CN115980286B - Method for detecting wastewater at different stages of sewage treatment plant by utilizing electronic nose - Google Patents
Method for detecting wastewater at different stages of sewage treatment plant by utilizing electronic nose Download PDFInfo
- Publication number
- CN115980286B CN115980286B CN202211542861.3A CN202211542861A CN115980286B CN 115980286 B CN115980286 B CN 115980286B CN 202211542861 A CN202211542861 A CN 202211542861A CN 115980286 B CN115980286 B CN 115980286B
- Authority
- CN
- China
- Prior art keywords
- odor
- electronic nose
- sample
- wastewater
- samples
- 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.)
- Active
Links
- 239000002351 wastewater Substances 0.000 title claims abstract description 62
- 239000010865 sewage Substances 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 28
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 51
- 238000001514 detection method Methods 0.000 claims abstract description 35
- 238000000605 extraction Methods 0.000 claims abstract description 9
- 238000003909 pattern recognition Methods 0.000 claims abstract description 7
- 238000012549 training Methods 0.000 claims abstract description 4
- 239000007789 gas Substances 0.000 claims description 28
- 238000012706 support-vector machine Methods 0.000 claims description 19
- 238000004458 analytical method Methods 0.000 claims description 16
- 238000010238 partial least squares regression Methods 0.000 claims description 15
- 238000010561 standard procedure Methods 0.000 claims description 15
- 230000004044 response Effects 0.000 claims description 14
- 230000003750 conditioning effect Effects 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 12
- XKMRRTOUMJRJIA-UHFFFAOYSA-N ammonia nh3 Chemical compound N.N XKMRRTOUMJRJIA-UHFFFAOYSA-N 0.000 claims description 10
- 230000005540 biological transmission Effects 0.000 claims description 10
- 238000005259 measurement Methods 0.000 claims description 10
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 claims description 9
- 238000004140 cleaning Methods 0.000 claims description 9
- 239000000126 substance Substances 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 239000001301 oxygen Substances 0.000 claims description 6
- 229910052760 oxygen Inorganic materials 0.000 claims description 6
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 claims description 5
- 229910052698 phosphorus Inorganic materials 0.000 claims description 5
- 239000011574 phosphorus Substances 0.000 claims description 5
- 229910052757 nitrogen Inorganic materials 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 238000011088 calibration curve Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 claims description 3
- 238000007599 discharging Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000004451 qualitative analysis Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000001052 transient effect Effects 0.000 claims description 3
- 238000013480 data collection Methods 0.000 claims description 2
- 238000007865 diluting Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 2
- 238000010801 machine learning Methods 0.000 abstract description 2
- 235000019645 odor Nutrition 0.000 description 62
- 210000001331 nose Anatomy 0.000 description 46
- 230000036541 health Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 238000002290 gas chromatography-mass spectrometry Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000011897 real-time detection Methods 0.000 description 2
- 239000002689 soil Substances 0.000 description 2
- 208000035473 Communicable disease Diseases 0.000 description 1
- 206010028980 Neoplasm Diseases 0.000 description 1
- 241000863480 Vinca Species 0.000 description 1
- 231100000570 acute poisoning Toxicity 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000001684 chronic effect Effects 0.000 description 1
- 231100000739 chronic poisoning Toxicity 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 208000010643 digestive system disease Diseases 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000003651 drinking water Substances 0.000 description 1
- 235000020188 drinking water Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 230000008595 infiltration Effects 0.000 description 1
- 238000001764 infiltration Methods 0.000 description 1
- 238000001868 liquid chromatography-fluorescence detection Methods 0.000 description 1
- 230000000149 penetrating effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 208000017520 skin disease Diseases 0.000 description 1
- 239000003039 volatile agent Substances 0.000 description 1
- 238000004065 wastewater treatment Methods 0.000 description 1
Landscapes
- Investigating Or Analyzing Materials By The Use Of Fluid Adsorption Or Reactions (AREA)
Abstract
A method for detecting wastewater at different stages of a sewage treatment plant by utilizing an electronic nose comprises the following steps: assembling an electronic nose system for detection to perform sample detection and model training: the assembly is suitable for detecting the electronic nose required; collecting a wastewater sample and an odor sample; collecting smell sample data; and carrying out feature extraction and pattern recognition on the data to obtain a machine learning model. The electronic nose is used for wastewater detection at different stages of a sewage treatment plant, and comprises the following steps: arranging the electronic nose system at different treatment stages of a sewage treatment plant; and (3) using an electronic nose system to detect the odor concentration and predict the water quality parameters so as to determine whether the emission is normal. The invention can realize high-efficiency, accurate and low-cost detection of wastewater at different stages of the sewage treatment plant by utilizing the electronic nose, and informs the detection result in real time, thereby solving the problems of high cost, complex operation, high instrument requirement, low detection efficiency and the like of the existing water quality parameter detection technology and providing a more feasible method for wastewater detection of the sewage treatment plant.
Description
Technical Field
The invention belongs to the technical field of sewage detection, and particularly relates to a method for detecting wastewater at different stages of a sewage treatment plant by using an electronic nose.
Background
Illegal discharge from sewage treatment plants is a key source of environmental pollution.
