Water quality eutrophication dynamic monitoring method based on multi-source sensor data fusion
By fusing data from multiple sensor sources, selecting feature points, and setting sampling parameters, the cost and stability issues of large-area water eutrophication monitoring were resolved, achieving efficient and economical water eutrophication assessment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 河南省水文水资源测报中心
- Filing Date
- 2025-12-12
- Publication Date
- 2026-05-12
AI Technical Summary
Existing technologies make it difficult to achieve accurate eutrophication monitoring in large water bodies, especially in local high-risk areas. Furthermore, the cost of sensor deployment is high, and the water flow velocity leads to unstable monitoring results.
A multi-source sensor data fusion method is adopted to generate benchmark indicators, screen feature points, set the number of sampling times and intervals, form a compensation function, and integrate monitoring data to assess water eutrophication, reduce the number of sensors and adapt to water flow velocity.
While ensuring overall monitoring accuracy, the cost of sensor deployment is reduced, high-risk areas of eutrophication are identified, and monitoring results are stabilized.
Smart Images
Figure CN122017166A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of water quality monitoring technology, specifically to a method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion. Background Technology
[0002] Eutrophication refers to the pollution phenomenon in which excessive amounts of nutrients such as nitrogen and phosphorus enter slow-flowing water bodies such as lakes and estuaries, causing abnormal proliferation of algae and plankton, leading to a decrease in dissolved oxygen and deterioration of water quality. Under natural conditions, this process is slow, but the input of industrial wastewater, domestic sewage, and agricultural non-point source pollution can accelerate this process, forming algal blooms or red tides.
[0003] Because the water monitoring area is large and the degree of eutrophication varies in different locations, it is difficult to obtain accurate monitoring results by monitoring only a single point, and it is also difficult to monitor local high-risk areas. Although multi-point monitoring can improve the monitoring accuracy and cover local high-risk areas, it requires the installation of sensors at every location in the water monitoring area, which requires a lot of money and is not very feasible. In addition, the water flow rate during monitoring can also cause the monitoring results to be unstable. Summary of the Invention
[0004] To address the aforementioned technical problems, this technical solution provides a dynamic monitoring method for water eutrophication based on multi-source sensor data fusion, which solves the problems mentioned in the background section.
[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A dynamic monitoring method for water eutrophication based on multi-source sensor data fusion includes: Generate at least one benchmark index for eutrophication detection, acquire a sensor for measuring the benchmark index, and use the sensor to perform the measurement. The water area to be tested is uniformly divided to obtain at least one test point. One of the test points is selected as the target test point, and the remaining test points are designated as non-target test points. Based on data processing, the first value of the benchmark index at the target test point and the second value of the benchmark index at the non-target test point are obtained. Based on the second value, at least one feature set is formed, and the feature set is composed of test points. Based on the feature set, at least one feature point is formed; Based on the measurement results at the feature points, a compensation function for the benchmark index is formed to identify the water flow at the feature points and obtain the characteristic velocity. Based on the characteristic velocity, the number of samplings and the sampling interval at the feature points are set. At the feature points, the actual values of the benchmark indicators are collected. Using the compensation function, the actual values of the benchmark indicators are combined to obtain the comprehensive value of the benchmark indicators in the water body to be tested. The comprehensive value and the actual values are used to assess the eutrophication of the water.
[0006] Preferably, generating at least one benchmark indicator for eutrophication detection includes the following steps: At least one benchmark indicator is defined as phosphorus content, nitrogen content, chlorophyll a content, transparency, permanganate index, dissolved oxygen, and algal density.
[0007] Preferably, the analysis to obtain the first value of the benchmark index at the target test point and the second value of the benchmark index at non-target test points includes the following steps: Set at least one characteristic time point and at least one data collection day, with the data collection days being consecutive working days. The number of characteristic times point is equal to the number of non-target test points. Randomly establish a one-to-one correspondence between the characteristic times point and the non-target test points. Randomly select a value as the first value of the benchmark index at the target test point; On the same day, the reference index at the target test point is measured at a characteristic time to obtain the reference value. On the same day, the reference index at the non-target test point corresponding to the characteristic time is measured at a characteristic time to obtain the flow value. The reference value and flow value obtained at the same characteristic time are then paired. Divide the first value by the reference value to obtain the adjustment ratio of the reference value. Multiply the adjustment ratio of the reference value by the corresponding flow value to obtain the second value of the benchmark index at the non-target test point. The second set of values obtained within the same collection day is obtained by summarizing the second values.
