Water eutrophication detection equipment based on intelligent sensing system

Abstract

The invention relates to water eutrophication detection equipment based on an intelligent sensing system, in particular to the technical field of environment detection and Internet of Things. The equipment comprises a data acquisition, processing, evaluation, output and maintenance module, water quality detection data is obtained through an online detection equipment probe, after abnormity processing, missing value filling and classified storage, an embedded algorithm model is adopted for water quality category and eutrophication evaluation, data are uploaded in real time and early warning is carried out, algorithm model updating and parameter configuration are supported, and the online detection equipment is used for online detection. The equipment realizes multi-parameter real-time monitoring and intelligent analysis, supports automatic early warning and remote maintenance, improves the detection efficiency and accuracy, effectively reduces the influence of environmental factors on a detection result, improves the equipment stability in combination with a nano protection technology, and provides reliable technical support for water pollution prevention and control.

Description

Technical Field

[0001] The present invention relates to a water eutrophication detection device based on an intelligent sensing system, and in particular to the technical fields of environmental detection and the Internet of Things. Background Art

[0002] With the acceleration of industrialization and urbanization, eutrophication of water bodies has become a key problem in global water environment governance. Traditional water monitoring technologies generally have defects such as limited single-point detection coverage and insufficient multi-parameter collaborative analysis capabilities. They are also significantly disturbed by environmental factors such as water flow disturbances and light changes, resulting in large deviations in detection data. Existing equipment mostly relies on a single sensor to measure water quality parameters and lacks a multi-dimensional data fusion and analysis mechanism, making it difficult to achieve accurate assessment of the degree of eutrophication of water bodies. At the same time, traditional detection equipment generally has problems such as rigid algorithm models, lagging parameter configuration and high maintenance costs, and cannot adapt to dynamically changing water quality monitoring needs.

[0003] Therefore, how to comprehensively utilize multi-source environmental data for water eutrophication detection, effectively reduce the errors caused by environmental factors, and thereby improve detection accuracy and make accurate assessments has become an urgent problem to be solved. In response to the above technical problems, the present invention proposes a water eutrophication detection device based on an intelligent sensing system. Summary of the Invention

[0004] The purpose of the present invention is to provide a water eutrophication detection device based on an intelligent sensing system to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: A water eutrophication detection device based on an intelligent sensing system includes a data acquisition module, a data processing module, an evaluation module, a result output module, and a maintenance module. The data acquisition module, data processing module, evaluation module, result output module, and maintenance module are operated by powering a solar battery. The data acquisition module is used to obtain water quality detection data through the online monitoring device probe equipped with the protection component; The data processing module is used to perform exception processing on the acquired water quality detection data, perform missing value processing on the data after exception processing, divide the water quality detection data after missing value processing into water quality data and enriched oxidation data, and store them in the local storage medium SD card; The evaluation module is used to obtain an evaluation basis by running an embedded algorithm model on the water quality data and eutrophication data in the local storage medium SD card through an embedded processor, make a water quality category evaluation and a eutrophication evaluation based on the evaluation basis, and store them in the local storage medium SD card; The result output module is used to upload the water quality data, enrichment and oxidation data, water quality category evaluation and eutrophication evaluation in the local storage medium SD card to the cloud database in real time through the MQTT protocol support module. At the same time, according to the water quality category evaluation or eutrophication evaluation, it is determined whether to issue an early warning device warning and a 4G module notification; The maintenance module is used to update the water quality category evaluation and eutrophication evaluation embedded algorithm models stored in the non-volatile memory EEPROM through the OTA update component, as well as the device parameter configuration.

[0006] Preferably, in the data acquisition module, the process of obtaining water quality test data through the online monitoring device probe equipped with the protection component is: The online detection device probe is a sensor device, which includes a sensor chamber for detecting water quality detection data. The sensor chamber includes: an electrochemical sensor for measuring the dissolved oxygen concentration in water through electrode reaction, an ultraviolet spectrum sensor for indirectly calculating the permanganate index through spectral analysis, a spectral analysis sensor for calculating total phosphorus and total nitrogen by measuring absorbance at a specific wavelength using a standard curve and spectrophotometry, a fluorescence sensor for inverting chlorophyll-a concentration through fluorescence intensity, an optical scattering sensor for determining water turbidity by measuring the degree of light scattering by suspended particles in water, and an ion-selective electrode for measuring ammonia nitrogen in a solution based on the response of an ion-selective membrane to specific ions; The online detection equipment probes, equipped with protective components including nano-titanium dioxide photocatalytic coating and IP68 waterproof shell, are deployed in the center of the river by means of poles and in the center of the lake and the sea surface by means of buoys. The water quality detection data including dissolved oxygen DO, permanganate index CODMn, total phosphorus TP, total nitrogen TN, ammonia nitrogen NH3-N, chlorophyll a, turbidity and transparency SD are collected at a set sampling frequency f.

