment static process experiment, analyzing the sample solubility chemical oxygen demand SCOD, ammonia nitrogen, phosphate, nitrate and volatile fatty acid VFA, and detecting volatile fatty acid, wherein the volatile fatty acid detection needs to measure indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid;
constructing a regression model of the main components of the volatile fatty acid;
constructing a partial least squares regression model of the volatile fatty acid;
and (5) constructing a volatile fatty acid BP neural network regression model.
2. The method of claim 1, wherein the performing of the static process experiment of wastewater biological treatment comprises obtaining samples and data on time.
3. The wastewater biological treatment static process experiment of claim 2, wherein the wastewater biological treatment static process experiment is repeated 5 times.
4. The static process experiment of the biological treatment of sewage and timely obtaining the sample according to claim 2, characterized in that it comprises adding 1L of the effluent of the primary sedimentation tank into the device, adding 1L of the return sludge, adding 5mg/L of phosphate and 70mg/L of sodium acetate to convert into COD, stirring for 1h under oxygen deficiency and 3.5h under aerobic condition; after one round of culture, standing and precipitating to take out supernatant, taking a sample after the first round, taking the sample once every 10min during an anoxic period, and obtaining the sample according to 1, 2, 3 and 3.5h during an aerobic period;
a total of 5 experiments were performed to obtain 45 samples.
5. The wastewater biological treatment static process experiment of claim 4, wherein the device comprises a 2.5L beaker.
6. The experimental apparatus of claim 4, wherein the apparatus comprises a stirrer, and the stirrer is required to stir in the anoxic zone and the aerobic zone.
7. The method of claim 1, wherein the construction of the regression model of the principal components of volatile fatty acids does not include a discriminant training set and a test set.
8. The method for predicting volatile fatty acid in sewage according to claim 1, wherein the main component regression model of volatile fatty acid is constructed by using indexes of acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid in detection and determination of soluble chemical oxygen demand SCOD, ammonia nitrogen, phosphate, nitrate and volatile fatty acid.
9. The method of claim 7, wherein a regression model of the principal components of the volatile fatty acids is constructed, and the dimension of the data set with a large number of relevant variables is reduced and regression analysis is performed while maintaining the current change of the data set as much as possible.
10. The method for predicting the volatile fatty acid in the sewage according to claim 1, wherein a partial least squares regression model of the volatile fatty acid is constructed by using the dissolved chemical oxygen demand SCOD, ammonia nitrogen, phosphate, nitrate and the volatile fatty acid.
11. The method of claim 10, wherein the method for predicting volatile fatty acids in wastewater comprises constructing a partial least squares regression model of volatile fatty acids, and dividing data into the following groups of variables according to a covariance maximization criterion: x-input variable and Y-output variable; and decomposing the input variable data matrix X and the output variable data matrix Y at the same time, and establishing a regression relation equation between the corresponding interpretation hidden variable and the reaction hidden variable.
12. The method for predicting volatile fatty acid in sewage according to claim 1, wherein a volatile fatty acid BP neural network regression model is constructed by using soluble chemical oxygen demand SCOD, ammonia nitrogen, phosphate, nitrate and volatile fatty acid.
13. The method of claim 12, wherein the building of the volatile fatty acid (BP) neural network regression model comprises learning complex correlations between independent and dependent variables based on processing elements of a highly interconnected system; artificial neural network methods extract concepts directly from historical data.