[0008] Preferably, the step of filtering to form at least one feature set based on the second numerical value includes the following steps: A sample water area is pre-acquired. Based on historical monitoring data of the sample water area, the number of days with pollution in the sample water area in a year is counted as the number of polluted days. The number of polluted days is divided by 365 to obtain the pollution probability. Take the maximum value of the second value in all the second value sets as the baseline value, subtract the contamination probability from 1 to get the retention ratio, and multiply the baseline value by the retention ratio to get the screening threshold value. The non-target test points corresponding to the second values in the second set of values that exceed the screening threshold are taken as reserve points. The reserve points in the second set of values are summarized to obtain the reserve point set. Based on empirical data, the allowable measurement error is obtained, the average of the second values in the second set of values is taken to obtain the preset value, and a subset of the second set of values is used as the set to be verified. The average of the second value in the set to be verified is taken to obtain the value to be verified. If the difference between the value to be verified and the preset value is less than the measurement allowable error, a set of values to be verified will be generated as a preliminary set. The initial set containing the set of reserve points is used as the reserve set, and the reserve set with the smallest number of elements is used as the feature set.
[0009] Preferably, forming at least one feature point based on the feature set includes the following steps: Take the union of at least one feature set to obtain the overall set, and use the test points in the overall set as feature points.
[0010] Preferably, the step of forming the compensation function for the benchmark index based on the measurement results at the feature points includes the following steps: The average value of the second values corresponding to the feature points in the second set of values is taken to obtain the feature values. The feature values are then paired with preset values and fitted to obtain the compensation function of the benchmark index. Here, the feature values are the independent variables and the preset values are the dependent variables.
[0011] Preferably, the identification of water flow at the feature point to obtain the feature velocity includes the following steps: At the current moment, perform contour recognition on the object at the feature point to obtain at least one first object contour. Take the first object contour with the largest contour as the first target contour and obtain the position of the center of the first target contour as the first position. After a preset time, contour recognition is performed on the object at the feature point to obtain at least one second object contour. The second object contour with the largest contour is taken as the second target contour. The position of the center of the second target contour is obtained as the second position. The preset time is set based on experience. The characteristic velocity is obtained by dividing the distance between the second position and the first position by a preset time.
[0012] Preferably, setting the number of samples and the sampling interval based on the feature velocity includes the following steps: Based on historical data, the minimum velocity of water flow at the feature point is obtained as the target value; The water surface area at the feature point is taken as the feature area. At least one identification point is evenly selected at the edge of the feature area to form at least one identification point combination. The identification point combination consists of two identification points. The distance between the identification points in the identification point combination is used as the parameter value of the identification point combination, and the maximum value of the parameter value is used as the feature distance. Divide the feature distance by the target value to obtain the first dwell time. Set the number of samples corresponding to the target value to 2, and set the sampling interval duration corresponding to the target value to the first dwell time. Divide the characteristic velocity by the target value to obtain the amplification factor. If the amplification factor is an integer, set the amplification factor to the amplification factor; otherwise, set the amplification factor to the integer part of the amplification factor plus 1. The number of samplings at the feature point is magnified by 2 times. The first dwell time is divided by the magnification factor to obtain the sampling interval at the feature point.
[0013] Preferably, the step of acquiring the actual value of the benchmark index at the feature point includes the following steps: The value of the benchmark index is collected at least once at the feature point to obtain at least one collected value. The number of collections is equal to the number of samplings, and the collection interval is equal to the sampling interval duration. The average value of the at least one collected value is taken to obtain the actual value of the benchmark index.
[0014] Preferably, the process of synthesizing the actual values of the benchmark indicators to obtain the comprehensive value of the benchmark indicators in the water body to be tested includes the following steps: The average value of the actual values of the benchmark indicators at all feature points is taken as the average value. The average value is then substituted into the compensation function of the benchmark indicators to obtain the comprehensive value of the benchmark indicators in the water body to be tested.
[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: By analyzing the first and second values, forming feature points, and setting the number of samplings and sampling intervals, monitoring points can be selected based on the water conditions. This ensures overall monitoring accuracy while minimizing the number of sensors required, thus controlling monitoring costs. Furthermore, by forming feature points, areas with high eutrophication risk can be identified for separate monitoring, thus balancing overall and local monitoring. Additionally, the monitoring method can be adaptively adjusted based on the water flow velocity at the monitoring location, ensuring the stability of monitoring at that location. Attached Figure Description
[0016] Figure 1 This is a flowchart illustrating the dynamic monitoring method for water eutrophication based on multi-source sensor data fusion according to the present invention. Figure 2 This is a flowchart illustrating the process of obtaining the first value of the benchmark index at the target test point and the second value of the benchmark index at non-target test points in the analysis of this invention. Figure 3 This is a schematic diagram of the process of forming at least one feature set based on a second numerical value according to the present invention; Figure 4 This is a schematic diagram illustrating the process of identifying water flow at feature points and obtaining characteristic velocities according to the present invention. Figure 5This is a schematic diagram illustrating the process of setting the number of samples and the sampling interval at a feature point based on feature velocity, according to the present invention. Detailed Implementation
[0017] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.
[0018] Reference Figure 1 As shown, the dynamic monitoring method for water eutrophication based on multi-source sensor data fusion includes: Generate at least one benchmark index for eutrophication detection, acquire a sensor for measuring the benchmark index, and use the sensor to perform the measurement. The water area to be tested is uniformly divided to obtain at least one test point. One of the test points is selected as the target test point, and the remaining test points are designated as non-target test points. Based on data processing, the first value of the benchmark index at the target test point and the second value of the benchmark index at the non-target test point are obtained. Based on the second value, at least one feature set is formed, and the feature set is composed of test points. Based on the feature set, at least one feature point is formed; Based on the measurement results at the feature points, a compensation function for the benchmark index is formed to identify the water flow at the feature points and obtain the characteristic velocity. Based on the characteristic velocity, the number of samplings and the sampling interval at the feature points are set. At the feature points, the actual values of the benchmark indicators are collected. Using the compensation function, the actual values of the benchmark indicators are combined to obtain the comprehensive value of the benchmark indicators in the water body to be tested. The comprehensive value and the actual values are used to assess the eutrophication of the water.
[0019] When monitoring water bodies, which cover large areas, the eutrophication levels vary significantly across different locations. Monitoring only one location makes it difficult to identify eutrophication in areas outside that location. However, monitoring all locations would require a large number of sensors, which is impractical. Therefore, it is necessary to carefully select the locations of the sensors, limiting their number as much as possible, and ensuring that the monitoring locations are in areas with high eutrophication risk. This way, if eutrophication occurs, it will almost certainly be detected in these locations, allowing for timely intervention. At the same time, it is also necessary to estimate the overall condition of the water body using the monitoring data from the sensors. Based on this, preventative water management can be implemented. Because the water body is large, the management effect will be insufficient without preventative measures. Therefore, a series of steps are subsequently set to limit the location of the sensors. When water is flowing, the faster the flow, the more unstable the monitoring results become, making it difficult to reflect the true situation. Therefore, it is necessary to limit the number of samplings based on the flow speed, and also to limit the sampling interval. This ensures that the sampled data can more accurately reflect the current state of the water body, and appropriate steps are set up to process it in subsequent steps. Here, the composite value is used to assess the overall condition of the water body under test, while the actual value is the value of the benchmark index at the feature point. It can assess the water condition at the feature point individually because all feature points include high-risk areas of eutrophication. Therefore, local eutrophication can be identified in a timely manner.
[0020] Generating at least one benchmark for eutrophication detection includes the following steps: At least one benchmark indicator is defined as phosphorus content, nitrogen content, chlorophyll a content, transparency, permanganate index, dissolved oxygen, and algal density.
[0021] Eutrophication is mainly caused by various nutrients. Therefore, when monitoring eutrophication, these nutrients are monitored, and the degree of eutrophication is judged based on the monitoring results. These judgments can be based on existing eutrophication standards, so as to determine whether the content of each nutrient exceeds the standard. Thus, the existence of eutrophication and the degree of eutrophication can be identified. In this plan, the judgment of eutrophication is not the focus, so it will not be elaborated on.
[0022] Reference Figure 2 As shown, the analysis to obtain the first value of the benchmark index at the target test point and the second value of the benchmark index at non-target test points includes the following steps: Set at least one characteristic time point and at least one data collection day, with the data collection days being consecutive working days. The number of characteristic times point is equal to the number of non-target test points. Randomly establish a one-to-one correspondence between the characteristic times point and the non-target test points. Randomly select a value as the first value of the benchmark index at the target test point; On the same day, the reference index at the target test point is measured at a characteristic time to obtain the reference value. On the same day, the reference index at the non-target test point corresponding to the characteristic time is measured at a characteristic time to obtain the flow value. The reference value and flow value obtained at the same characteristic time are then paired. Divide the first value by the reference value to obtain the adjustment ratio of the reference value. Multiply the adjustment ratio of the reference value by the corresponding flow value to obtain the second value of the benchmark index at the non-target test point. The second set of values obtained within the same collection day is obtained by summarizing the second values.
[0023] Here, to filter monitoring locations, it is first necessary to obtain the overall situation of the water area to be measured. Therefore, many non-target test points are set up. This allows for the selection of the required locations from among the non-target test points. However, it is impossible to set up sensors at all non-target test points due to the sheer number and high cost. Consequently, it is impossible to acquire data from non-target test points at the same time. Since the data are not from the same time, they are not comparable. Therefore, it is necessary to obtain data under the same standard through some means. Here, the target test point is used as a reference. Each time, the target test point and a non-target test point can be measured simultaneously. This allows us to obtain the adjustment ratio of the target test point's value change to the first value. The flow value is then transformed according to the adjustment ratio. When the data of the non-target test point is the second value, the data of the target test point is always the first value. Therefore, the second values of the non-target test points are all under the same standard and are thus comparable, allowing for selection.
[0024] Reference Figure 3 As shown, based on the second numerical value, the process of filtering to form at least one feature set includes the following steps: A sample water area is pre-acquired. Based on historical monitoring data of the sample water area, the number of days with pollution in the sample water area in a year is counted as the number of polluted days. The number of polluted days is divided by 365 to obtain the pollution probability. Take the maximum value of the second value in all the second value sets as the baseline value, subtract the contamination probability from 1 to get the retention ratio, and multiply the baseline value by the retention ratio to get the screening threshold value. The non-target test points corresponding to the second values in the second set of values that exceed the screening threshold are taken as reserve points. The reserve points in the second set of values are summarized to obtain the reserve point set. Based on empirical data, the allowable measurement error is obtained, the average of the second values in the second set of values is taken to obtain the preset value, and a subset of the second set of values is used as the set to be verified. The average of the second value in the set to be verified is taken to obtain the value to be verified. If the difference between the value to be verified and the preset value is less than the measurement allowable error, a set of values to be verified will be generated as a preliminary set. The initial set containing the set of reserve points is used as the reserve set, and the reserve set with the smallest number of elements is used as the feature set.
[0025] The second values obtained within the same collection day are summarized to obtain a second value set. At least one collection day will form at least one second value set. First, an operation is performed on a second value set. Since the second values are all generated by non-target test points, each second value has its corresponding non-target test point. When selecting non-target test points, priority should be given to non-target test points with higher values. However, a corresponding screening threshold needs to be formed for selection. Since the probability of pollution can be obtained, the proportion of values that show eutrophication is similar. The retention ratio is the proportion of values that do not show eutrophication. Therefore, the baseline value is multiplied by the retention ratio to obtain the screening threshold. Values greater than the screening threshold are locations with a high risk of eutrophication. Therefore, these values are summarized to form a reserve point set. The reserve points in the reserve point set can be used to predict high-risk areas. In addition, a preliminary set needs to be formed. The average value of the second numerical values of the points in the preliminary set needs to be close to the overall situation of the water area to be measured, i.e., the preset value. Then, the overall situation can be predicted based on the situation in the preliminary set. However, it should be noted that a sensor needs to be set at each selected location. Therefore, while meeting the requirements, the fewer the number of sensors, the more economical it is. Thus, the preliminary set containing the set of preliminary points is used as the preliminary set, and the preliminary set with the smallest number of elements is used as the feature set. Using the feature points in the feature set as the monitoring location is the most appropriate. However, since the feature set is formed by the second numerical set obtained within a collection day, and the data collected in a single day has limitations, i.e., it may not be comprehensive enough, the union of the feature sets formed by multiple collection days is taken, and the test points in the overall set are used as feature points. Then, the monitoring results using the feature points are more reliable.
[0026] Forming at least one feature point based on the feature set includes the following steps: Take the union of at least one feature set to obtain the overall set, and use the test points in the overall set as feature points.
[0027] Based on the measurement results at the feature points, the compensation function for the benchmark index is formed through the following steps: The average value of the second values corresponding to the feature points in the second set of values is taken to obtain the feature values. The feature values are then paired with preset values and fitted to obtain the compensation function of the benchmark index. Here, the feature values are the independent variables and the preset values are the dependent variables.
[0028] Although the feature values at the feature points are relatively close to the preset values, there are still deviations. Therefore, in order to further improve the measurement accuracy, a compensation function is formed, which allows for subsequent adjustments to the values at the feature points.
[0029] Reference Figure 4As shown, identifying the water flow at a feature point and obtaining the feature velocity involves the following steps: At the current moment, perform contour recognition on the object at the feature point to obtain at least one first object contour. Take the first object contour with the largest contour as the first target contour and obtain the position of the center of the first target contour as the first position. After a preset time, contour recognition is performed on the object at the feature point to obtain at least one second object contour. The second object contour with the largest contour is taken as the second target contour. The position of the center of the second target contour is obtained as the second position. The preset time is set based on experience. The characteristic velocity is obtained by dividing the distance between the second position and the first position by a preset time.
[0030] When performing feature velocity recognition, the main method is to identify objects by their movement in the water. Since the water is flowing and objects rotate with the water flow, object recognition cannot be performed by whether the contours are consistent. It can be known that within the image range, the contour of the largest object is always the largest. Therefore, the first position and the second position can be regarded as the positions of the same object at different times, and thus its velocity can be obtained.
[0031] Reference Figure 5 As shown, setting the number of samples and the sampling interval at a feature point based on the feature velocity includes the following steps: Based on historical data, the minimum velocity of water flow at the feature point is obtained as the target value; The water surface area at the feature point is taken as the feature area. At least one identification point is evenly selected at the edge of the feature area to form at least one identification point combination. The identification point combination consists of two identification points. The distance between the identification points in the identification point combination is used as the parameter value of the identification point combination, and the maximum value of the parameter value is used as the feature distance. Divide the feature distance by the target value to obtain the first dwell time. Set the number of samples corresponding to the target value to 2, and set the sampling interval duration corresponding to the target value to the first dwell time. Divide the characteristic velocity by the target value to obtain the amplification factor. If the amplification factor is an integer, set the amplification factor to the amplification factor; otherwise, set the amplification factor to the integer part of the amplification factor plus 1. The number of samplings at the feature point is magnified by 2 times. The first dwell time is divided by the magnification factor to obtain the sampling interval at the feature point.
[0032] Generally speaking, the target value is when the water flow velocity is the minimum. At this time, only two samplings are needed because the situation at the feature point will hardly fluctuate. The time interval is the time it takes for the water to enter the feature area and leave the feature area. When the characteristic velocity is determined, the number of samplings needs to be set proportionally. However, the number of samplings is an integer, so it needs to be rounded down. As the number of samplings increases, the sampling interval needs to be reduced accordingly, which is also generated according to the magnification factor.
[0033] At the feature points, the actual values of the benchmark indicators are collected through the following steps: The value of the benchmark index is collected at least once at the feature point to obtain at least one collected value. The number of collections is equal to the number of samplings, and the collection interval is equal to the sampling interval duration. The average value of the at least one collected value is taken to obtain the actual value of the benchmark index.
[0034] The process of synthesizing the actual values of the benchmark indicators to obtain the composite value of the benchmark indicators in the water body to be tested includes the following steps: The average value of the actual values of the benchmark indicators at all feature points is taken as the average value. The average value is then substituted into the compensation function of the benchmark indicators to obtain the comprehensive value of the benchmark indicators in the water body to be tested.
[0035] Furthermore, this solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is invoked, it executes the aforementioned dynamic monitoring method for water eutrophication based on multi-source sensor data fusion.
[0036] It is understandable that the storage medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid-state drive (SSD).
[0037] In summary, the advantages of this invention are as follows: by analyzing and obtaining the first and second values, forming feature points, and setting the number of samplings and the sampling interval, the monitoring points can be screened according to the water conditions. This ensures overall monitoring accuracy while minimizing the number of sensors required, thereby controlling monitoring costs. Furthermore, by forming feature points, areas with high eutrophication risk can be identified for separate monitoring, thus balancing overall and local monitoring. Additionally, the monitoring method can be adaptively adjusted according to the water flow velocity at the monitoring location, thereby ensuring the stability of monitoring at the monitoring location.
[0038] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.
Claims
1. A method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion, characterized in that, include: Generate at least one benchmark index for eutrophication detection, acquire a sensor for measuring the benchmark index, and use the sensor to perform the measurement. The water area to be tested is uniformly divided to obtain at least one test point. One of the test points is selected as the target test point, and the remaining test points are designated as non-target test points. Based on data processing, the first value of the benchmark index at the target test point and the second value of the benchmark index at the non-target test point are obtained. Based on the second value, at least one feature set is formed, and the feature set is composed of test points. Based on the feature set, at least one feature point is formed; Based on the measurement results at the feature points, a compensation function for the benchmark index is formed to identify the water flow at the feature points and obtain the characteristic velocity. Based on the characteristic velocity, the number of samplings and the sampling interval at the feature points are set. At the feature points, the actual values of the benchmark indicators are collected. Using the compensation function, the actual values of the benchmark indicators are combined to obtain the comprehensive value of the benchmark indicators in the water body to be tested. The comprehensive value and the actual values are used to assess the eutrophication of the water.
2. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 1, characterized in that, The generation of at least one benchmark indicator for eutrophication detection includes the following steps: At least one benchmark indicator is defined as phosphorus content, nitrogen content, chlorophyll a content, transparency, permanganate index, dissolved oxygen, and algal density.
3. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 2, characterized in that, The analysis to obtain the first value of the benchmark index at the target test point and the second value of the benchmark index at non-target test points includes the following steps: Set at least one characteristic time point and at least one data collection day, with the data collection days being consecutive working days. The number of characteristic times point is equal to the number of non-target test points. Randomly establish a one-to-one correspondence between the characteristic times point and the non-target test points. Randomly select a value as the first value of the benchmark index at the target test point; On the same day, the reference index at the target test point is measured at a characteristic time to obtain the reference value. On the same day, the reference index at the non-target test point corresponding to the characteristic time is measured at a characteristic time to obtain the flow value. The reference value and flow value obtained at the same characteristic time are then paired. Divide the first value by the reference value to obtain the adjustment ratio of the reference value. Multiply the adjustment ratio of the reference value by the corresponding flow value to obtain the second value of the benchmark index at the non-target test point. The second set of values obtained within the same collection day is obtained by summarizing the second values.
4. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 3, characterized in that, The process of filtering to form at least one feature set based on the second numerical value includes the following steps: A sample water area is pre-acquired. Based on historical monitoring data of the sample water area, the number of days with pollution in the sample water area in a year is counted as the number of polluted days. The number of polluted days is divided by 365 to obtain the pollution probability. Take the maximum value of the second value in all the second value sets as the baseline value, subtract the contamination probability from 1 to get the retention ratio, and multiply the baseline value by the retention ratio to get the screening threshold value. The non-target test points corresponding to the second values in the second set of values that exceed the screening threshold are taken as reserve points. The reserve points in the second set of values are summarized to obtain the reserve point set. Based on empirical data, the allowable measurement error is obtained, the average of the second values in the second set of values is taken to obtain the preset value, and a subset of the second set of values is used as the set to be verified. The average of the second value in the set to be verified is taken to obtain the value to be verified. If the difference between the value to be verified and the preset value is less than the measurement allowable error, a set of values to be verified will be generated as a preliminary set. The initial set containing the set of reserve points is used as the reserve set, and the reserve set with the smallest number of elements is used as the feature set.
5. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 4, characterized in that, The process of forming at least one feature point based on the feature set includes the following steps: Take the union of at least one feature set to obtain the overall set, and use the test points in the overall set as feature points.
6. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 5, characterized in that, The process of forming a compensation function for the benchmark index based on the measurement results at the feature points includes the following steps: The average value of the second values corresponding to the feature points in the second set of values is taken to obtain the feature values. The feature values are then paired with preset values and fitted to obtain the compensation function of the benchmark index. Here, the feature values are the independent variables and the preset values are the dependent variables.
7. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 6, characterized in that, The process of identifying the water flow at feature points and obtaining the feature velocity includes the following steps: At the current moment, perform contour recognition on the object at the feature point to obtain at least one first object contour. Take the first object contour with the largest contour as the first target contour and obtain the position of the center of the first target contour as the first position. After a preset time, contour recognition is performed on the object at the feature point to obtain at least one second object contour. The second object contour with the largest contour is taken as the second target contour. The position of the center of the second target contour is obtained as the second position. The preset time is set based on experience. The characteristic velocity is obtained by dividing the distance between the second position and the first position by a preset time.
8. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 7, characterized in that, The step of setting the number of samples and the sampling interval at a feature point based on feature velocity includes the following steps: Based on historical data, the minimum velocity of water flow at the feature point is obtained as the target value; The water surface area at the feature point is taken as the feature area. At least one identification point is evenly selected at the edge of the feature area to form at least one identification point combination. The identification point combination consists of two identification points. The distance between the identification points in the identification point combination is used as the parameter value of the identification point combination, and the maximum value of the parameter value is used as the feature distance. Divide the feature distance by the target value to obtain the first dwell time. Set the number of samples corresponding to the target value to 2, and set the sampling interval duration corresponding to the target value to the first dwell time. Divide the characteristic velocity by the target value to obtain the amplification factor. If the amplification factor is an integer, set the amplification factor to the amplification factor; otherwise, set the amplification factor to the integer part of the amplification factor plus 1. The number of samplings at the feature point is magnified by 2 times. The first dwell time is divided by the magnification factor to obtain the sampling interval at the feature point.
9. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 8, characterized in that, The process of collecting the actual values of the benchmark index at the feature points includes the following steps: The value of the benchmark index is collected at least once at the feature point to obtain at least one collected value. The number of collections is equal to the number of samplings, and the collection interval is equal to the sampling interval duration. The average value of the at least one collected value is taken to obtain the actual value of the benchmark index.
10. The method for dynamic monitoring of water eutrophication based on multi-source sensor data fusion according to claim 9, characterized in that, The process of synthesizing the actual values of the benchmark indicators to obtain the comprehensive value of the benchmark indicators in the water body to be tested includes the following steps: The average value of the actual values of the benchmark indicators at all feature points is taken as the average value. The average value is then substituted into the compensation function of the benchmark indicators to obtain the comprehensive value of the benchmark indicators in the water body to be tested.