The illegal discharge of sewage will bring various harm, will have direct influence on the health of the masses, and various diseases such as infectious diseases, digestive diseases, skin diseases, chronic or acute poisoning and the like can be caused by drinking water or food chains, even cancer is caused, and the health of the masses is threatened. Various pollutions can be generated to the environment, so that soil is polluted, and the soil is not suitable for cultivation; the water quality of the river is deteriorated, the normal utilization of the water body is affected, and the underground water is polluted by infiltration and other ways. The illegal discharge of sewage brings potential risks to human health and ecosystems, and it is necessary to accurately detect sewage quality and take measures in time.
The wastewater quality of a sewage treatment plant is determined by levels including Chemical Oxygen Demand (COD), ammonia Nitrogen (AN), total Nitrogen (TN), and Total Phosphorus (TP). The parameters comprise basic information of wastewater quality, are main basis for water quality monitoring, utilization and pollution control, and are helpful for determining whether wastewater meets the discharge requirement.
Currently, commonly used detection methods for wastewater quality and odor concentration include detection methods such as gas chromatography-mass spectrometry, thermal desorption-gas chromatography-mass spectrometry, gas chromatography-flame ionization detector, high performance liquid chromatography-fluorescence detection method and the like. The methods have good stability and sensitivity, but have the defects of high cost, large volume, complex pretreatment, long analysis time and the like, and cannot realize real-time detection application. Furthermore, these methods can only analyze one volatile compound in the gas under test in each test, and cannot describe the odor characteristics of the wastewater. Therefore, the conventional detection method is not suitable for on-site rapid characterization of wastewater in a sewage treatment plant.
Electronic nose is a new type of detection device that mimics biological smell, and responds to a variety of volatile substances through a sensor array. The electronic nose has the characteristics of low cost, miniaturization, simplicity, convenience, high efficiency, real-time in-situ and the like, and the technology is applied to various fields such as environment, food, agriculture, military, other scientific researches and the like. In addition, there are also some potential applications of electronic noses in wastewater analysis, and researchers have combined algorithms to evaluate concentration levels of odors in sewage treatment plants through electronic noses. In view of the advantages, the electronic nose is suitable for comprehensively describing the smell of wastewater and penetrating into the operation information of the sewage treatment plant contained in the excavated smell.
The treatment of wastewater odor in sewage treatment plants is related to environmental protection, the health of people is necessary to be paid attention to, and effective measures are taken to monitor the wastewater odor. Only if the condition of the discharged wastewater is accurately detected in real time, the wastewater treatment of the sewage treatment plant can be accurately finished by taking the odor concentration and the water quality parameter as the attention points.
Therefore, in order to solve the defects of high detection cost, large detection instrument volume, complex sample pretreatment, long analysis time, incapability of achieving real-time detection and the like of the existing wastewater quality detection method, an electronic nose system for wastewater detection at different stages of a sewage treatment plant is invented according to the relation between an electronic nose response signal and a detection result of a Chinese standard method. By combining the discharge condition of the sewage treatment plant, the construction of the recognition model is completed by detecting the wastewater discharged by the sewage treatment plant and collecting the response signals of the sensor, the real-time accurate prediction of the odor concentration and the water quality parameter condition is further realized, the early warning is actively carried out according to the prediction result, the discharge is controlled, the illegal discharge of the sewage treatment plant is avoided to a great extent, the health and the environmental safety of residents are effectively ensured, and the sewage treatment work takes effect rapidly.
Disclosure of Invention
The invention aims to provide a method for detecting wastewater at different stages of a sewage treatment plant by using an electronic nose, which can solve the problems of high detection cost, long analysis time, complex pretreatment and the like in the prior art.
A method for detecting wastewater at different stages of a sewage treatment plant by utilizing an electronic nose comprises the following steps:
1.1 Assembly of an electronic nose system suitable for detecting requirements: the electronic nose system consists of a data transmission interface, a data acquisition card, a conditioning circuit board, an air inlet pipe, a sensor array, a sensor cavity, a connecting pipe, an air outlet pipe and an air pump; the data acquisition card is positioned below the conditioning circuit board, the conditioning circuit board is positioned below the sensor cavity, and the sensor cavity and the air pump are sequentially arranged from left to right; the inlet of the sensor chamber is connected with the air inlet pipe, the outlet of the sensor chamber is connected with the inlet of the air pump through the connecting pipe, and the outlet of the air pump is connected with the air outlet pipe; the gas sensor array is fixed on the inner wall of the sensor cavity in a surrounding way and is connected with the data acquisition card through the conditioning circuit board, and the data acquisition card is provided with a data transmission interface; the gas to be tested enters the electronic nose system through the gas inlet pipe, is discharged out of the electronic nose system through the gas outlet pipe, and response signals generated by the sensor array are collected by the data acquisition card and are output to the computer end through the data transmission interface; the sensor array uses a plurality of gas sensors;
1.2 A wastewater sample and an odor sample are obtained, comprising the following steps:
1.2.1 Collecting wastewater samples and odor samples at different positions of a sewage treatment plant, collecting the wastewater samples at positions P1, P3, P4, P5 and P6 by using plastic bottles, and collecting m+n samples at each position, wherein m samples are used for qualitative analysis, n samples are used for detection analysis of water quality parameters, the n samples are collected in n times, and the time interval between every two collection is 5 days, so that the water quality parameters of each sample are ensured to be different;
1.2.2 An air pump is used for collecting odor samples at the P2 position into an odor collecting bag and diluting the odor samples to K concentrations, so that the odor samples are convenient for subsequent detection and use;
1.3 Using an electronic nose system for sample analysis, comprising the steps of:
1.3.1 The sampling frequency of the electronic nose system is set to be f, each sampling is divided into two stages of odor sample data acquisition and sensor array cleaning, and the time of the two stages is set to be t 1 and t 2 respectively;
1.3.2 Before testing, standing the wastewater sample for 15 minutes to generate headspace gas;
1.3.3 Connecting a plastic bottle mouth filled with a wastewater sample with the air inlet pipe 4, conveying a headspace gas into the electronic nose cavity 6 by using the air pump 9, enabling the headspace gas to react with the sensor array 5, obtaining smell sample data, and cleaning the electronic nose cavity 6 by using clean air before the next measurement is carried out so as to ensure that a reaction curve of the sensor array 5 returns to a base line position;
1.3.4 Directly connecting an odor collecting bag filled with an odor sample to the air inlet pipe 4, conveying the odor sample to the electronic nose cavity 6 by using the air pump 9, enabling the odor sample to react with the sensor array 5 to obtain odor sample data, and cleaning the electronic nose cavity 6 by using clean air before the next measurement is carried out so as to ensure that the reaction curve of the sensor array 5 returns to a base line position;
1.4 According to the Chinese standard method, detecting the four water quality parameters of chemical oxygen demand, ammonia nitrogen, total nitrogen and total phosphorus of the wastewater sample and the concentration of the odor sample;
1.5 Data preprocessing is performed by using the odor sample data obtained in the step 1.3.3 and the step 1.3.4: extracting features of the original data obtained by the gas sensor array (5), and extracting transient features of the smell sample data by using Fourier transformation and polynomial curve fitting; extracting steady-state features of the odor sample data using the maximum value, the average value, and the integral value;
1.6 Using Linear Discriminant Analysis (LDA), support Vector Machine (SVM) and partial least squares regression (PLS) to perform pattern recognition and data analysis on the data after feature extraction, performing pattern recognition result analysis, comprising the steps of:
1.6.1 Determining that the characteristics of wastewater samples at different positions in the sewage treatment plant are different, using the maximum voltage value of the response signals of the sensor as input data to be applied to parameters of an LDA model, and obtaining a visual LDA graph of odor samples from five sampling sites in the sewage treatment plant, wherein m samples exist in each category;
1.6.2 Using the characteristic values extracted by different characteristic extraction methods in the pretreatment stage as input data of an SVM support vector machine, training an SVM model for qualitative identification, and determining the source of a wastewater sample;
1.6.3 Quantitative prediction of wastewater by using a two-stage regression model constructed by PLS regression method, further accurately evaluating odor concentration and water quality parameters of a sewage treatment plant, comprising the following steps:
1.6.3.1 A regression model of the first stage is constructed for determining whether the odor concentration and water quality parameters of the wastewater are within the proper ranges. The samples for predicting the water quality parameters are collected from five positions P1, P3, P4, P5 and P6 respectively for n times, each sample is repeatedly detected by an electronic nose, the average value V max of repeated detection is calculated, the samples for predicting the odor concentration are odor samples which are collected from the position P2 and diluted to K different concentrations, a plurality of repeated detection samples are arranged under each concentration, and the repeated response values of each sensor are averaged to be used as the input parameters of a PLS model;
1.6.3.2 Setting a water quality parameter data input matrix of H multiplied by N elements, wherein the columns comprise V max values of each of N sensors in the electronic nose, and the rows comprise average V max values of H sample points; similarly, setting a K multiplied by N odor concentration input matrix; the input data obtained by using the Chinese standard method are stored in two matrixes H multiplied by 1 and K multiplied by 1, and each row represents the detection result of the two matrixes H multiplied by 1;
1.6.3.3 According to the data, calculating a regression coefficient, and then calculating the output result of the regression model as odor concentration and water quality parameters;
1.6.3.4 Constructing a second-stage regression model for more accurate parameter prediction, constructing a model by adopting the same method as the first stage, further reducing the sample range by combining the output result obtained in the previous step and the maximum emission limit, and constructing a two-stage regression model for odor concentration prediction and a two-stage regression model for predicting a plurality of water quality parameters;
1.6.3.5 A PLS method is used for constructing a two-stage regression model, a linear relation is formed between an electronic nose response signal and a Chinese standard method detection result, a global calibration curve of odor concentration and water quality parameters of a sewage treatment plant is obtained, and the model is used for predicting the odor concentration and water quality parameters discharged by the electronic nose in the sewage treatment process;
1.7 Arranging the electronic nose system at different treatment stages of a sewage treatment plant, detecting odor concentration and predicting water quality parameters, and if no over-standard parameters are detected, normally discharging; if the standard exceeding parameter is detected, an alarm is sent out and the discharge is stopped in a control room of the sewage treatment plant.
Compared with the prior art, the invention has the beneficial effects that:
1. the method for detecting the odor concentration and the water quality parameters of the sewage treatment plant by utilizing the electronic nose system and the machine learning algorithm can accurately determine the source of the wastewater, quantitatively forecast the source of the wastewater, has high precision and quick detection, and is suitable for rapidly and qualitatively determining the wastewater in the sewage treatment plant on site.
2. Compared with the prior art, the electronic nose system can detect the odor concentration and water quality parameter conditions in real time, realizes accurate dynamic monitoring of the discharge conditions of the sewage treatment plant, and is convenient for further executing control operation.
3. By detecting the wastewater discharged by the sewage treatment plant, predicting the odor concentration and water quality parameter conditions, and actively carrying out early warning and discharge control according to the prediction result, the illegal discharge of the sewage treatment plant is avoided to a great extent, and the health and environmental safety of residents are effectively ensured.
Drawings
FIG. 1 is a unitary frame diagram of the present invention;
FIG. 2 is a schematic diagram of an electronic nose system;
FIG. 3 is a schematic diagram of a sampling position;
FIG. 4 is a flow chart of data processing;
FIG. 5 is a visual LDA graph of odor samples from five sampling sites within a sewage treatment plant;
FIG. 6 is a correlation between the smell concentration prediction result of the first stage regression model and the measurement result of the standard method;
FIG. 7 is a correlation between the predicted chemical oxygen demand of the first stage regression model and the measured chemical oxygen demand of the standard method;
FIG. 8 is a correlation between ammonia nitrogen prediction results of the first stage regression model and the results obtained by the standard method;
FIG. 9 is a correlation between the total nitrogen prediction result of the first stage regression model and the standard method measurement result;
FIG. 10 is a correlation between the total phosphorus prediction results of the first stage regression model and the standard method measurement results.
Wherein: 1-a data transmission interface; 2-a data acquisition card; 3-conditioning a circuit board; 4-an air inlet pipe; 5-sensor arrays; 6-chamber; 7-connecting pipes; 8-an air outlet pipe; 9-an air pump.
Detailed Description
A method for detecting wastewater at different stages of a sewage treatment plant by utilizing an electronic nose system comprises the following steps:
1) Assembling an electronic nose system suitable for detecting requirements, as shown in fig. 2: the electronic nose system consists of a data transmission interface 1, a data acquisition card 2, a conditioning circuit board 3, an air inlet pipe 4, a sensor array 5, a sensor cavity 6, a connecting pipe 7, an air outlet pipe 8 and an air pump 9; the data acquisition card 2 is positioned below the conditioning circuit board 3, the conditioning circuit board 3 is positioned below the sensor cavity 6, and the sensor cavity 6 and the air pump 9 are sequentially arranged from left to right; the inlet of the sensor chamber 6 is connected with the air inlet pipe 4, the outlet of the sensor chamber 6 is connected with the inlet of the air pump 9 through the connecting pipe 7, and the outlet of the air pump 9 is connected with the air outlet pipe 8; the gas sensor array 5 is fixedly arranged on the inner wall of the sensor cavity 6 in a surrounding manner, is connected with the data acquisition card 2 through the conditioning circuit board 3, and is provided with the data transmission interface 1 on the data acquisition card 2; the gas to be tested enters the electronic nose system through the gas inlet pipe 4, is discharged out of the electronic nose system through the gas outlet pipe 8, and response signals generated by the sensor array 5 are collected by the data collection card 2 and are output to the computer end through the data transmission interface 1; the number of the sensors used in the sensor array 5 is TGS2612、TGS2611、TGS2620、TGS2603、TGS2602、TGS2610、TGS2600、GSBT11、MS1100、MP135、MP901、MP-9、MP-3B、MP-4、MP-5、MP-2、MP503、MP801、MP905、MP402、WSP1110、WSP2110、WSP7110、MP-7、TGS2612、TGS2611、TGS2620、MP-3B、MP702、TGS2610、TGS2600、TGS2618-COO in total;
2) The method comprises the following steps of:
2.1 Collecting wastewater samples and odor samples at different treatment positions of a sewage treatment plant in Ji Lin Sheng vinca, wherein the sampling positions are shown in figure 3, a plastic bottle is used for collecting wastewater samples at positions P1, P3, P4, P5 and P6, 47 samples are collected at each position, 40 samples are used for qualitative analysis, 7 samples are used for detection and analysis of water quality parameters, 7 samples are collected in 7 times, and the time between every two collection is 5 days, so that the water quality parameters of each sample are ensured to be different;
2.2 An air pump is used for collecting odor samples at the P2 position into an odor collecting bag, and the odor samples are diluted to 15 concentrations, so that the subsequent detection and use are facilitated;
3) Sample analysis using an electronic nose system, comprising the steps of:
3.1 The sampling frequency of the electronic nose system is set to be 100 Hz, each sampling is divided into two stages of odor sample data acquisition and sensor array cleaning, and the time of the two stages is respectively set to be 1 minute and 15 minutes;
3.2 Before testing, standing the wastewater sample for 15 minutes to generate headspace gas;
3.3 Connecting a plastic bottle mouth filled with a wastewater sample with the air inlet pipe 4, conveying a headspace gas into the electronic nose cavity 6 by using the air pump 9, enabling the headspace gas to react with the sensor array 5 to obtain smell sample data, cleaning the electronic nose cavity 6 by using clean air before the next measurement is carried out so as to ensure that a reaction curve of the sensor array 5 returns to a base line position, and detecting each sample by using the electronic nose for 4 times in quantitative analysis;
3.4 Directly connecting an odor collecting bag filled with an odor sample to the air inlet pipe 4, conveying the odor sample to the electronic nose cavity 6 by using the air pump 9, enabling the odor sample to react with the sensor array 5 to obtain odor sample data, and cleaning the electronic nose cavity 6 by using clean air before the next measurement is carried out so as to ensure that the reaction curve of the sensor array 5 returns to a base line position;
4) Detecting the four water quality parameters of chemical oxygen demand, ammonia nitrogen, total nitrogen and total phosphorus of the wastewater sample and the concentration of the odor sample according to a Chinese standard method, wherein the measurement results of the water quality parameters of the obtained wastewater sample are shown in the following table 1;
TABLE 1 Water quality parameter determination results for wastewater samples
| Sample of | COD | AN | TN | TP | Sample of | COD | AN | TN | TP | |
| NO.1 | 117 | 55.1 | 61.29 | 2.9 | NO.19 | 335 | 14.9 | 28 | 3.3 | |
| NO.2 | 690 | 20.2 | 54.2 | 9.3 | NO.20 | 99 | 1.28 | 9.89 | 0.69 | |
| NO.3 | 23 | 0.613 | 12.8 | 0.21 | NO.21 | 148 | 5.83 | 7.08 | 1.3 | |
| NO.4 | 41 | 0.148 | 11.3 | 0.42 | NO.22 | 45 | 0.698 | 6.95 | 0.4 | |
| NO.5 | 276 | 14.7 | 46.1 | 1.7 | NO.23 | 462 | 14.5 | 15.3 | 1.6 | |
| NO.6 | 36 | 0.871 | 3.96 | 0.28 | NO.24 | 55 | 0.726 | 5.99 | 0.59 | |
| NO.7 | 33 | 0.504 | 3.55 | 0.32 | NO.25 | 73 | 0.94 | 3.69 | 0.52 | |
| NO.8 | 674 | 19.3 | 21.5 | 1.9 | NO.26 | 174 | 14.8 | 18.64 | 2.4 | |
| NO.9 | 32 | 0.701 | 3.96 | 0.35 | NO.27 | 121 | 6 | 17.56 | 1.7 | |
| NO.10 | 483 | 15.6 | 15.8 | 3.7 | NO.28 | 364 | 8.94 | 15.38 | 3.2 | |
| NO.11 | 42 | 0.619 | 3.14 | 0.29 | NO.29 | 121 | 4.09 | 15.1 | 1.4 | |
| NO.12 | 55 | 0.646 | 3.88 | 0.29 | NO.30 | 523 | 4.91 | 11.64 | 3.6 | |
| NO.13 | 213 | 18.4 | 23.3 | 2.1 | NO.31 | 129 | 0.846 | 5.62 | 1.9 | |
| NO.14 | 50 | 1.33 | 6.89 | 0.46 | NO.32 | 205 | 6.99 | 12.97 | 2.5 | |
| NO.15 | 418 | 20.9 | 31.7 | 5.9 | NO.33 | 125 | 2.08 | 3.43 | 1.6 | |
| NO.16 | 34 | 2.28 | 8.92 | 0.2 | NO.34 | 31 | 0.613 | 11.4 | 0.17 | |
| NO.17 | 327 | 20.1 | 34.7 | 2.8 | NO.35 | 23 | 0.52 | 12.5 | 0.21 | |
| NO.18 | 84 | 0.984 | 3.71 | 0.67 | Maximum emission limit | 50 | 5 | 15 | 0.5 |
5) Performing data preprocessing by using the smell sample data obtained in the step 3.3) and the step 3.4): extracting features of the original data obtained by the gas sensor array (5), and extracting transient features of the smell sample data by using Fourier transformation and polynomial curve fitting; extracting steady-state features of the odor sample data using the maximum value, the average value, and the integral value;
6) The method for performing pattern recognition and data analysis on the data after feature extraction by using Linear Discriminant Analysis (LDA), support Vector Machine (SVM) and partial least squares regression (PLS) and performing pattern recognition result analysis comprises the following steps:
6.1 Determining that the characteristics of wastewater samples at different positions in the sewage treatment plant are different, using the maximum voltage value of the response signal of the sensor as input data to be applied to parameters of an LDA model, and obtaining a visual LDA graph of odor samples from five sampling positions in the sewage treatment plant, wherein 40 samples exist in each category, as shown in figure 5;
6.2 Using the feature values extracted by different feature extraction methods in the pretreatment stage as input data of an SVM support vector machine, and using a radial basis function to establish a classifier model; and (5) training an SVM model for qualitative identification, and determining the source of the wastewater sample. In the classification process, ten-fold cross-validation was used to evaluate the performance of the tool, with the recognition rates shown in Table 2 below. The result shows that different characteristic extraction methods have different recognition rates, but the recognition rates are all above 95.33%; the recognition rate of the SVM model based on the average value characteristic extraction method can reach 98.83%; this result shows that wastewater samples from five sites can be identified by the SVM model;
TABLE 2 LDA and SVM Classification results
| FT | PCF | IV | Max | Mean | |
| LDA | 94.67 | 97 | 97.83 | 97.67 | 98.33 |
| SVM | 95.33 | 96.33 | 98.5 | 98 | 98.83 |
6.3 Quantitative prediction of wastewater by using a two-stage regression model constructed by PLS regression method, further accurately evaluating odor concentration and water quality parameters of a sewage treatment plant, comprising the following steps:
6.3.1 A regression model of the first stage is constructed for determining whether the odor concentration and water quality parameters of the wastewater are within the proper ranges. The samples for predicting the water quality parameters are collected seven times from five positions P1, P3, P4, P5 and P6 respectively, each sample is detected 4 times by an electronic nose, and the average value V max of repeated tests is calculated; the samples for odor concentration prediction were odor samples collected from the P2 position and diluted to 15 concentrations, each with 4 repeated detection samples, and the four repeated response values of each sensor were averaged as input parameters to the PLS model;
6.3.2 A 35 x 32 element water quality parameter data input matrix is set, the columns comprise 32 Vmax values of each sensor in the electronic nose, and the rows comprise 35 sample points of V max values; similarly, a 15×32 odor concentration input matrix is set; the input data obtained by using the Chinese standard method are stored in two matrixes of 35 multiplied by 1 and 15 multiplied by 1;
6.3.3 According to the above data, the regression coefficient is calculated, and then the output result of the regression model is calculated as the odor concentration and water quality parameters, and the result is shown in fig. 6 to 10. The points represent modeled sample data, the lines represent fitting results, and both the predicted concentration and water quality parameters are distributed around the lines. The fitting result is determined by a correlation coefficient (R 2) and a Root Mean Square Error (RMSE), and the R 2 values of the odor concentration and water quality parameter model are between 0.9920 and 0.9983, so that the prediction result has a good linear relation; the root mean square error ranges from 0.0745 to 15.5895, indicating that the expected error range is around the true value from the prediction result of the regression model, but accounts for 8.46% -31.18% of the maximum limit emission range, followed by further use of the second stage regression model;
6.3.4 A second stage regression model is constructed for more accurate parameter predictions. Establishing a model by adopting the same method as that of the first stage, establishing a two-stage regression model of the predicted odor concentration by using a sample with the odor concentration of <15, and establishing two-stage regression models of a plurality of predicted water quality parameters by using a sample with the COD of <100 mg/L, AN of <10 mg/L, TN of <25 mg/L, TP of <1 mg/L, wherein the correlation coefficient results of the two-stage regression models are shown in Table 3;
TABLE 3 correlation coefficient results for regression models
| Smell concentration | COD | AN | TN | TP | |
| R2 | 1 | 0.9999 | 0.9999 | 0.9988 | 0.9997 |
| RMSE | 0.0029 | 0.1781 | 0.028 | 0.1889 | 0.0029 |
6.3.5 A PLS method is used for constructing a two-stage regression model, a linear relation is formed between an electronic nose response signal and a Chinese standard method detection result, a global calibration curve of odor concentration and water quality parameters of a sewage treatment plant is obtained, and the model is used for predicting the odor concentration and water quality parameters discharged by the electronic nose in the sewage treatment process;
7) Arranging the electronic nose system at different treatment stages of a sewage treatment plant, detecting odor concentration and predicting water quality parameters, and if no over-standard parameters are detected, normally discharging; if the standard exceeding parameter is detected, an alarm is sent out and the discharge is stopped in a control room of the sewage treatment plant.
Claims (1)
1. A method for detecting wastewater at different stages of a sewage treatment plant by utilizing an electronic nose is characterized by comprising the following steps of: comprises the following steps:
1.1 Assembly of an electronic nose system suitable for detecting requirements: the electronic nose system consists of a data transmission interface (1), a data acquisition card (2), a conditioning circuit board (3), an air inlet pipe (4), a sensor array (5), a sensor cavity (6), a connecting pipe (7), an air outlet pipe (8) and an air pump (9); the data acquisition card (2) is positioned below the conditioning circuit board (3), the conditioning circuit board (3) is positioned below the sensor cavity (6), and the sensor cavity (6) and the air pump (9) are sequentially arranged from left to right; the inlet of the sensor chamber (6) is connected with the air inlet pipe (4), the outlet of the sensor chamber (6) is connected with the inlet of the air pump (9) through the connecting pipe (7), and the outlet of the air pump (9) is connected with the air outlet pipe (8); the gas sensor array (5) is fixed on the inner wall of the sensor cavity (6) in a surrounding manner, is connected with the data acquisition card (2) through the conditioning circuit board (3), and is provided with the data transmission interface (1) on the data acquisition card (2); the gas to be tested enters the electronic nose system through the gas inlet pipe (4), is discharged out of the electronic nose system through the gas outlet pipe (8), and response signals generated by the sensor array (5) are collected by the data collection card (2) and are output to the computer end through the data transmission interface (1); the sensor array (5) uses a plurality of gas sensors;
1.2 A wastewater sample and an odor sample are obtained, comprising the following steps:
1.2.1 Collecting wastewater samples and odor samples at different positions of a sewage treatment plant, collecting the wastewater samples at positions P1, P3, P4, P5 and P6 by using plastic bottles, and collecting m+n samples at each position, wherein m samples are used for qualitative analysis, n samples are used for detection analysis of water quality parameters, the n samples are collected in n times, and the time interval between every two collection is 5 days, so that the water quality parameters of each sample are ensured to be different;
1.2.2 An air pump is used for collecting odor samples at the P2 position into an odor collecting bag and diluting the odor samples to K concentrations, so that the odor samples are convenient for subsequent detection and use;
1.3 Using an electronic nose system for sample analysis, comprising the steps of:
1.3.1 The sampling frequency of the electronic nose system is set to be f, each sampling is divided into two stages of odor sample data acquisition and sensor array cleaning, and the time of the two stages is set to be t 1 and t 2 respectively;
1.3.2 Before testing, standing the wastewater sample for 15 minutes to generate headspace gas;
1.3.3 Connecting a plastic bottle mouth filled with a wastewater sample with an air inlet pipe (4), conveying a headspace gas into an electronic nose cavity (6) by using an air pump (9), enabling the headspace gas to react with a sensor array (5), obtaining smell sample data, and cleaning the electronic nose cavity (6) by clean air before the next measurement is carried out so as to ensure that a reaction curve of the sensor array (5) returns to a base line position;
1.3.4 Directly connecting an odor collecting bag filled with an odor sample to an air inlet pipe (4), conveying the odor sample into an electronic nose cavity (6) by an air pump (9), enabling the odor sample to react with a sensor array (5), obtaining odor sample data, and cleaning the electronic nose cavity (6) by clean air before the next measurement is carried out so as to ensure that a reaction curve of the sensor array 5 returns to a base line position;
1.4 According to the Chinese standard method, detecting the four water quality parameters of chemical oxygen demand, ammonia nitrogen, total nitrogen and total phosphorus of the wastewater sample and the concentration of the odor sample;
1.5 Data preprocessing is performed by using the odor sample data obtained in the step 1.3.3 and the step 1.3.4: extracting features of the original data obtained by the gas sensor array (5), and extracting transient features of the smell sample data by using Fourier transformation and polynomial curve fitting; extracting steady-state features of the odor sample data using the maximum value, the average value, and the integral value;
1.6 Using Linear Discriminant Analysis (LDA), support Vector Machine (SVM) and partial least squares regression (PLS) to perform pattern recognition and data analysis on the data after feature extraction, performing pattern recognition result analysis, comprising the steps of:
1.6.1 Determining that the characteristics of wastewater samples at different positions in a sewage treatment plant are different, using the maximum voltage value of a sensor response signal as input data to be applied to parameters of an LDA model, and obtaining a visual LDA graph of odor samples from five sampling sites in the sewage treatment plant, wherein m samples exist in each class;
1.6.2 Using the characteristic values extracted by different characteristic extraction methods in the pretreatment stage as input data of an SVM support vector machine, training an SVM model for qualitative identification, and determining the source of a wastewater sample;
1.6.3 Quantitative prediction of wastewater by using a two-stage regression model constructed by PLS regression method, further accurately evaluating odor concentration and water quality parameters of a sewage treatment plant, comprising the following steps:
1.6.3.1 Constructing a regression model of the first stage, wherein the regression model is used for determining whether the odor concentration and the water quality parameter of the wastewater are in a proper range, samples for predicting the water quality parameter are respectively collected from five positions P1, P3, P4, P5 and P6 for n times, each sample is repeatedly detected by an electronic nose, an average value V max of repeated tests is calculated, the samples for predicting the odor concentration are diluted odor samples with K different concentrations, each concentration is provided with a plurality of repeated detection samples, and a plurality of repeated response values of each sensor are averaged to be used as input parameters of the PLS model;
1.6.3.2 Setting a water quality parameter data input matrix of H multiplied by N elements, wherein the columns comprise V max values of each of N sensors in the electronic nose, and the rows comprise average V max values of H sample points; similarly, setting a K multiplied by N odor concentration input matrix; the input data obtained by using the Chinese standard method are stored in two matrixes H multiplied by 1 and K multiplied by 1, and each row represents the detection result of the two matrixes H multiplied by 1;
1.6.3.3 According to the data, calculating a regression coefficient, and then calculating the output result of the regression model as odor concentration and water quality parameters;
1.6.3.4 Constructing a second-stage regression model for more accurate parameter prediction, constructing a model by adopting the same method as the first stage, further reducing the sample range by combining the output result obtained in the previous step and the maximum emission limit, and constructing a two-stage regression model for odor concentration prediction and a two-stage regression model for predicting a plurality of water quality parameters;
1.6.3.5 A PLS method is used for constructing a two-stage regression model, a linear relation is formed between an electronic nose response signal and a Chinese standard method detection result, a global calibration curve of odor concentration and water quality parameters of a sewage treatment plant is obtained, and the model is used for predicting the odor concentration and water quality parameters discharged by the electronic nose in the sewage treatment process;