[0007] Preferably, in the data processing module, the process of processing outliers and missing values ​​of the acquired water quality test data and the process of obtaining water quality data and enriched oxygen data are as follows: For the acquired water quality test data, a box plot is drawn, and the upper and lower limits in the box plot are used to identify extreme values ​​in the water quality test data, and the water quality test data exceeding the upper and lower limits are deleted; The box plot is a tool for visualizing data distribution by displaying the quartiles and median of the data. The upper limit in the box plot is: upper limit = Q3 + 1.5 * IQR, and the lower limit in the box plot is: lower limit = Q1 - 1.5 * IQR, where Q3 is the upper quartile representing the 75th percentile of the data, Q1 is the lower quartile representing the 25th percentile of the data, and IQR is the interquartile range representing the range of variation of the data, and IQR = Q3 - Q1; The water quality test data that has been processed for outliers are used to estimate missing values ​​according to the linear relationship using the linear interpolation formula to fill in the missing values; The linear interpolation formula is: ; in, is the missing value to be filled at time point t, is the moment t-1 before time point t, and t is the time point corresponding to the data that needs to fill the missing value. is the time point before t, is the moment after time point t, is the known data value at time point t+1; The water quality test data with outlier and missing value processing were divided into water quality data including dissolved oxygen, permanganate index, total phosphorus, total nitrogen and ammonia nitrogen, and enriched oxygenation data including chlorophyll a, permanganate index, transparency, total phosphorus and total nitrogen.

[0008] Preferably, in the evaluation module, the process of obtaining the evaluation basis by running the embedded algorithm model through the embedded processor for the water quality data and the enrichment and oxidation data in the local storage medium SD card is as follows: Water quality data is read through local storage media, and the "Surface Water Environmental Quality Standard" (GB3838-2002) or "Seawater Quality Standard" (GB3097-1997) stored in the local storage media is automatically associated. The excess multiple is obtained through the water quality category evaluation embedded algorithm model, which is used to measure the water quality category evaluation. The "Surface Water Environmental Quality Standard" (GB3838-2002) is the core standard for surface water quality management in my country. According to the water quality, surface water is divided into Class I to Class V. The surface water classification thresholds are: Class I water has dissolved oxygen greater than 7.5 mg / L, permanganate index less than 2 mg / L, total phosphorus less than 0.02 mg / L, total nitrogen less than 0.2 mg / L and ammonia nitrogen less than 0.15 mg / L; Class II water has dissolved oxygen greater than 6mg / L, permanganate index less than 4mg / L, total phosphorus less than 0.1mg / L, total nitrogen less than 0.5mg / L and ammonia nitrogen less than 0.5mg / L; Class III water has dissolved oxygen greater than 5mg / L, permanganate index less than 6mg / L, total phosphorus less than 0.2mg / L, total nitrogen less than 1.0mg / L and ammonia nitrogen less than 1.0mg / L; Class IV water has dissolved oxygen greater than 3mg / L, permanganate index less than 10mg / L, total phosphorus less than 0.3mg / L, total nitrogen less than 1.5mg / L and ammonia nitrogen less than 1.5mg / L; Class V water has dissolved oxygen greater than 2mg / L, permanganate index less than 15mg / L, total phosphorus less than 0.4mg / L, total nitrogen less than 2.0mg / L and ammonia nitrogen less than 2.0mg / L; The Seawater Quality Standard (GB3097-1997) is the core standard for my country's seawater quality management. It divides seawater into four categories based on water quality. The thresholds for seawater classification are as follows: Class I seawater has a dissolved oxygen level greater than 6 mg / L, a permanganate index less than 2 mg / L, a total phosphorus level less than 0.015 mg / L, a total nitrogen level less than 0.2 mg / L, and ammonia nitrogen level less than 0.02 mg / L; Class II seawater has a dissolved oxygen level greater than 5 mg / L, a permanganate index less than 3 mg / L, a total phosphorus level less than 0. Class III seawater has dissolved oxygen greater than 4 mg / L, permanganate index less than 4 mg / L, total phosphorus less than 0.045 mg / L, total nitrogen less than 0.4 mg / L, and ammonia nitrogen less than 0.04 mg / L; Class IV seawater has dissolved oxygen greater than 3 mg / L, permanganate index less than 5 mg / L, total phosphorus less than 0.045 mg / L, total nitrogen less than 0.5 mg / L, and ammonia nitrogen less than 0.04 mg / L; Eutrophication data is read from local storage media, and a comprehensive TLI value is obtained through the eutrophication evaluation embedded algorithm to measure the eutrophication evaluation; The water quality category evaluation embedded algorithm model is:

[0009] Among them, the highest water quality category standard value is the Class V standard value for surface water or the Class IV standard value for seawater; The eutrophication assessment embedded algorithm for obtaining the comprehensive TSI value is:

[0010] Among them, TLI is the comprehensive TSI value, is the weight coefficient, i includes chlorophyll a, permanganate index, transparency, total phosphorus and total nitrogen, the weight of chlorophyll a (Chl-a) is 0.4, the permanganate index (COD Mn ) The weight of transparency (SD) is 0.1 The weight of total phosphorus (TP) is 0.1. The weight of total nitrogen (TN) is 0.2 is 0.2, It is a single-item nutritional status index including chlorophyll a, permanganate index, transparency, total phosphorus and total nitrogen; The individual nutritional status indices including chlorophyll a, permanganate index, transparency, total phosphorus and total nitrogen are:

[0011]

[0012]

[0013]

[0014] Among them, TSI (TP) is the nutritional status index of the total phosphorus single indicator, TSI (TN) is the nutritional status index of the total nitrogen single indicator, TSI (SD) is the nutritional status index of the transparency single indicator, TSI (CODMn) is the nutritional status index of the permanganate index single indicator, and TSI (Chl-a) is the nutritional status index of the chlorophyll a single indicator.

[0015] Preferably, in the evaluation module, the process of making water quality category evaluation and eutrophication evaluation based on the results is: The dissolved oxygen, permanganate index, total phosphorus, total nitrogen, and ammonia nitrogen in the water quality data are evaluated using the water quality category assessment embedded algorithm model to obtain their respective exceedance multiples. If any exceedance multiple is greater than 0, the water quality category is directly determined. The seawater quality category is evaluated as Class IV, and the surface water quality category is evaluated as Class V. If multiple exceedance multiples are greater than 0, the data with the largest exceedance multiple is used as the basis for judgment; If the exceedance multiples are all less than zero, the dissolved oxygen, permanganate index, total phosphorus, total nitrogen, and ammonia nitrogen in the water quality data are matched with the surface water classification thresholds given in the "Surface Water Environmental Quality Standard" (GB3838-2002) or the seawater classification thresholds given in the "Seawater Quality Standard" (GB3097-1997), and the worst category among all the data is taken as the water quality category evaluation result; For the eutrophication data including chlorophyll a, transparency and turbidity, a comprehensive TLI value is obtained through the eutrophication assessment embedded algorithm model, and the eutrophication assessment is obtained based on the comprehensive TLI value; If the comprehensive TLI value is less than or equal to 30, the eutrophication is assessed as oligotrophic; If the comprehensive TLI value is greater than 30 and less than 50, the eutrophication is assessed as mesotrophic; If the comprehensive TLI value is greater than 50 and less than 60, the eutrophication is assessed as mild eutrophication; If the comprehensive TLI value is greater than 60 and less than 70, the eutrophication is assessed as moderate eutrophication; If the comprehensive TLI value is greater than or equal to 70, the eutrophication is assessed as severe eutrophication.

[0016] Preferably, in the result output module, the process of uploading the water quality data, eutrophication data, water quality category evaluation and eutrophication evaluation stored in the local storage medium SD card to the cloud database in real time through the MQTT protocol support module is: The data stored in the local storage medium, including water quality data, eutrophication data, water quality category evaluation, and eutrophication evaluation, are encapsulated in JSON format and compressed using the Zlib library. The compressed data is then uploaded to the cloud database via the MQTT protocol supported by the MQTT protocol support module. The Zlib library is a data compression library suitable for embedded devices and low-resource environments; the MQTT protocol is a lightweight communication protocol based on a publish or subscribe model; and the JSON format is a lightweight data exchange format.

[0017] Preferably, in the result output module, the process of determining whether to perform an early warning device warning and a 4G module notification according to the water quality category evaluation or the eutrophication evaluation is as follows: Set the water quality category evaluation and eutrophication evaluation warning thresholds. Each time the device completes data collection and evaluation, it immediately determines whether the surface water quality category evaluation is greater than Category IV, the seawater quality category evaluation is greater than Category III, or the eutrophication evaluation is mild, moderate, or severe. If the surface water quality category evaluation is greater than Category IV, the seawater quality category evaluation is greater than Category III, or the eutrophication evaluation is mild, moderate, or severe, the device buzzer of the warning device will sound continuously for 10 seconds, repeated at intervals of 5 seconds, until manual intervention or data returns to normal, the LED indicator of the warning device will flash in synchronization with the buzzer, and the device will send a text message to the administrator via the 4G module. The text message content contains device location information, water quality data, eutrophication data, water quality category evaluation, and eutrophication evaluation.

[0018] Preferably, in the maintenance module, the process of updating the water quality category evaluation and eutrophication evaluation embedded algorithm model is as follows: The administrator updates the embedded algorithm models for water quality category evaluation and eutrophication evaluation based on actual situations, including the discovery of defects in the algorithm model, changes in water quality monitoring standards, and the acquisition of better algorithm improvement plans. The administrator selects a new algorithm model and specifies the device equipped with the OTA update component for the updated algorithm model. The administrator sends an update package to the device equipped with the OTA update component for the specified updated algorithm model through the cloud platform. After the device receives the update package, it stores the update package in the device's non-volatile memory EEPROM. At the same time, the device backs up the algorithm model currently running on the device, and the device writes the new algorithm model data in the update package to the device's program storage area, overwriting the original algorithm model. The device automatically starts to verify the new algorithm model using the built-in test data. If the verification passes, it sends an update success message to the cloud platform and deletes the backed-up algorithm model and update package. If the verification fails, the device restores the original algorithm model based on the backup data and sends an update failure message to the cloud platform. The OTA update component is a hardware interface that supports the over-the-air (OTA) technology.

[0019] Preferably, in the maintenance module, the process of configuring device parameters is as follows: Remote configuration of equipment parameters is achieved through the cloud platform. The online detection equipment probe is configured to collect water quality test data at a frequency f ranging from 15 minutes to 24 hours per time. Water quality category evaluation and eutrophication evaluation warning thresholds are set. The water quality category evaluation warning for drinking water sources can be set to a Class III trigger warning, and the water quality category evaluation warning for bathing beaches can be set to a Class II trigger warning. For electrochemical sensors and spectral analysis sensors, calibration coefficients are configured to correct the accuracy of sensor measurement data and the measurement range is used to determine the effective measurement parameter value range. The measurement range can be dynamically adjusted according to actual conditions. The device receives new parameter configuration information from the cloud platform through the OTA update component, saves the new parameter configuration information to the device's non-volatile memory EEPROM, and restarts the program according to the new parameter configuration.

[0020] Preferably, in the intelligent prediction module, based on real-time meteorological data and water quality data and oxygen enrichment data in the local storage medium SD card, the embedded processor runs a pre-trained algal bloom prediction model to predict the algal bloom risk in the next 48 hours as follows: Real-time meteorological data including water temperature, light intensity, wind speed and direction, and rainfall are obtained from the cloud platform through the 4G module. At the same time, water quality data and oxygen enrichment data of the past 72 hours are read from the SD card. The real-time meteorological data and the water quality data and oxygen enrichment data of the latest 72 hours are input into the pre-trained algal bloom prediction model. Based on this, the input layer forms a 72×8 time series feature matrix with a tensor format of [1,72,8]. The 64-dimensional hidden features are extracted through two LSTM hidden layers with 64 neurons in each layer and the activation function of ReLU. Finally, the fully connected layer of the output layer is mapped into a 48-dimensional probability sequence and normalized to the [0,1] interval through the Sigmoid function to obtain the hourly algal bloom probability sequence for the next 48 hours. ; If there are three consecutive hours , the algal bloom risk in the next 48 hours is high risk algal bloom; If there is any 5, the algal bloom risk in the next 48 hours is low risk algal bloom; If there is any 5, the algal bloom risk in the next 48 hours is a potential algal bloom risk.

[0021] Preferably, under the condition that the evaluation module is replaced by the intelligent prediction module, the result output module uploads the algal bloom risk within the next 48 hours obtained by the intelligent prediction module to the cloud database in real time through the MQTT protocol support module, and the process of determining whether to issue an early warning device warning and a 4G module notification based on the algal bloom risk within the next 48 hours obtained by the intelligent prediction module is as follows: The hourly algal bloom probability sequence for the next 48 hours generated by the intelligent prediction module, the algal bloom risk within the next 48 hours, and the corresponding timestamp are encapsulated in JSON format, compressed by the Zlib library, and uploaded to the cloud database at QoS1 level using the MQTT protocol support module. At the same time, if the algal bloom risk within the next 48 hours is high-risk algal bloom, the warning device buzzer will sound and the LED light will flash at a high frequency. At the same time, a text message containing the device location information and the risk period corresponding to the high-risk algal bloom will be sent to the administrator through the 4G module. The risk period corresponding to the high-risk algal bloom is within 48 hours. A continuous period of more than or equal to three hours.

[0022] Preferably, under the condition that the evaluation module is replaced by the intelligent prediction module, the process of the maintenance module updating the pre-trained algal bloom prediction model stored in the non-volatile memory EEPROM through the OTA update component is as follows: The administrator updates the pre-trained algal bloom prediction model based on actual conditions, including discovering defects in the pre-trained algal bloom prediction model or obtaining a better algorithm improvement solution. The administrator selects a new model and specifies a device equipped with an OTA update component for the updated model. The administrator sends an update package to the device equipped with the OTA update component for the specified updated model through the cloud platform. When the device receives the update package, it stores the update package in the device's non-volatile memory EEPROM. At the same time, the device backs up the model currently running on the device. The device writes the new model data in the update package to the device's program storage area, overwriting the original model. The device automatically starts to verify the new model using the built-in test data of the device. If the verification passes, it sends an update success message to the cloud platform and deletes the backed-up model and update package. If the verification fails, the device restores the original model based on the backup data and sends an update failure message to the cloud platform.

[0023] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art: 1. The present invention realizes multi-parameter collaborative monitoring and intelligent data processing. It realizes the integration of multiple types of sensors and the synchronous acquisition of parameters through the sensor compartment of the online monitoring equipment probe, breaking through the limitation of traditional equipment relying on a single sensor. The collected data is automatically identified based on the box plot method. The linear interpolation method is combined to fill the missing data, reducing manual intervention. The data preprocessing efficiency is improved by more than 40%, ensuring the data quality of the input algorithm. At the same time, through the nano-titanium dioxide photocatalytic coating and IP68 waterproof design, it effectively resists biological attachment, corrosion and underwater environmental interference, improves the long-term stability of the sensor in complex water bodies, and reduces the measurement error to below ±0.1mg / L.

[0024] 2. The present invention enables precise assessment of water quality and eutrophication, as well as remote maintenance of equipment. The embedded algorithm for water quality classification evaluation is directly linked to the "Surface Water Environmental Quality Standard" (GB3838-2002) or the "Seawater Quality Standard" (GB3097-1997). By calculating the excess multiples and determining the worst category, it achieves automated water quality grading with a classification accuracy of up to 98%. The embedded algorithm for eutrophication evaluation integrates the weights of chlorophyll a, total phosphorus, total nitrogen, permanganate index, and transparency to calculate the eutrophication level in real time, shortening the evaluation time from several hours required for traditional manual analysis to seconds. Furthermore, the equipment supports online updating of algorithm models to repair defects or adapt to new national standards, eliminating the need for on-site equipment disassembly, reducing maintenance costs by 60%. Furthermore, the sensor sampling frequency f, alarm threshold, and calibration coefficient can be remotely configured to flexibly meet different monitoring needs.

[0025] 3. The present invention highly integrates data acquisition, data processing, evaluation, result output and maintenance modules. Through the linkage of various hardware and embedded algorithm models, it constructs a comprehensive and efficient water quality evaluation and eutrophication evaluation rapid and accurate control system, which realizes real-time and accurate monitoring and evaluation of water quality and eutrophication conditions of different water bodies, quickly responds to abnormal situations and issues early warnings. At the same time, the model can be continuously optimized to adapt to the changing monitoring environment, and can serve as a command platform for rapid and accurate control of water quality evaluation and eutrophication evaluation.

[0026] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0027] Figure 1 Schematic diagram of the functional modules of the present invention; Figure 2 It is a structural diagram of the present invention; Figure 3 It is a working diagram of the equipment of the present invention. Specific embodiments

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0029] Example 1, as Figure 1 The water eutrophication detection device based on the intelligent sensing system includes a data acquisition module, a data processing module, an evaluation module, a result output module and a maintenance module, and each module works together to complete the water eutrophication detection and evaluation.

[0030] A data acquisition module is used to obtain water quality test data through an online monitoring device probe equipped with a protective component; The data processing module is used to perform exception processing on the acquired water quality test data, perform missing value processing on the data after exception processing, divide the water quality test data after missing value processing into water quality data and enriched oxidation data, and store them in the local storage medium SD card; The evaluation module is used to obtain an evaluation basis by running an embedded algorithm model on the water quality data and eutrophication data in the local storage medium SD card through the embedded processor, make a water quality category evaluation and a eutrophication evaluation based on the evaluation basis, and store them in the local storage medium SD card; The result output module is used to upload the water quality data, enrichment and oxidation data, water quality category evaluation and eutrophication evaluation in the local storage medium SD card to the cloud database in real time through the MQTT protocol support module. At the same time, according to the water quality category evaluation or eutrophication evaluation, it determines whether to issue an early warning device warning and 4G module notification; The maintenance module is used to update the water quality category evaluation and eutrophication evaluation embedded algorithm models stored in the non-volatile memory EEPROM and configure device parameters through OTA update components.

[0031] Furthermore, the working principle of the present invention is described below by way of examples: Assuming a tributary of the Yangtze River, which is a source of drinking water, an online detection equipment probe was deployed in the center of the lake using a buoy type, and an online detection equipment probe was deployed in the center of the river using a pole type. Water quality test data was collected every hour for 7 consecutive days. Some of the raw data collected are as follows: DO was 5.8mg / L, 6.2mg / L and 5.5mg / L in three hours; COD MnWithin three hours, they were 4.2 mg / L, 3.8 mg / L and 4.5 mg / L respectively; TP within three hours were 0.12 mg / L, 0.10 mg / L and 0.15 mg / L respectively; TN within three hours were 1.8 mg / L, 1.6 mg / L and 2.0 mg / L respectively; NH3-N within three hours were 1.2 mg / L, 1.0 mg / L and 1.3 mg / L respectively; Chl-a within three hours were 15.6 μg / L, 18.2 μg / L and 20.5 μg / L respectively; turbidity within three hours were 20.5 NTU, 22.1 NTU and 19.8 NTU respectively; transparency within three hours were 1.2 m, 1.1 m and 1.3 m respectively.

[0032] Through the box plot, TN is 2.0 mg / L, exceeding the Class V standard of 2.0 mg / L, marked as an outlier and deleted, and missing value filling is performed. The missing DO value is filled by linear interpolation. For example, the data for the third hour is missing, and the interpolation value is 5.7 mg / L based on the previous and subsequent data. The processed data is classified. The water quality data includes: DO is 5.7 mg / L, CODMn is 4.2 mg / L, TP is 0.12 mg / L, TN is 1.8 mg / L and NH3-N is 1.2 mg / L; the rich oxygen data includes: Chl-a is 15.6 μg / L, CODMn is 4.2 mg / L, TP is 0.12 mg / L, TN is 1.8 mg / L and transparency is 1.2m.

[0033] The water quality data and eutrophication data were respectively obtained by the water quality category evaluation embedded algorithm model and the eutrophication evaluation embedded algorithm model, and the results were matched with the threshold values ​​specified in the "Surface Water Environmental Quality Standard" (GB3838-2002) to obtain DO as Class IV and COD as Class IV. Mn The worst category is Class IV, TP is Class III, TN is Class IV and NH3-N is Class IV. The water quality category of a tributary of the Yangtze River is Class IV. The comprehensive TLI value is obtained through the eutrophication evaluation embedded algorithm model: TLI=0.4×53.2+0.2×58.5+0.2×70.1+0.1×45.3+0.1×41.2=57.8. Based on this, the eutrophication evaluation of a tributary of the Yangtze River is slightly eutrophic.

[0034] The local water quality data, enrichment and oxidation data, water quality category evaluation, and eutrophication evaluation are encapsulated in JSON and compressed to 500 bytes. They are then uploaded to the cloud via the MQTT protocol. Since the water quality category is IV, the device buzzer alarm is triggered, the LED flashes, and a text message notification is sent. The text message content includes: device ID: 1234, location: 30.2°N, 120.5°E, water quality category: IV, eutrophication level: mild eutrophication, comprehensive TLI value: 57.8, and time: 2025-01-01 08:00.

[0035] The administrator discovered a deviation in the TN calculation model and pushed a revised algorithm. The device received the update package and automatically replaced the old model after verification. At the same time, the sampling frequency f was adjusted from 1 hour to 30 minutes based on actual conditions. After the calibration coefficient was updated, the DO measurement error was reduced from ±0.2 mg / L to ±0.1 mg / L.

[0036] The above is a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily think of modifications or replacements within the technical scope disclosed in the present invention, which should be included in the scope of protection of the present invention.

[0037] Example 2, as Figure 3 The workflow of a water eutrophication detection device based on an intelligent sensing system is as follows: Solar batteries power the equipment, ensuring the normal operation of all hardware modules and starting the detection process. The online monitoring equipment probe containing the sensor compartment and protective components collects data on surface water or seawater. The multiple sensors in the sensor compartment simultaneously obtain water quality test data including dissolved oxygen DO, permanganate index CODMn, total phosphorus TP, total nitrogen TN, ammonia nitrogen NH3-N, chlorophyll a (Chl-a), turbidity and transparency SD. After processing, the collected water quality test data is stored in the local storage medium SD card. The embedded processor calls the data from the SD card and runs the water quality category evaluation algorithm and eutrophication evaluation by associating it with the GB3838 or GB3097 standards. The algorithm outputs the water quality category evaluation and eutrophication evaluation to the local storage medium SD card. Through the MQTT protocol support module, the water quality test data and evaluation results are uploaded to the cloud platform and stored in the cloud database. At the same time, when the embedded processor detects that the water quality category evaluation or eutrophication evaluation exceeds the standard, it controls the 4G module to send a text message containing the location, data, and evaluation results to the administrator. By controlling the early warning equipment, it triggers the buzzer to sound and the LED indicator to flash, realizing on-site warning. When the cloud platform sends an algorithm or parameter update instruction, the OTA update component receives the update package, backs up the original model through the non-volatile memory EEPROM, writes the new model and verifies it.

Claims

1. A water eutrophication detection device based on an intelligent sensing system, characterized in that: include: A data acquisition module, a data processing module, an evaluation module, a result output module and a maintenance module, wherein the data acquisition module, the data processing module, the evaluation module, the result output module and the maintenance module are operated by powering a solar battery; The data acquisition module is used to obtain water quality detection data through the online monitoring device probe equipped with the protection component; The data processing module is used to perform exception processing on the acquired water quality detection data, perform missing value processing on the data after exception processing, divide the water quality detection data after missing value processing into water quality data and enriched oxidation data, and store them in the local storage medium SD card; The evaluation module is used to obtain an evaluation basis by running an embedded algorithm model on the water quality data and eutrophication data in the local storage medium SD card through an embedded processor, make a water quality category evaluation and a eutrophication evaluation based on the evaluation basis, and store them in the local storage medium SD card; The result output module is used to upload the water quality data, enrichment and oxidation data, water quality category evaluation and eutrophication evaluation in the local storage medium SD card to the cloud database in real time through the MQTT protocol support module. At the same time, according to the water quality category evaluation or eutrophication evaluation, it is determined whether to issue an early warning device warning and a 4G module notification; The maintenance module is used to update the water quality category evaluation and eutrophication evaluation embedded algorithm models stored in the non-volatile memory EEPROM through the OTA update component, as well as the device parameter configuration.

2. The water eutrophication detection device based on the intelligent sensing system according to claim 1, characterized in that: In the data acquisition module, the process of obtaining water quality test data through the online monitoring device probe equipped with the protection component is as follows: The online monitoring equipment probe equipped with the protective component is equipped with a sensor compartment, and is also equipped with a protective component including a nano-titanium dioxide photocatalytic coating and an IP68 waterproof shell. The sensor compartment contains an electrochemical sensor, an ultraviolet spectrum sensor, a spectrum analysis sensor, a fluorescence sensor, an optical scattering sensor and an ion selective electrode; The online detection equipment probe equipped with protective components is deployed in the center of the lake, the center of the river or the sea surface to collect water quality detection data including dissolved oxygen DO, permanganate index CODMn, total phosphorus TP, total nitrogen TN, ammonia nitrogen NH3-N, chlorophyll a, turbidity and transparency SD at a set sampling frequency f.

3. The water eutrophication detection device based on the intelligent sensing system according to claim 2, characterized in that: In the data processing module, the process of processing abnormal values ​​and missing values ​​of the acquired water quality detection data and obtaining water quality data and enrichment and oxidation data is as follows: For the acquired water quality test data, we draw a box plot and use the upper and lower limits in the box plot to identify extreme values ​​in the water quality test data. We delete the water quality test data that exceeds the upper and lower limits, and fill the missing values ​​with the water quality test data that have been processed for outliers through linear interpolation. The water quality test data after missing value processing is divided into water quality data and enriched oxygen data, wherein the water quality data includes dissolved oxygen, permanganate index, total phosphorus, total nitrogen and ammonia nitrogen, and the enriched oxygen data includes chlorophyll a, permanganate index, transparency, total phosphorus and total nitrogen; The box plot is a tool for visualizing the distribution of data by showing the quartiles and median of the data.

4. The water eutrophication detection device based on the intelligent sensing system according to claim 3 is characterized in that: In the evaluation module, the process of obtaining the evaluation basis by running the embedded algorithm model on the water quality data and the enrichment and oxidation data in the local storage medium SD card through the embedded processor is as follows: Water quality data and eutrophication data are read from the local storage medium SD card, and the water quality data is automatically associated with the surface water environmental quality standards or seawater quality standards stored in the local storage medium. The dissolved oxygen, permanganate index, total phosphorus, total nitrogen and ammonia nitrogen in the water quality data are embedded in the algorithm model through the water quality category evaluation to obtain the respective exceedance multiples; The comprehensive TLI value is obtained by embedding the eutrophication assessment algorithm model into the eutrophication data including chlorophyll a, permanganate index, transparency, total phosphorus and total nitrogen; Based on this, the evaluation and judgment basis including the comprehensive TLI value and the multiples of dissolved oxygen, permanganate index, total phosphorus, total nitrogen and ammonia nitrogen exceeding the standard were obtained.

5. The water eutrophication detection device based on the intelligent sensing system according to claim 4 is characterized in that: In the evaluation module, the process of making water quality category evaluation and eutrophication evaluation based on the evaluation judgment basis is as follows: For each of the multiples exceeding the standard for dissolved oxygen, permanganate index, total phosphorus, total nitrogen, and ammonia nitrogen, if any of the multiples exceeding the standard is greater than 0, the water quality category is directly determined. The water quality category of seawater is evaluated as Category IV, and the water quality category of surface water is evaluated as Category V. If multiple multiples exceeding the standard are greater than 0, the data with the largest multiple exceeding the standard is used as the basis for judgment; If the exceedance multiples are all less than zero, the dissolved oxygen, permanganate index, total phosphorus, total nitrogen, and ammonia nitrogen in the water quality data are matched with the surface water classification thresholds given in the surface water environmental quality standards or the seawater classification thresholds given in the seawater quality standards, and the worst category among all the data is taken as the water quality category evaluation result; Eutrophication evaluation is made based on the comprehensive TLI value. If the comprehensive TLI value is less than or equal to 30, the eutrophication evaluation is oligotrophic; If the comprehensive TLI value is greater than 30 and less than 50, the eutrophication is assessed as mesotrophic; If the comprehensive TLI value is greater than 50 and less than 60, the eutrophication is assessed as mild eutrophication; If the comprehensive TLI value is greater than 60 and less than 70, the eutrophication is assessed as moderate eutrophication; If the comprehensive TLI value is greater than or equal to 70, the eutrophication is assessed as severe eutrophication.

6. The water eutrophication detection device based on the intelligent sensing system according to claim 5, characterized in that: In the result output module, the water quality data, eutrophication data, water quality category evaluation and eutrophication evaluation in the local storage medium SD card are uploaded to the cloud database in real time through the MQTT protocol support module, and the process of determining whether to issue an early warning device warning and a 4G module notification is as follows: The data stored in the local storage medium SD card, including water quality data, enrichment and oxidation data, water quality category evaluation, and eutrophication evaluation, are encapsulated in JSON format and compressed using the Zlib library. The compressed data is uploaded to the cloud database via the MQTT protocol supported by the MQTT protocol support module; Set the water quality category evaluation and eutrophication evaluation warning thresholds. When the surface water quality category evaluation is greater than Category IV, the seawater quality category evaluation is greater than Category III, or the eutrophication evaluation is mild, moderate, or severe eutrophication, the device buzzer of the warning device sounds, the LED indicator flashes, and at the same time, the device sends a text message to the administrator via the 4G module.

7. The water eutrophication detection device based on the intelligent sensing system according to claim 6, characterized in that: In the maintenance module, the process of updating the water quality category evaluation and eutrophication evaluation embedded algorithm models stored in the non-volatile memory EEPROM and configuring the equipment parameters is as follows: Based on defects in the algorithm model, changes in water quality monitoring standards, and the availability of better improvement plans, the administrator updates the algorithm models embedded in water quality classification evaluation and eutrophication evaluation. The administrator selects the new algorithm model, specifies the device, and sends the update package to the device OTA update component through the cloud platform. After receiving the update package, the device stores it in the EEPROM, backs up the current algorithm model, writes the new algorithm model into the program storage area to overwrite the original algorithm model, and automatically starts the built-in test data to verify the new algorithm model. If the verification is successful, an update success message is sent and the backup and update package are deleted. If the verification fails, the original algorithm model is restored according to the backup and an update failure message is sent. Configure the online detection equipment probe on the cloud platform to collect water quality test data at a frequency f, and set water quality category evaluation and eutrophication evaluation warning thresholds; The device OTA update component receives the new parameter configuration including the collection frequency f, water quality category evaluation warning threshold and eutrophication evaluation warning threshold issued by the cloud platform and saves it to EEPROM. The device restarts the program according to the new parameter configuration.

8. The water eutrophication detection device based on the intelligent sensing system according to claim 1, characterized in that: The evaluation module can be replaced by an intelligent prediction module, which is used to predict the risk of algal blooms within the next 48 hours based on real-time meteorological data and water quality data and oxygen enrichment data stored in the local storage medium SD card, and to run a pre-trained algal bloom prediction model through an embedded processor; The pre-trained algal bloom prediction model includes: The input layer is used to convert real-time meteorological data and water quality data and oxygen enrichment data within 72 hours in the local storage medium SD card into a time series matrix; Two LSTM hidden layers are used to extract 64-dimensional hidden features based on the time series matrix; The output layer is used to map the 64-dimensional hidden features to a 48-dimensional probability sequence and generate the algal bloom probability sequence for the next 48 hours through the Sigmoid activation function.

9. The water eutrophication detection device based on the intelligent sensing system according to claim 8, characterized in that: The process of running the pre-trained algal bloom prediction model on the embedded processor to predict the algal bloom risk within the next 48 hours is as follows: Read the latest 72-hour water quality and oxygen enrichment data from the local storage medium SD card, and obtain real-time meteorological data through the 4G module. Input the real-time meteorological data and the latest 72-hour water quality and oxygen enrichment data into the pre-trained algal bloom prediction model to obtain the algal bloom probability sequence for the next 48 hours. ; If there are three consecutive hours , the algal bloom risk in the next 48 hours is high risk algal bloom; If there is any 5, the algal bloom risk in the next 48 hours is low risk algal bloom; If there is any 5, the algal bloom risk in the next 48 hours is a potential algal bloom risk.

10. The water eutrophication detection device based on the intelligent sensing system according to claim 9, characterized in that: When the evaluation module is replaced by the intelligent prediction module, the result output module is used to upload the algal bloom risk within the next 48 hours obtained by the intelligent prediction module to the cloud database in real time through the MQTT protocol support module. At the same time, based on the algal bloom risk within the next 48 hours obtained by the intelligent prediction module, it is determined whether to issue an early warning device warning and a 4G module notification; The maintenance module is used to update the pre-trained algal bloom prediction model stored in the non-volatile memory EEPROM through the OTA update component.

Images