14. The method of claim 7, wherein the regression model is constructed using R language software as a model construction tool.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111612107.8A CN114354867A (en) | 2021-12-27 | 2021-12-27 | Method for predicting volatile fatty acid in sewage |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111612107.8A CN114354867A (en) | 2021-12-27 | 2021-12-27 | Method for predicting volatile fatty acid in sewage |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114354867A true CN114354867A (en) | 2022-04-15 |
Family
ID=81102361
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111612107.8A Pending CN114354867A (en) | 2021-12-27 | 2021-12-27 | Method for predicting volatile fatty acid in sewage |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114354867A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105045951A (en) * | 2015-05-27 | 2015-11-11 | 华南理工大学 | Soft-measurement method for volatile fatty acid in effluent of anaerobic wastewater treatment system |
| CN107132325A (en) * | 2017-04-14 | 2017-09-05 | 华南理工大学 | A kind of flexible measurement method of the Anaerobic Waste Treatment System water outlet volatile fatty acid based on particle cluster algorithm and SVMs |
| WO2018116507A1 (en) * | 2016-12-19 | 2018-06-28 | フィル－ジャパン ワールドワイド マネジメント サービス， インコーポレイテッド | Activated sludge method for biological simultaneous removal of nitrogen and phosphorous |
| CN109273058A (en) * | 2018-09-20 | 2019-01-25 | 中轻国环（北京）环保科技有限公司 | A kind of composite algorism for the exceeded early warning of anaerobic processes volatile fatty acid |
-
2021
- 2021-12-27 CN CN202111612107.8A patent/CN114354867A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105045951A (en) * | 2015-05-27 | 2015-11-11 | 华南理工大学 | Soft-measurement method for volatile fatty acid in effluent of anaerobic wastewater treatment system |
| WO2018116507A1 (en) * | 2016-12-19 | 2018-06-28 | フィル－ジャパン ワールドワイド マネジメント サービス， インコーポレイテッド | Activated sludge method for biological simultaneous removal of nitrogen and phosphorous |
| CN107132325A (en) * | 2017-04-14 | 2017-09-05 | 华南理工大学 | A kind of flexible measurement method of the Anaerobic Waste Treatment System water outlet volatile fatty acid based on particle cluster algorithm and SVMs |
| CN109273058A (en) * | 2018-09-20 | 2019-01-25 | 中轻国环（北京）环保科技有限公司 | A kind of composite algorism for the exceeded early warning of anaerobic processes volatile fatty acid |
Non-Patent Citations (2)
| Title |
|---|
| 刘博;万金泉;黄明智;马邕文;王艳;: "基于PCA-LSSVM的厌氧废水处理***出水VFA在线预测模型", 环境科学学报, no. 06 * |
| 梁浩;黄圆萍;沈广辉;韩鲁佳;杨增玲;: "在线旁路近红外实时监测粪污厌氧发酵挥发性脂肪酸含量", 农业工程学报, no. 10 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Petersen | Calibration, identifiability and optimal experimental design of activated sludge models | |
| Vanrolleghem et al. | On-line monitoring equipment for wastewater treatment processes: state of the art | |
| AU2021101438A4 (en) | Adaptive control method and system for aeration process | |
| Lee et al. | Neural network modeling for on-line estimation of nutrient dynamics in a sequentially-operated batch reactor | |
| Vanrolleghem et al. | Transient response of aerobic and anoxic activated sludge activities to sudden substrate concentration changes | |
| Han et al. | Hierarchical neural network modeling approach to predict sludge volume index of wastewater treatment process | |
| CN110378533A (en) | A kind of intelligence aeration management method based on big data analysis | |
| Scholes et al. | A review of practical tools for rapid monitoring of membrane bioreactors | |
| CN115754207A (en) | Simulation method and system for biological sewage treatment process | |
| CN112782975A (en) | Sewage treatment aeration intelligent control method and system based on deep learning | |
| CN115470702A (en) | Sewage treatment water quality prediction method and system based on machine learning | |
| CN101556270A (en) | Device for detecting dynamic specific oxygen utilization rate of activated sludge system | |
| CN101539564B (en) | Method for detecting dynamic specific oxygen utilization rate of activated sludge system | |
| CN102183910B (en) | Method for detecting specific oxygen utilization rate of activated sludge microorganism online based on frequency control | |
| Lesage et al. | Calibration and application of a 1-D model for oxidation ditches | |
| CN101928064A (en) | Method for simulating paper-making wastewater treatment by activated sludge process | |
| CN201330211Y (en) | Working parameter self-optimizing simulation system for sewage treatment plant | |
| CN107686160B (en) | Sewage treatment method and system based on SBR reactor | |
| CN114354867A (en) | Method for predicting volatile fatty acid in sewage | |
| CN116263850A (en) | Online sewage water quality early warning method combining offline simulation data | |
| CN201508636U (en) | Teaching experimental apparatus for detecting dynamic ratio oxygen consumption rate in activated sludge system | |
| Wan-liang et al. | Soft-sensing method for wastewater treatment based on BP neural network | |
| Moussa et al. | Model-based evaluation of nitrogen removal in a tannery wastewater treatment plant | |
| CN112322483A (en) | Microbial activity determination system and microbial activity determination method | |
| Choubert et al. | Nitrogen removal from urban wastewater by activated sludge process operated over the conventional carbon loading rate limit at low temperature |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |