A water olfaction instrument based on carbon nanotube sensors and its application

Abstract

This invention belongs to the field of water quality testing equipment, specifically relating to an online water odor detector based on a carbon nanotube sensor and its application. The online water odor detector based on a carbon nanotube sensor includes a treatment reaction tank, an inlet pipe and an outlet pipe connected to the treatment reaction tank, a cleaning pipe, a water odor detection sensor, and a PLC controller. The water odor detection sensor is electrically connected to the PLC controller. The water odor detection sensor includes an array of odor gas sensors, which is formed by surface modification of graphene material with conductive MOF material to create an array of olfactory chips for detecting odor substances. This invention provides a rapid and convenient method for detecting the concentration of odor substances in water.

Description

Technical Field

[0001] This invention belongs to the field of water quality testing equipment, specifically relating to an online water olfaction instrument based on a carbon nanotube sensor and its application. Background Technology

[0002] Odor and taste testing of water bodies is a crucial aspect of water resource management, ecological protection, and public health maintenance. On one hand, humans have a high sensitivity to odorous substances; even extremely low concentrations in water can significantly affect its sensory characteristics, thus impacting drinking water quality and reducing public confidence in its safety. On the other hand, odorous substances primarily originate from the metabolic activities of microorganisms such as cyanobacteria and actinomycetes, and elevated concentrations typically indicate eutrophication or harmful algal blooms. Currently, standards such as the "Standards for Drinking Water Quality (GB5749-2022)" set limits on the concentration of odorous substances in water bodies. Therefore, odor and taste testing of source water bodies, treated water from water treatment plants, and water in distribution networks has become an important part of current water quality testing practices.

[0003] Currently, the concentration of odor substances in water bodies is mostly in the ng / L range, and solid-phase microextraction (SPE) combined with gas chromatography-mass spectrometry (GC-MS) is commonly used for detection. However, this method is cumbersome, has a long detection cycle, high equipment costs, and cannot achieve good real-time on-site monitoring. Therefore, there is an urgent need to develop an online water odor monitoring device and technology that combines high sensitivity and rapid response capabilities, meeting the requirements of lightweight, low-cost, and rapid identification. Summary of the Invention

[0004] In view of the above problems, the present invention provides an online odor identification instrument for water odor substances based on carbon nanotube sensors, which can realize rapid identification and on-site monitoring of the type and level of water odor.

[0005] An online water odor detector based on a carbon nanotube sensor includes a treatment reaction tank, an inlet pipe and an outlet pipe connected to the treatment reaction tank; it also includes a pipe cleaning device, a water odor detection sensor, and a PLC controller, wherein the water odor detection sensor is electrically connected to the PLC controller; the water odor detection sensor includes a graphene-based carbon nanotube odor sensor array, which is composed of olfactory chips formed by surface modification of graphene material with conductive MOF material arranged in an array; the pipe cleaning device is connected to the treatment reaction tank through a cleaning pipe.

[0006] To ensure automated and consistent sampling, the treatment reaction tank is equipped with an inlet and an outlet at opposite corners. The inlet is located on the upper side of the treatment reaction tank, and the inlet pipe and cleaning pipe are connected to the inlet via a tee. The outlet is located at the lower edge of the treatment reaction tank, and the outlet pipe is connected to the outlet. The inlet pipe and cleaning pipe are powered by peristaltic pumps to deliver the water sample to be tested and the cleaning water, respectively.

[0007] The treatment reaction tank has a hollow internal structure and is a sealed stainless steel metal container. A water odor detection sensor is fixed at the top of the treatment reaction tank, and the odor sensor array in the water odor detection sensor is exposed inside the treatment reaction tank and located above the inlet plane. An upper heating device is superimposed on the water odor detection sensor. A lower heating device and a magnetic stirrer are installed at the bottom of the treatment reaction tank.

[0008] The PLC controller is used to receive water sample odor data transmitted by the water odor detection sensor, process, analyze and store the data in real time, and determine the odor level (using pattern recognition and machine learning algorithms to classify concentrations).

[0009] Preferred,

[0010] It also includes a housing, a reaction tank, a PLC controller, and its PLC display, all fixed inside the housing.

[0011] The water odor detection sensor also includes a temperature sensor and a humidity sensor; the temperature sensor module and humidity sensor module built into the temperature sensor and humidity sensor are electrically connected to the data processing module.

[0012] The inlet pipe is connected to an inlet flow meter; the inlet flow meter is connected to a data processing module, which transmits the monitored inlet water data to the data processing module for processing, and displays the data on the display screen through the display module.

[0013] The drying device includes a miniature fan connected to an upper heating device, which provides a heat source for the drying device; the control module controls the opening and closing of the drying module, and at the same time, the control module controls the opening and closing of the upper heating module.

[0014] The control module controls the opening and closing of the carbon nanotube water odor detection module by receiving instructions from the timing module; the graphene-based carbon nanotube odor sensor array transmits the detected odor data to the data processing module; the temperature sensor module transmits the detected temperature data to the data processing module; and the humidity sensor module transmits the detected humidity data to the data processing module.

[0015] The temperature data received in the data processing module is used to determine the threshold for whether to use the odor data detected by the graphene-based carbon nanotube odor sensor array; the humidity data received in the data processing module is used to determine the threshold for whether to open the water inlet pipeline solenoid valve module after the drying device has finished working.

[0016] Specifically,

[0017] The inlet water pipeline is made of polytetrafluoroethylene (PTFE) and is connected to an inlet water flow meter. The inlet water is controlled by a diaphragm pump at a flow rate of 500 mL / min, and the injection volume is controlled by a liquid level detection method installed in the reaction tank. Before injection, the inlet water pipeline is flushed with water at a flow rate of 500 mL / min for 2 minutes. Then, gas cleaning is performed using a diaphragm pump at a flow rate of 1.2 L / min, with the gas being air filtered through activated carbon, for a cleaning time of 5 minutes.

[0018] A heating device is connected to one side of the upper end of the water odor detection sensor, and its air outlet is located at the edge of the odor sensor array in the water odor detection sensor; a lower heating device and a magnetic stirrer are installed at the bottom of the treatment reaction tank, and the temperature is controlled at 35-50℃ to ensure that the odor substances in the water are stably volatilized to the top of the reaction tank, so that the sample acquisition conditions are consistent.

[0019] The water odor detection sensor also includes a temperature sensor and a humidity sensor; the temperature sensor module and humidity sensor module built into the temperature sensor and humidity sensor are electrically connected to the data processing module.

[0020] This invention employs a graphene-based water odor detection sensor. It utilizes conductive MOF material to modify the surface of graphene, forming an olfactory chip for detecting odorous substances. This chip is then arranged in an array to form a graphene-based carbon nanotube odor sensor array, constituting the water odor detection sensor. The graphene-based carbon nanotube odor sensor array is a modular arrangement, consisting of 1-8 olfactory chips. Each olfactory chip has several data channels (preferably 7), allowing for the simultaneous transmission of multiple data points to a data processing center.

[0021] One olfactory chip can respond to a variety of odor substances to different degrees. Typically, four olfactory chips form a 2×2 array, with each array having seven channels, resulting in 4×7=28 responses, which have a broad spectrum.

[0022] The PLC controller includes a control module and a power supply module, a display module, a data processing module, a communication module, a storage module, and a timing module, which are electrically connected to the control module respectively.

[0023] It also includes an inlet water pipe solenoid valve module, an outlet water pipe solenoid valve module, a cleaning pipe solenoid valve module, an upper heating module, a lower heating module, a drying module, a magnetic stirring module, a carbon nanotube water odor detection module, and a peristaltic pump module, all of which are electrically connected to the PLC controller control module.

[0024] The inlet water pipeline solenoid valve module, outlet water pipeline solenoid valve module, and cleaning pipeline solenoid valve module are used to control the opening and closing of the inlet water pipeline, outlet water pipeline, and cleaning pipeline, respectively. The solenoid valve modules are uniformly controlled by the control module in the PLC controller; the inlet water pipeline solenoid valve module and the cleaning pipeline solenoid valve module switch between opening and closing.

[0025] The cleaning module is used to control the cleaning pipeline to clean the treatment reaction tank;

[0026] The upper heating module is used for upper heating treatment of the water sample to be tested;

[0027] The lower heating module is used for lower heating treatment of the water sample to be tested;

[0028] The drying module is used to process the heating and drying of the reaction tank.

[0029] The magnetic stirring module is used to stir the water body to be tested at a certain rotation speed;

[0030] The water odor detection module is used to detect odor substances in the treated water and transmit the signal to the data processing module.

[0031] The peristaltic pump module is used to control the peristaltic pump to power the delivery of the water sample to be tested and the clean water;

[0032] The peristaltic pump module contains two independent power modules, and the control module in the PLC controller controls the delivery of the water sample to be tested and the water sample to be cleaned.

[0033] The data processing module is used to receive the water sample odor signal transmitted by the carbon nanotube water odor detection module, process, analyze and store the signal in real time, and classify the odor level.

[0034] The odorous substances mentioned are four common odorous substances found in drinking water sources: dimethyl isoborneol, geosmin, dimethyl disulfide, and dimethyl trisulfide.

[0035] The odor rating is the concentration range of the four odor substances in water. The range for rating 1 is set at 0-10 ng / L; the range for rating 2 is set at 11-50 ng / L; the range for rating 3 is set at 51-100 ng / L; and the range for rating 4 is set at 101 ng / L and above.

[0036] The data acquisition mode of the water odor detection sensor is as follows:

[0037] The graphene-based carbon nanotube odor sensor array will collect raw response data of the water sample to be extracted after a preset number of cycles, and construct the corresponding sensor signal based on the raw response data;

[0038] The data processing module includes a machine learning model based on linear discriminant analysis; the method for constructing the machine learning model based on linear discriminant analysis and processing data is as follows:

[0039] Linear discriminant analysis is used to analyze the labeled data, find the optimal coordinate system in high-dimensional space, and determine the optimal projection direction. In the obtained optimal coordinate system, when a new sensor signal is received, the sensor signal is projected according to the optimal projection direction, and the odor level is determined based on its position in space. The odor data (sensor signals) monitored by the graphene-based carbon nanotube odor sensor array received in the data processing module are matched to the corresponding odor level and displayed on the screen through the display module.

[0040] To train a sensor array to distinguish the categories and levels of odor substances, the sensor signals were arranged sequentially into a 154-dimensional vector, and linear discriminant analysis (LDA) was performed on it. LDA is a classic supervised machine learning method whose core objective is to construct and optimize an objective function to obtain the most favorable projection direction for classification, thereby achieving effective data classification. It can be used in scenarios such as odor substance identification using sensor arrays. This method uses the in-class divergence matrix (S... w ) and inter-category scatter matrix (S b Construct an objective function J(w), and by optimizing the projection direction vector w to maximize J(w), the direction most conducive to classification can be obtained.

[0041]

[0042]

[0043]

[0044] Where S w Let C be the within-class scatter matrix, C be the number of classes (four classes in total), X be the feature vector of a single sample, and μ be the within-class scatter matrix. c Let S be the sample mean vector of the Cth class. b Let μ be the scatter matrix between categories, and μ be the global mean vector of all samples. c -μ is the deviation vector between the mean of class C and the global mean, J(w) is the objective function, representing the ratio of "between-class divergence" to "within-class divergence", and W T It is the transpose of W, WT S b W is S b The projective divergence in the W direction, W T S w W is S w The projective divergence in the W direction, W C Let N be the set of samples of class C. C Let w be the number of samples in class C, and w be the projection direction vector.

[0045] LDA analysis was performed on the sensor data, and the four categories of odor substances were labeled with different colors. The scatter plots of the category distribution after 2D and 3D dimensionality reduction are shown below. Figure 6 As shown.

[0046] Principal component analysis showed that the cumulative contribution rate of the first two principal components reached 95.5%, and the cumulative contribution rate of the first three principal components reached 98.5%, indicating that the signal features after dimensionality reduction still retained most of the information of the original signal, and the sensor array has the ability to identify odor-causing substances (odor substances).

[0047] To further verify whether the sensor array has the ability to identify the concentrations of the aforementioned substances, a single-substance concentration gradient was constructed: 5, 10, 15, 20, 30, 40, 45, 50, 55, 60, 70, 80, 90, 95, 100, 105, 120, 150, 180, 200, 300, 400, and 500 ng / L. The collected sensor array signals were input into a linear discriminant analysis model, and the output layer was modified to output the concentrations of the odor-causing substances. The predicted concentrations of the four odor-causing substances are shown below. Figure 7 As shown, the prediction results for the four levels are as follows: Figure 8 As shown.

[0048] In addition, the present invention also provides the application of an online water odor detector based on carbon nanotube sensors for real-time on-site detection at water sources.

[0049] Considering the inconvenience of water sample transportation, the uncertainty of water sample preservation, the variability of human factors during water sample pretreatment and detection, and the problems of low detection signal response and poor detection process stability in existing water odor detection equipment, this application provides an online water odor detector based on a carbon nanotube sensor. The detector includes an inlet / outlet water module, a cleaning module, a solenoid valve module, a treatment reaction tank module, an upper heating module, a lower heating module, a drying module, a magnetic stirring module, a carbon nanotube water odor detection module, and a data processing module. It enables automated detection of odor substances in water and features a self-cleaning function, allowing for deep cleaning of the treatment reaction tank module after each detection. The coordinated use of the inlet / outlet water module, cleaning module, solenoid valve module, treatment reaction tank module, upper heating module, lower heating module, drying module, and magnetic stirring module replicates the laboratory pretreatment process for odor detection. Combined with the carbon nanotube water odor detection module and data processing module, it enables continuous detection of odors in water.

[0050] Compared to laboratory testing methods, this method offers faster detection speeds, lower requirements for on-site testing environments, and automated and stable sampling capabilities, enabling real-time on-site monitoring of water sources. Currently, waterworks routinely test for odor substances at water sources twice a year, a relatively low frequency that hinders timely identification of risk points. The device proposed in this application can achieve timely response and rapid identification of risk points at water sources through long-term, high-frequency real-time on-site monitoring.

[0051] The technical effects and advantages of this invention are as follows:

[0052] 1. Detecting odors and tastes in water. Existing odor and taste detection instruments detect odor and taste substances in gases. However, the concentration of odor and taste substances in water is very low (ng / L) and has a certain degree of solubility, making it difficult to volatilize into the gas phase, thus requiring very high detection accuracy.

[0053] 2. The detection method differs. It employs pattern recognition and machine learning to categorize concentrations, classifying odor and taste substance concentrations in water into four zones: 0-10 ng / L, 11-50 ng / L, 51-100 ng / L, and above 101 ng / L. This differs from the traditional method of directly detecting concentrations, thus improving detection speed and reducing costs.

[0054] 3. Enables on-site monitoring. The on-site environment is harsh and easily affected by temperature, humidity, and particulate matter in the water, thus increasing the difficulty of detection. The equipment designed in this patent fully considers environmental factors and reduces environmental interference.

[0055] 4. Online testing has been achieved. The existing testing instruments require samples to be collected and sent to the laboratory for testing, while this instrument realizes the functions of automatic sample collection, automatic sample injection, automatic cleaning, and automatic remote transmission of test data. Attached Figure Description

[0056] Figure 1 This is a schematic diagram of the online water olfaction instrument based on a carbon nanotube sensor, which is involved in this invention.

[0057] Figure 2 This is a schematic diagram of the PLC controller module structure involved in the present invention.

[0058] Figure 3 This is a schematic diagram of the control module structure involved in the present invention.

[0059] Figure 4 This is a schematic diagram of the data processing module structure involved in the present invention.

[0060] Figure 5 This is a schematic diagram of the graphene-based carbon nanoparticle odor sensor array structure involved in this invention.

[0061] Figure 6 The results are two-dimensional and three-dimensional principal component diagrams of the linear discrimination results of four odor substances.

[0062] Figure 7 This is the result of linear discriminant analysis predicting the concentrations of four odorants.

[0063] Figure 8 It is a confusion matrix with four levels.

[0064] The reference numerals in the above figures are explained as follows: 1. Housing; 2. PLC display; 3. PLC controller; 4. Water odor detection sensor; 5. Inlet pipe; 6. Cleaning pipe; 7. Outlet pipe; 8. Drying device; 9. Upper heating device; 10. Lower heating device; 11. Treatment reaction tank; 13. Inlet pipe solenoid valve module; 14. Cleaning pipe solenoid valve module; 15. Outlet pipe solenoid valve module; 16. Magnetic stirrer; 17. Peristaltic pump; 51. Odor sensor array; 52. Odor chip. Detailed Implementation

[0065] The present invention will be further described in detail below with reference to the embodiments given in the accompanying drawings.

[0066] An online water odor detector based on a carbon nanotube sensor includes a treatment reaction tank 11, an inlet pipe 5, an outlet pipe 7, a pipe cleaning device, a water odor detection sensor 4, and a PLC controller 3, all connected to the treatment reaction tank 11. The water odor detection sensor 4 is electrically connected to the PLC controller 3. The water odor detection sensor 4 includes an odor sensor array 51, which is formed by surface modification of graphene material with conductive MOF material to create an array of odor chips 52 for detecting odor substances. The pipe cleaning device is connected to the treatment reaction tank 11 via a cleaning pipe 6.

[0067] Preferably, the device also includes a housing 1, a processing reaction tank 11, a PLC controller 3 and its PLC display 2, which are fixed inside the housing 1.

[0068] The treatment reaction tank 11 is provided with an inlet and an outlet at opposite corners. The inlet is located on the upper side of the treatment reaction tank 11. The inlet pipe 5 and the cleaning pipe 6 are connected to the inlet via a tee. The outlet is located at the lower edge of the treatment reaction tank 11. The outlet pipe 7 is connected to the outlet. The inlet pipe 5 and the cleaning pipe 6 are powered by a peristaltic pump 17 to transport the water sample to be tested and the cleaning water.

[0069] The interior of the treatment reaction tank 11 is a hollow structure. A water odor detection sensor 4 is fixed at the top of the treatment reaction tank 11. The odor sensor array 51 in the water odor detection sensor 4 is exposed inside the treatment reaction tank 11. An upper heating device is superimposed on the water odor detection sensor 4.

[0070] The PLC controller 3 is used to receive water sample odor data transmitted by the water odor detection sensor 4, process, analyze and store the data in real time, calculate the odor value of the water sample, and classify the odor level.

[0071] The main body of the sample chamber (processing reaction cell 11) is made of stainless steel and has an adjustable volume of 100-300 mL.

[0072] Temperature Heating: The bottom of the sample chamber is heated by an electric heating film, with the temperature controlled by a PID temperature control module, achieving an optimal temperature of 35-50℃. A carbon-based sensor is installed at the top of the sample chamber, also heated by an electric heating film, with the optimal temperature controlled by a PID temperature control module, achieving an optimal temperature of the bottom set temperature + 5℃. Advantages: On one hand, it promotes the volatilization of odor substances in the water into the headspace, improving detection accuracy; on the other hand, it prevents water vapor from adhering to the sample chamber due to high ambient humidity; and during the water heating process, water vapor condenses at the top, affecting the sensor's sensitivity and lifespan.

[0073] Automatic sample introduction: The sample introduction tube is made of polytetrafluoroethylene (PTFE) (to prevent the adsorption of odorous substances). The water inlet is controlled by a diaphragm pump with a flow rate of 500 mL / min, and the sample volume is controlled by the liquid level in the sample chamber. The sample introduction tube uses a stainless steel metal screen (1000 mesh) to intercept particulate matter in the water.

[0074] Cleaning mode: Water rinsing before sample injection, flow rate 500 mL / min, time 2 min; then gas cleaning, using a diaphragm pump, flow rate 1.2 L / min, gas is air filtered through activated carbon, gas cleaning time 5 min.

[0075] It should be noted that the carbon nanotube water odor detection module of this application uses a graphene-based material sensor array (refer to CN202210420150.2). The graphene-based surface continuously adsorbs odor substances released by the water during heating. The adsorbed molecules cause changes in the graphene carrier concentration. Different functionalized materials modify odor substances through different mechanisms, forming unique signal fingerprints. Through the differential response algorithm of the sensor array, accurate calculation of odor substance detection data is achieved. This can also be achieved through patent CN202410030411.9.

[0076] In one embodiment, see Figure 1 As shown, the water body online olfaction instrument based on carbon nanotube sensors adopts a compact structured layout, making the device highly portable and enabling flexible switching between application scenarios such as water sources and water treatment plants.

[0077] The PLC display 2 includes information on the detected water sample, the odor level of the actual test, temperature, humidity, historical test data, odor level fluctuation curve, etc., which can realize real-time display of the water body test site.

[0078] In one embodiment, see Figure 1 As shown, the processing reaction tank 11 is a sealed stainless steel metal container, with the top connected to the upper heating device 9 and the bottom connected to the lower heating device 10, adopting an integrated design; the lower part of the processing reaction tank 11 is connected to a metal support, adopting an elevated arrangement; a magnetic stirrer 16 is installed at the bottom of the processing reaction tank.

[0079] The water inlet pipe 5 and the cleaning pipe 6 are connected by a tee and share a common water inlet, and are controlled by the control module in the PLC controller 3.

[0080] The solenoid valve module 13 for the inlet pipeline and the solenoid valve module 14 for the cleaning pipeline are switched on and off.

[0081] The peristaltic pump module contains two independent power modules, and the control module in the PLC controller 3 controls the delivery of the water sample to be tested and the water sample to be cleaned in the pipeline.

[0082] The drying device 8 uses the heat from the upper heating device 9 to input hot air into the processing reaction tank 11 module to dry the internal space.

[0083] The upper heating device 9 is closely connected to the carbon nanotube water odor detection sensor 4. Its main function is to keep the surface of the sensor array in the carbon nanotube water odor detection sensor 4 dry by continuously outputting heat from the upper heating device 9.

[0084] The upper heating device 9 adopts a temperature adjustable range, which can be flexibly adjusted within a temperature range of 4550℃ according to the usage scenario and needs;

[0085] In one embodiment, see Figure 2 As shown, the data processing and analysis module uses a PLC controller 3, which integrates a control module, a power supply module, a display module, a data processing module, a communication module, and a storage module. The power supply module, display module, data processing module, communication module, and storage module are all electrically connected to the control module, and the display module is also connected to the PLC display 2.

[0086] In one embodiment, see Figure 3 As shown, the inlet flow meter module is connected to the data processing module, which transmits the monitored inlet water data to the data processing module for processing, and displays it on the PLC display 2 through the display module;

[0087] The graphene-based carbon nanotube odor sensor array 51, the temperature sensor module, and the humidity sensor module are connected to the data processing module.

[0088] The graphene-based carbon nanotube odor sensor array 51 transmits the detected odor data to the data processing module; the temperature sensor module transmits the detected temperature data to the data processing module; the humidity sensor module transmits the detected humidity data to the data processing module.

[0089] The temperature data received in the data module is used as a threshold to determine whether to use the odor data detected by the graphene-based carbon nanoparticle odor sensor array 51.

[0090] The humidity data received in the data module is used to determine whether the water pipeline solenoid valve module has reached its threshold after the drying device 8 has finished working.

[0091] In one embodiment, see Figure 3 As shown, the data processing module includes a machine algorithm model based on a neural network;

[0092] The method for constructing the neural network-based machine algorithm model is to generate multi-scale training sample data corresponding to the water sample; and to train the neural network using the multi-scale training sample data to obtain the corresponding neural network model.

[0093] In obtaining the corresponding neural network model, the odor levels are divided according to the odor concentration range. The odor data (sensor signals) monitored by the graphene-based carbon nanotube odor sensor array received in the data processing module are matched to the corresponding odor levels and displayed on the display screen through the display module.

[0094] In one embodiment, see Figure 4 As shown, the inlet pipeline solenoid valve module 13, the cleaning pipeline solenoid valve module 14, and the outlet pipeline solenoid valve module 15 are respectively connected to the control module, and the control module controls the opening and closing of the inlet pipeline solenoid valve module 13, the cleaning pipeline solenoid valve module 14, and the outlet pipeline solenoid valve module 15.

[0095] The upper heating module and the lower heating module are respectively connected to the control module, and the control module controls the opening and closing of the upper heating module and the lower heating module.

[0096] The drying module is connected to the control module. The control module controls the drying module to turn on and off. At the same time, the control module needs to control the opening and closing of the upper heating module.

[0097] The carbon nanotube water odor detection module is connected to the control module, and the control module controls the opening and closing of the carbon nanotube water odor detection module by receiving instructions from the timing module.

[0098] In one embodiment, see Figure 5 As shown, the graphene-based carbon nanotube odor sensor array 51 is a modular array, typically consisting of an array of 4 olfactory chips 52. However, depending on the actual detection requirements, the array of olfactory chips 52 can be flexibly combined in the range of 1 to 8 chips.

[0099] The olfactory chip 52 has 10 data channels, which can transmit 10 detection data to the data processing center at one time.

[0100] After the heating module starts working and reaches the set temperature, the graphene-based carbon nanoparticle odor sensor array 51 is activated and begins to take readings. It takes continuous readings within 20 minutes, and the data processing module performs data processing and analysis after the sensor readings stabilize.

[0101] The method for extracting the characteristic parameters of the graphene-based odor sensor array is as follows: a preset number of cycles is set, and raw response data of the water sample to be extracted is collected under the preset number of cycles. Based on the raw response data, the corresponding sensor signal is constructed (curve is plotted).

[0102] By calculating the peak prominence, peak width, original signal value, noise-reduced signal value, number of peaks, and the preset number of cycles of the sensor signal curve, the signal-to-noise ratio parameters of each sensor are determined, and the sensor with the maximum signal-to-noise ratio is selected as the target sensor.

[0103] The target baseline timing is determined based on the signal characteristics of the target sensor. This timing baseline is then synchronized to the signal curves of all sensors in the graphene-based carbon nanotube odor sensor array to establish an independent sensor baseline for each sensor.

[0104] By comparing the original response data with the corresponding sensor baseline, multidimensional feature parameters of the water sample to be extracted are obtained.

[0105] In this embodiment, the method for detecting odor substances in water is as follows: The inlet flow meter module is adjusted to set the inlet flow rate of the water to be tested. The inlet flow rate setting range is 30-100 mL. The control module sends an opening command to the inlet pipeline solenoid valve 13 module. After the inlet flow rate reaches the set value, the control module controls the inlet pipeline solenoid valve module 13 to close. Simultaneously, it controls the upper heating module, lower heating module, and magnetic stirring module to open. The temperature sensor transmits temperature data back to the data processing module. After the temperature reaches the set threshold, the timer module is activated, with the timer set to 20 seconds. After the timed period ends, the control module activates the carbon nanotube water odor detection module. The graphene-based carbon nanotube odor sensor array 51 continuously monitors the odor data. Once the detection data stabilizes, it is transmitted back to the data processing module. The data processing module processes and analyzes the received odor data, matches it to the corresponding odor level, and displays it on the screen. The detection is then complete. The control module then deactivates the upper heating module, lower heating module, and magnetic stirring module. It also activates the outlet water pipe solenoid valve module 15 and the cleaning pipe solenoid valve module 14, and activates the timer module, setting the timer to 10 minutes. After the timed period ends, the control module deactivates the outlet water pipe solenoid valve module 15 and the cleaning pipe solenoid valve module 14. Simultaneously, it activates the upper heating module and the drying module. The humidity sensor transmits humidity data back to the data processing module. Once the humidity reaches the set threshold, the control module deactivates the upper heating module and the drying module, completing the cleaning process.

[0106] Finally, it should be noted that although the above embodiments have been described in the description and drawings of this patent, this does not limit the scope of patent protection of this patent. Any technical solutions that are based on the essential concept of this patent and are generated by replacing or modifying the equivalent structure or equivalent process described in the description and drawings of this patent, as well as the direct or indirect implementation of the technical solutions of the above embodiments in other related technical fields, are all included within the scope of patent protection of this patent.

Claims

1. An online water olfaction instrument based on a carbon nanotube sensor, comprising a treatment reaction tank, an inlet pipe and an outlet pipe connected to the treatment reaction tank, characterized in that: It also includes a pipeline cleaning device, a water odor detection sensor, and a PLC controller. The water odor detection sensor is electrically connected to the PLC controller. The water odor detection sensor includes a graphene-based carbon nanotube odor sensor array, which is composed of olfactory chips formed by surface modification of graphene material with conductive MOF material arranged in an array. The pipeline cleaning device is connected to the treatment reaction tank through a cleaning pipeline. The treatment reaction tank is equipped with an inlet and an outlet at opposite corners. The inlet is located on the upper side of the treatment reaction tank, and the inlet pipe and the cleaning pipe are connected to the inlet via a tee. The outlet is located at the lower edge of the treatment reaction tank, and the outlet pipe is connected to the outlet. The inlet pipe and the cleaning pipe are powered by peristaltic pumps to transport the water sample to be tested and the cleaning water, respectively. The processing reaction tank has a hollow internal structure. A water odor detection sensor is fixed at the top of the processing reaction tank, and the graphene-based carbon nanoparticle odor sensor array in the water odor detection sensor is exposed inside the processing reaction tank. An upper heating device is superimposed on the water odor detection sensor. A lower heating device and a magnetic stirrer are installed at the bottom of the processing reaction tank. The PLC controller is used to receive water sample odor data transmitted by the water odor detection sensor, process, analyze and store the data in real time, and determine the odor level. The PLC controller includes a control module, and a power supply module, a display module, a data processing module, a communication module, a storage module, and a timing module, which are electrically connected to the control module respectively. The data processing module includes a machine learning model based on linear discriminant analysis; the method for constructing the machine learning model based on linear discriminant analysis and processing data is as follows: Linear discriminant analysis is used to analyze the labeled data to find the optimal coordinate system in high-dimensional space and determine the optimal projection direction. When the sensor signal is received in the corresponding optimal coordinate system, it is projected according to the optimal projection direction. The odor level is determined based on the position in space. The odor data monitored by the graphene-based carbon nanometer odor sensor array received in the data processing module is matched to the corresponding odor level and displayed on the display screen through the display module. In the data processing module, the odor substances in the water are four common odor substances found in drinking water sources: dimethyl isoborneol, geosmin, dimethyl disulfide, and dimethyl trisulfide. The odor level is the concentration range of the four odor substances: Level 1 is set at 0-10 ng / L; Level 2 is set at 10-50 ng / L; Level 3 is set at 50-100 ng / L; and Level 4 is set at 100 ng / L and above.

2. The online water olfaction instrument according to claim 1, characterized in that: A drying device is connected to one side of the upper end of the water odor detection sensor, and the air outlet of the drying device is arranged at the edge of the graphene-based carbon nanoparticle odor sensor array.

3. The online water olfaction instrument according to claim 1, characterized in that: The PLC controller also includes an inlet water pipe solenoid valve module, an outlet water pipe solenoid valve module, a cleaning pipe solenoid valve module, an upper heating module, a lower heating module, a drying module, a magnetic stirring module, a carbon nanotube water odor detection module, and a peristaltic pump module, all electrically connected to the PLC controller control module.

4. The online water olfaction device according to claim 3, characterized in that: The inlet water pipeline solenoid valve module, outlet water pipeline solenoid valve module, and cleaning pipeline solenoid valve module are used to control the opening and closing of the inlet water pipeline, outlet water pipeline, and cleaning pipeline, respectively. The inlet water pipeline solenoid valve module, outlet water pipeline solenoid valve module, and cleaning pipeline solenoid valve module are uniformly controlled by the control module in the PLC controller, and the inlet water pipeline solenoid valve module and the cleaning pipeline solenoid valve module switch between opening and closing. The cleaning pipeline solenoid valve module is used to control the cleaning pipeline for cleaning the treatment reaction tank; The upper heating module is used for upper heating treatment of the water sample to be tested; The lower heating module is used for lower heating treatment of the water sample to be tested; The drying module is used to process the heating and drying of the reaction tank. The magnetic stirring module is used to stir the water body to be tested at a certain rotation speed; The carbon nanotube water odor detection module is used to detect odor substances in the treated water, construct corresponding sensor signals, and send them to the data processing module. The peristaltic pump module is used to control the peristaltic pump to power the delivery of the water sample to be tested and the clean water; The peristaltic pump module contains two independent power modules, and the control module in the PLC controller controls the delivery of the water sample to be tested and the water sample to be cleaned in the pipeline. The data processing module is used to receive the sensor signals of water sample odor transmitted by the water odor detection module, process, analyze and store the sensor signals in real time, and use pattern recognition and machine learning algorithms to divide the concentration and determine the odor level.

5. The online water olfaction device according to claim 2, characterized in that: The water odor detection sensor also includes a temperature sensor and a humidity sensor; the temperature sensor module and humidity sensor module built into the temperature sensor and humidity sensor are electrically connected to the data processing module. The inlet pipe is connected to an inlet flow meter; the inlet flow meter is connected to a data processing module, which transmits the monitored inlet water data to the data processing module for processing, and displays the data on the display screen through the display module. The drying device includes a miniature fan connected to an upper heating device, which provides a heat source for the drying device; the control module controls the opening and closing of the drying module, and at the same time, the control module controls the opening and closing of the upper heating module. The control module controls the opening and closing of the carbon nanotube water odor detection module by receiving instructions from the timing module; the graphene-based carbon nanotube odor sensor array transmits the detected odor data to the data processing module; the temperature sensor module transmits the detected temperature data to the data processing module; and the humidity sensor module transmits the detected humidity data to the data processing module. The temperature data received in the data processing module is used to determine the threshold for whether to use the odor data detected by the graphene-based carbon nanotube odor sensor array; the humidity data received in the data processing module is used to determine the threshold for whether to open the water inlet pipeline solenoid valve module after the drying device has finished working.

6. The online water olfaction device according to claim 1, characterized in that: The method for extracting characteristic parameters of the graphene-based carbon nanotube odor sensor array is as follows: a preset number of cycles is set, and raw response data of the water sample to be extracted is collected at the preset number of cycles. Based on the raw response data, the corresponding sensor signal is constructed.

7. The online water olfaction device according to claim 6, characterized in that: The sensor signals output by the carbon nanotube water odor detection module are arranged sequentially to construct a 154-dimensional vector. Each sample's corresponding sensor signal is equivalent to a point in the 154-dimensional space. This is then analyzed using the in-class scatter matrix S. w and the scatter matrix S between categories b Construct an objective function J(w), and by optimizing the projection direction vector w to maximize J(w), the direction most conducive to classification can be obtained; Where S w Let S be the within-class scatter matrix, C be the number of classes, X be the feature vector of a single sample, μc be the mean vector of the samples in the Cth class, and S be the within-class scatter matrix. b Let μ be the scatter matrix between categories, and μ be the global mean vector of all samples. c -μ is the deviation vector between the mean of class C and the global mean, J(w) is the objective function, representing the ratio of "between-class divergence" to "within-class divergence", and W T It is the transpose of W, W T S b W is S b The projective divergence in the W direction, W T S w W is S w The projective divergence in the W direction, W C Let N be the set of samples of class C. C Let w be the number of samples in class C, and w be the projection direction vector. Linear analytical reasoning (LDA) was performed on the odor detection sensor data of water bodies, and the four categories of odor substances were marked with different colors to generate two-dimensional and three-dimensional dimensionality-reduced scatter plots of category distribution.

8. The online water olfaction instrument according to claim 1, characterized in that: The inlet water pipeline is made of polytetrafluoroethylene (PTFE) and is controlled by a diaphragm pump at a flow rate of 500 mL / min. The injection volume is controlled by a liquid level detection system installed in the reaction tank. Before injection, the inlet water pipeline is flushed with water at a flow rate of 500 mL / min for 2 minutes. Then, gas cleaning is performed using a diaphragm pump at a flow rate of 1.2 L / min. The gas used is air filtered through activated carbon, and the gas cleaning time is 5 minutes.

9. The online water odor detector according to any one of claims 1-8 is used for real-time on-site monitoring of the odor of water at water sources.

Images