1.7 Arranging the electronic nose system at different treatment stages of a sewage treatment plant, detecting odor concentration and predicting water quality parameters, and if no over-standard parameters are detected, normally discharging; if the standard exceeding parameter is detected, an alarm is sent out and the discharge is stopped in a control room of the sewage treatment plant.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211542861.3A CN115980286B (en) | 2022-12-03 | 2022-12-03 | Method for detecting wastewater at different stages of sewage treatment plant by utilizing electronic nose |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211542861.3A CN115980286B (en) | 2022-12-03 | 2022-12-03 | Method for detecting wastewater at different stages of sewage treatment plant by utilizing electronic nose |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115980286A CN115980286A (en) | 2023-04-18 |
| CN115980286B true CN115980286B (en) | 2024-08-23 |
Family
ID=85968998
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211542861.3A Active CN115980286B (en) | 2022-12-03 | 2022-12-03 | Method for detecting wastewater at different stages of sewage treatment plant by utilizing electronic nose |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115980286B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117521008B (en) * | 2023-11-09 | 2024-07-16 | 北京德众国良环保科技有限公司 | Sewage treatment monitoring control method and system based on data fusion |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105938116A (en) * | 2016-06-20 | 2016-09-14 | 吉林大学 | Gas sensor array concentration detection method based on fuzzy division and model integration |
| CN109239207A (en) * | 2018-07-23 | 2019-01-18 | 中山大学 | Odor identification method, apparatus and electric nasus system based on electronic nose |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7531137B2 (en) * | 2004-04-30 | 2009-05-12 | Honeywell International Inc. | Substance detection system |
| CN112268993A (en) * | 2020-09-14 | 2021-01-26 | 杭州电子科技大学 | Method for rapidly monitoring river channel water quality parameters based on electronic nose technology |
| CN112268994A (en) * | 2020-09-14 | 2021-01-26 | 杭州电子科技大学 | Method for quickly identifying black and odorous water body based on electronic nose technology |
| KR102538360B1 (en) * | 2021-04-14 | 2023-05-30 | 호서대학교 산학협력단 | Fast gas classification method in transient state using Recurrent neural network in an electronic nose system equipped with a gas sensor array |
| CN115047144B (en) * | 2022-05-30 | 2023-03-31 | 吉林大学 | Method for detecting drunk driving by using miniaturized electronic nose |
-
2022
- 2022-12-03 CN CN202211542861.3A patent/CN115980286B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105938116A (en) * | 2016-06-20 | 2016-09-14 | 吉林大学 | Gas sensor array concentration detection method based on fuzzy division and model integration |
| CN109239207A (en) * | 2018-07-23 | 2019-01-18 | 中山大学 | Odor identification method, apparatus and electric nasus system based on electronic nose |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115980286A (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Stuetz et al. | Assessment of odours from sewage treatment works by an electronic nose, H2S analysis and olfactometry | |
| Blanco-Rodríguez et al. | Development of an electronic nose to characterize odours emitted from different stages in a wastewater treatment plant | |
| Spinelle et al. | Field calibration of a cluster of low-cost commercially available sensors for air quality monitoring. Part B: NO, CO and CO2 | |
| CN114444259B (en) | Rain and sewage pipe network tracing and tracking system and method | |
| CN109444232A (en) | A kind of multichannel intelligent polluted gas monitoring device and diffusion source tracing method | |
| CN108665119B (en) | Water supply pipe network abnormal working condition early warning method | |
| Caron et al. | Identification of indoor air quality events using a K-means clustering analysis of gas sensors data | |
| CN115685853A (en) | Water environment pollution analysis management system and method based on big data | |
| CN115980286B (en) | Method for detecting wastewater at different stages of sewage treatment plant by utilizing electronic nose | |
| CN116340723B (en) | Rural water pollution quick tracing method and system based on big data | |
| CN117195135B (en) | Water pollution anomaly traceability detection method and system | |
| CN114544894A (en) | PMF model-based source analysis uncertainty analysis method and system | |
| CN116384268A (en) | Pollution source tracing method, device and equipment based on water pollution | |
| CN116448988A (en) | Industrial park soil pollution monitoring system and method | |
| Chen et al. | Towards accuracy recognition and content estimation of typical pesticides in groundwater via electronic nose | |
| KR100827647B1 (en) | Management apparatus for landfill | |
| CN117875573B (en) | Water environment evaluation method and system based on biotoxicity and bioavailability | |
| CN118275636A (en) | Method for analyzing heavy metal pollution source and pollution boundary of underground water by utilizing water system sediment | |
| Deshmukh et al. | MSW odor quantification using electronic nose and chemical analyzers: Relative exploration of prediction capabilities and robust model development | |
| CN111007220B (en) | Method for agile monitoring of water quality of live pig breeding sewage | |
| CN115290858A (en) | Soil pollution detection and analysis method and system | |
| Tan et al. | E-nose screening of pesticide residue on chilli and double-checked analysis through different data-recognition algorithms | |
| CN118209605B (en) | Arsenic intelligent monitoring method and system based on electrochemical sensor | |
| CN118396386B (en) | Method system and system for identifying and evaluating high risk links based on black and odor returning of urban water body | |
| Rivera Martinez et al. | Using metal oxide gas sensors for the estimate of methane controlled releases: reconstruction of the methane mole fraction time-series and quantification of the release rates and locations |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |