Water environment monitoring methods and systems, electronic equipment and storage media

By constructing a sample set and training an image recognition model, the problem of low efficiency in cross-regional anomaly analysis in online water environment monitoring systems was solved, enabling rapid identification and accurate source tracing, and improving the level of intelligence in water environment monitoring and management.

CN123068951APending Publication Date: 2026-11-10CORE VISION (BEIJING) TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611196913.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-07
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing online water environment monitoring systems are unable to quickly detect the transmission patterns of water quality and quantity anomalies across regions and time periods, have low anomaly analysis efficiency, and cannot achieve a closed-loop system for early warning and source tracing.

Method used

A sample set was constructed and an image recognition model was trained. By analyzing the spatiotemporal distribution characteristics and topological relationships of water environment monitoring indicators, anomaly identification and source tracing were achieved. The causes of anomalies were determined by combining the analysis of influencing factors.

Benefits of technology

It enables rapid identification, real-time early warning, and precise source tracing of water environment anomalies, improving the intelligence level of water environment monitoring and management, shortening the response time, and increasing management efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123068951A_ABST
    Figure CN123068951A_ABST
Patent Text Reader

Abstract

This disclosure relates to a water environment monitoring method and system, electronic device, and storage medium. The method includes: acquiring first time-series data of water environment monitoring indicators at different monitoring points within a target water environment monitoring area; constructing a sample set based on the topological relationship between different monitoring points and the first time-series data; training an initial image recognition model based on the sample set to obtain a trained target image recognition model; and determining the recognition result of the heat map to be measured based on a real-time updated heat map within the target water environment monitoring area and the target image recognition model. The recognition result includes anomaly recognition results and source tracing results used to trigger early warning information. Embodiments of this disclosure can achieve rapid identification, real-time early warning, and accurate source tracing of water environment anomalies, improving the intelligence level of water environment monitoring and management.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to a water environment monitoring method and system, electronic equipment and storage medium. Background Technology

[0002] Currently, online water environment monitoring systems are widely used in water environment management, water conservancy projects, and other fields, capable of collecting massive amounts of time-series monitoring data in real time. However, online water environment monitoring systems in related technologies have the following problems in data processing and application:

[0003] First, the monitoring data is mostly presented in the form of time-series curves of a single monitoring point and a single indicator, making it difficult to quickly discover the transmission patterns of water quality and quantity anomalies across regions and time periods.

[0004] Second, anomaly analysis often relies on manual statistical comparison of time series data or the use of simple threshold judgment methods. The processing and analysis of massive amounts of data is difficult and inefficient, and it is impossible to capture the complex correlation features in the data, which easily leads to missed or false alarms.

[0005] Third, most online water environment monitoring systems can only provide early warnings of anomalies, but cannot quickly locate the source of the anomaly at the same time as the warning, resulting in low efficiency of subsequent operation and maintenance work and failing to form a closed-loop system covering the entire process of early warning, diagnosis, source tracing, and disposal. Summary of the Invention

[0006] This disclosure presents a water environment monitoring method and system, electronic equipment, and storage medium.

[0007] According to one aspect of this disclosure, a water environment monitoring method is provided, the method comprising: acquiring first time-series data of water environment monitoring indicators at different monitoring points within a target water environment monitoring area, wherein the water environment monitoring indicators include at least one of water quality indicators, flow rate, and liquid level; constructing a sample set based on the topological relationship between different monitoring points and the first time-series data, the sample set including sample heat maps corresponding to the water environment monitoring indicators, the sample heat maps being used to reflect the spatiotemporal distribution characteristics of the monitoring values ​​of the corresponding water environment monitoring indicators; training an initial image recognition model based on the sample set to obtain a trained target image recognition model; determining the recognition result of the heat map to be measured based on the real-time updated heat map to be measured within the target water environment monitoring area and the target image recognition model, the recognition result including anomaly identification results and source tracing results of the target water environment monitoring area, the anomaly identification results and the source tracing results being used to trigger early warning information for the target water environment monitoring area, the anomaly identification results including a first anomaly type and an anomaly area location, the source tracing results including the anomaly source point and the confidence level of the anomaly source point.

[0008] In one possible implementation, obtaining the first time-series data of water environment monitoring indicators at different monitoring points within the target water environment monitoring area includes: selecting at least one water environment monitoring indicator for evaluating the target water environment monitoring area from a variety of preset initial water environment monitoring indicators; and obtaining the first time-series data of the water environment monitoring indicator at different monitoring points within the target water environment monitoring area for each selected water environment monitoring indicator.

[0009] In one possible implementation, at least one water environment monitoring indicator for evaluating the target water environment monitoring area is selected from a plurality of preset initial water environment monitoring indicators. This includes: determining the environmental disturbance resistance of each initial water environment monitoring indicator based on the water body type of the target water environment monitoring area, wherein the environmental disturbance resistance indicates the ability of the initial water environment monitoring indicator to resist environmental disturbances in the target water environment monitoring area; determining the discrimination of each initial water environment monitoring indicator based on different historical second anomaly types existing in the target water environment monitoring area, wherein the discrimination indicates the strength of specificity exhibited by different initial water environment monitoring indicators in identifying different second anomaly types; determining a score for each initial water environment monitoring indicator based on the environmental disturbance resistance and the discrimination; and selecting at least one water environment monitoring indicator for evaluating the target water environment monitoring area from a plurality of initial water environment monitoring indicators based on the score.

[0010] In one possible implementation, a sample set is constructed based on the topological relationship between different monitoring points and the first time-series data, including: determining the upstream and downstream positional relationship between different monitoring points based on the hydraulic flow direction information and spatial connectivity of different monitoring points within the target water environment monitoring area; using different acquisition times arranged chronologically in the first time-series data as the vertical axis of the sample heat map, and using different monitoring points arranged according to the upstream and downstream positional relationship within the target water environment monitoring area as the horizontal axis of the sample heat map; for any monitoring point and any acquisition time, determining the pixel value of the sample heat map in the sample set at the coordinate position defined by the horizontal axis corresponding to the monitoring point and the vertical axis corresponding to the acquisition time, based on the monitoring value of the water environment monitoring index at that acquisition time.

[0011] In one possible implementation, the initial image recognition model includes at least a first recognition sub-model and a second recognition sub-model. The first recognition sub-model is used to capture the abnormal transmission characteristics of the monitoring values ​​of the water environment monitoring indicators in the time dimension from the sample heat map. The abnormal transmission characteristics include the time of abnormal occurrence and the abnormal transmission speed. The second recognition sub-model is used to capture the spatial location characteristics between different monitoring points from the sample heat map. The spatial location characteristics include the upstream and downstream positional relationships between different monitoring points.

[0012] In one possible implementation, the method further includes: if the recognition result meets preset conditions, determining new samples based on the recognition result; and training the target image recognition model based on the new samples to obtain an optimized target image recognition model.

[0013] In one possible implementation, the method further includes: acquiring influencing factors of anomalies in the target water environment monitoring area, the influencing factors including at least one of rainfall events and hydraulic facility operation events; and performing correlation analysis between the source tracing results and the influencing factors to determine the cause information of the anomalies.

[0014] According to one aspect of this disclosure, a water environment monitoring system is provided, the system comprising: an acquisition module, configured to acquire first time-series data of water environment monitoring indicators at different monitoring points within a target water environment monitoring area, wherein the water environment monitoring indicators include at least one of water quality indicators, flow rate, and liquid level; a construction module, configured to construct a sample set based on the topological relationship between different monitoring points and the first time-series data, the sample set including sample heat maps corresponding to the water environment monitoring indicators, the sample heat maps being used to reflect the spatiotemporal distribution characteristics of the monitoring values ​​of the corresponding water environment monitoring indicators; and a training module, configured to... The sample set is used to train an initial image recognition model to obtain a trained target image recognition model. The recognition module is used to determine the recognition result of the heat map to be measured based on the real-time updated heat map of the target water environment monitoring area and the target image recognition model. The recognition result includes the anomaly recognition result and the source tracing result of the target water environment monitoring area. The anomaly recognition result and the source tracing result are used to trigger the early warning information of the target water environment monitoring area. The anomaly recognition result includes a first anomaly type and the location of the anomaly area. The source tracing result includes the location of the anomaly source point and the confidence level of the anomaly source point.

[0015] In one possible implementation, the acquisition module is used to: select at least one water environment monitoring indicator for evaluating the target water environment monitoring area from a variety of preset initial water environment monitoring indicators; and for each selected water environment monitoring indicator, acquire the first time-series data of the water environment monitoring indicator at different monitoring points within the target water environment monitoring area.

[0016] In one possible implementation, at least one water environment monitoring indicator for evaluating the target water environment monitoring area is selected from a plurality of preset initial water environment monitoring indicators. This includes: determining the environmental disturbance resistance of each initial water environment monitoring indicator based on the water body type of the target water environment monitoring area, wherein the environmental disturbance resistance indicates the ability of the initial water environment monitoring indicator to resist environmental disturbances in the target water environment monitoring area; determining the discrimination of each initial water environment monitoring indicator based on different historical second anomaly types existing in the target water environment monitoring area, wherein the discrimination indicates the strength of specificity exhibited by different initial water environment monitoring indicators in identifying different second anomaly types; determining a score for each initial water environment monitoring indicator based on the environmental disturbance resistance and the discrimination; and selecting at least one water environment monitoring indicator for evaluating the target water environment monitoring area from a plurality of initial water environment monitoring indicators based on the score.

[0017] In one possible implementation, the construction module is used to: determine the upstream and downstream positional relationship between different monitoring points based on the hydraulic flow direction information and spatial connection relationship of different monitoring points within the target water environment monitoring area; use different collection times arranged chronologically in the first time series data as the vertical axis of the sample heat map, and use different monitoring points arranged according to the upstream and downstream positional relationship within the target water environment monitoring area as the horizontal axis of the sample heat map; for any monitoring point and any collection time, determine the pixel value of the sample heat map in the sample set at the coordinate position defined by the horizontal axis corresponding to the monitoring point and the vertical axis corresponding to the collection time, based on the monitoring value of the water environment monitoring index of the monitoring point at the collection time.

[0018] In one possible implementation, the initial image recognition model includes at least a first recognition sub-model and a second recognition sub-model. The first recognition sub-model is used to capture the abnormal transmission characteristics of the monitoring values ​​of the water environment monitoring indicators in the time dimension from the sample heat map. The abnormal transmission characteristics include the time of abnormal occurrence and the abnormal transmission speed. The second recognition sub-model is used to capture the spatial location characteristics between different monitoring points from the sample heat map. The spatial location characteristics include the upstream and downstream positional relationships between different monitoring points.

[0019] In one possible implementation, the training module is further configured to: determine new samples based on the recognition results when the recognition results meet preset conditions; and train the target image recognition model based on the new samples to obtain an optimized target image recognition model.

[0020] In one possible implementation, the identification module is further configured to: acquire influencing factors of anomalies in the target water environment monitoring area, the influencing factors including at least one of rainfall events and hydraulic facility operation events; and perform correlation analysis between the source tracing results and the influencing factors to determine the cause information of the anomalies.

[0021] According to one aspect of this disclosure, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to invoke the instructions stored in the memory to perform the method described above.

[0022] According to one aspect of this disclosure, a computer-readable storage medium is provided that stores computer program instructions thereon, which, when executed by a processor, implement the above-described method.

[0023] The water environment monitoring method of this disclosure can acquire first time-series data of water environment monitoring indicators at different monitoring points within a target water environment monitoring area, wherein the water environment monitoring indicators include at least one of water quality indicators, flow rate, and liquid level; construct a sample set based on the topological relationship between different monitoring points and the first time-series data, the sample set including sample heat maps corresponding to the water environment monitoring indicators, the sample heat maps being used to reflect the spatiotemporal distribution characteristics of the monitoring values ​​of the corresponding water environment monitoring indicators; train an initial image recognition model based on the sample set to obtain a trained target image recognition model; determine the recognition result of the heat map to be measured based on the real-time updated heat map to be measured within the target water environment monitoring area and the target image recognition model, the recognition result including anomaly identification results and source tracing results of the target water environment monitoring area, the anomaly identification results and the source tracing results being used to trigger early warning information for the target water environment monitoring area, the anomaly identification results including a first anomaly type and an anomaly area location, the source tracing results including the anomaly source point and the confidence level of the anomaly source point.

[0024] In this way, the time-series data of water environment monitoring indicators from isolated monitoring points can be transformed into sample heat maps that can reflect cross-regional correlations. This is conducive to intuitively presenting the spatiotemporal distribution characteristics and anomaly transmission trends of water environment monitoring indicators, making it easier for target image recognition models to better learn the spatiotemporal distribution characteristics of water environment monitoring indicators. This is beneficial for achieving rapid identification, real-time early warning, and accurate source tracing of water environment anomalies, and improves the level of intelligence in water environment monitoring and governance.

[0025] Furthermore, the trained target image recognition model can output anomaly identification results and source tracing results. By combining anomaly identification with upstream and downstream point tracing, it can not only trigger early warnings and alerts in real time based on anomaly identification results, but also quickly locate the source of the anomaly and its transmission path, realizing the integration of early warning and source tracing. This provides precise guidance for water environment emergency response, shortens response time, and improves governance efficiency.

[0026] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Other features and aspects of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0027] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the specification, serve to illustrate the technical solutions of this disclosure.

[0028] Figure 1 A flowchart illustrating the water environment monitoring method provided in this embodiment is shown.

[0029] Figure 2 This is a rendering of a sample heatmap provided in an embodiment of the present disclosure.

[0030] Figure 3 A block diagram of a water environment monitoring system provided in an embodiment of this disclosure is shown.

[0031] Figure 4 This diagram illustrates a block diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0032] Various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0033] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0034] In this document, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0035] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0036] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, data stored, data displayed, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant regions.

[0037] Figure 1 A flowchart illustrating a water environment monitoring method provided in an embodiment of this disclosure is shown. In one possible implementation, the entity executing the water environment monitoring method may be a water environment monitoring device. For example, the water environment monitoring method may be executed by a terminal device, a server, or other electronic equipment. The terminal device may be a user equipment (UE), mobile device, user terminal, terminal, cellular phone, cordless phone, personal digital assistant (PDA), handheld device, computing device, vehicle-mounted device, or wearable device, etc. In some possible implementations, the water environment monitoring method may be implemented by a processor calling computer-readable instructions stored in memory.

[0038] like Figure 1 As shown, the water environment monitoring method includes:

[0039] In step S11, the first time-series data of water environment monitoring indicators at different monitoring points within the target water environment monitoring area are obtained, wherein the water environment monitoring indicators include at least one of water quality indicators, flow rate and liquid level;

[0040] In step S12, a sample set is constructed based on the topological relationship between different monitoring points and the first time-series data. The sample set includes sample heat maps corresponding to the water environment monitoring indicators. The sample heat maps are used to reflect the spatiotemporal distribution characteristics of the monitoring values ​​of the corresponding water environment monitoring indicators. The spatiotemporal distribution characteristics may include the color characteristics of the abnormal area (e.g., the color intensity mapped by the indicator value of the water environment monitoring indicator), spatial location characteristics (e.g., upstream and downstream monitoring points of the abnormal transmission), and temporal evolution characteristics (e.g., the time of the abnormality, the speed of the abnormal transmission, etc.).

[0041] In step S13, an initial image recognition model is trained based on the sample set to obtain a trained target image recognition model;

[0042] In step S14, based on the real-time updated heat map to be measured within the target water environment monitoring area and the target image recognition model, the recognition result of the heat map to be measured is determined. The recognition result includes the anomaly recognition result and the source tracing result of the target water environment monitoring area. The anomaly recognition result and the source tracing result are used to trigger the early warning information of the target water environment monitoring area. The anomaly recognition result includes a first anomaly type and the location of the anomaly area. The source tracing result includes the location of the anomaly source point and the confidence level of the anomaly source point.

[0043] The water environment monitoring method provided in this disclosure can transform the time-series data of water environment monitoring indicators from isolated monitoring points into sample heat maps that can reflect cross-regional correlations. This is beneficial for intuitively presenting the spatiotemporal distribution characteristics and anomaly transmission trends of water environment monitoring indicators, facilitating the target image recognition model to better learn the spatiotemporal distribution characteristics of water environment monitoring indicators, and enabling rapid identification, real-time early warning, and accurate source tracing of water environment anomalies, thereby improving the level of intelligence in water environment monitoring and management.

[0044] Furthermore, the trained target image recognition model can output anomaly identification results and source tracing results. By combining anomaly identification with upstream and downstream point tracing, it can not only trigger early warnings and alerts in real time based on anomaly identification results, but also quickly locate the source of the anomaly and its transmission path, realizing the integration of early warning and source tracing. This provides precise guidance for water environment emergency response, shortens response time, and improves governance efficiency.

[0045] In one possible implementation, the target water environment monitoring area may include areas such as rivers, lakes, seas, reservoirs, groundwater, and drainage pipe networks. This disclosure does not limit the specific location of the target water environment monitoring area.

[0046] In one possible implementation, the water environment monitoring indicators may include at least one of water quality indicators, flow rate, and liquid level. The water quality indicators may include at least one of chemical oxygen demand (COD), oxidation-reduction potential (ORP), hardness, turbidity, pH, total phosphorus content, ammonia nitrogen content, permanganate index, conductivity, total suspended solids, biological oxygen demand, total organic carbon, sulfate content, chloride content, dissolved oxygen content, dissolved iron content, dissolved manganese content, dissolved copper content, dissolved zinc content, nitrate content, nitrite content, total nitrogen content, fluoride content, selenium content, total arsenic content, total mercury content, total cadmium content, chromium content, total lead content, total cyanide, volatile phenol content, coliform bacteria content, and sulfide content. The embodiments of this disclosure do not specifically limit the categories of water quality indicators.

[0047] It should be understood that in the same target water environment monitoring area, different water environment monitoring indicators may have their own corresponding sample heat maps, and the embodiments disclosed herein do not impose specific limitations on this.

[0048] In one possible implementation, in step S11, initial time-series data of water environment monitoring indicators can be collected using monitoring equipment deployed at different monitoring points within the target water environment monitoring area. This data serves as the first time-series data for the water environment monitoring indicator. The collection frequency can be set according to actual needs, and the embodiments of this disclosure do not impose specific limitations on it. For example, to ensure the continuity and reliability of the first time-series data, the collection frequency can be set to a frequency of not less than 4 times per hour.

[0049] The monitoring equipment may include a quantum dot spectrometer, which can acquire first-time-series data of water quality indicators. The quantum dot spectrometer may include a quantum dot spectral probe, which measures incident light (e.g., light transmitted or scattered after passing through a predetermined area of ​​water sample) based on the physical and optical properties of nanocrystals to obtain spectral information of the incident light. For example, the quantum dot spectral probe may include a nanocrystal chip made of various nanocrystals, comprising a specific arrangement of various nanocrystals (e.g., a nanocrystal array). Each nanocrystal has different light absorption or emission characteristics. Different types of semiconductor nanocrystals can be made of different materials and sizes, allowing the nanocrystal chip to modulate its response to wavelengths over a wide wavelength range to obtain a spectrum adjusted for incident light over a wider wavelength range.

[0050] Light transmitted or scattered through water can be affected by substances in the water (e.g., suspended solids, pollutants), thus acquiring specific spectral information. A quantum dot spectrometer can acquire this spectral information in real time, which represents the water quality information at the monitoring point. Based on this spectral information, the monitoring values ​​of water quality indicators can be determined. For example, by observing the absorption intensity of different wavelengths of light by a water sample, spectral information of different frequency bands of light can be obtained, and the monitoring values ​​of water quality indicators can be calculated from this spectral information. This disclosure does not limit the working principle of the quantum dot spectrometer.

[0051] In this example, a quantum dot spectrometer can determine water quality indicators by analyzing the light absorption characteristics of various substances in the water. For instance, it can analyze the light intensity of light at a specific wavelength using spectral information to obtain the concentration of substances corresponding to that specific wavelength range (monitored values ​​of water quality indicators). Alternatively, the quantum dot spectrometer can infer water quality indicators using a neural network. For example, spectral information can be input into the neural network, which can then infer the concentration of various substances (water quality indicators). This disclosure does not limit the method used to determine the monitored values ​​of water quality indicators.

[0052] In one possible implementation, a quantum dot spectrometer can be used to measure water quality indicators (e.g., COD) in real time. The monitoring values ​​of the water quality indicator obtained at multiple times at the same monitoring point can be compiled into a time series data set of the water quality indicator, which can be used as the first time series data. Compared to the method of sampling water and then performing laboratory tests to measure water quality indicators, measuring indicators using a quantum dot spectrometer allows for online, in-situ, high-frequency, and real-time measurement. For example, water quality indicator time series data can be collected at a sampling frequency of 4 times per hour as the first time series data set.

[0053] In one possible implementation, a water level gauge can be used to measure the liquid level at different monitoring points within the target water environment monitoring area in real time. The liquid level monitoring values ​​obtained at multiple times at the same monitoring point can form the first time series data corresponding to the liquid level index.

[0054] In one possible implementation, a flow sensor can be used to measure the flow rate at different monitoring points within the target water environment monitoring area in real time. The flow rate monitoring values ​​obtained at multiple times at the same monitoring point can be used to form the first time series data corresponding to the flow rate index.

[0055] In one possible implementation, initial time-series data of water environment monitoring indicators can be collected by monitoring equipment deployed at different monitoring points within the target water environment monitoring area, and the initial time-series data can be processed to obtain the processed first time-series data.

[0056] For example, data collected during the period of abnormality of the monitoring device can be removed from the initial time-series data to obtain second time-series data; data collected during the period when the monitoring device is out of water can be removed from the second time-series data to obtain third time-series data; over-range processing can be performed on the third time-series data to obtain fourth time-series data; noise reduction processing can be performed on the fourth time-series data to obtain fifth time-series data; time resampling can be performed on the fifth time-series data to obtain sixth time-series data; and missing data can be filled in the sixth time-series data to obtain the first time-series data.

[0057] For example, during a period of monitoring equipment malfunction, the initial time-series data collected by the monitoring equipment is also abnormal. Therefore, data collected during the period of monitoring equipment malfunction can be removed from the initial time-series data so that the data collected during the period of monitoring equipment malfunction is not included in the statistics. The circumstances under which the monitoring equipment is abnormal may include: hardware failure of the monitoring equipment, deviation of the measured value of the monitoring equipment from the true value due to environmental abnormalities (such as excessively high or low ambient temperature), etc. The embodiments of this disclosure do not impose specific limitations on the circumstances under which the monitoring equipment is abnormal.

[0058] After removing the data collected during the period of monitoring equipment malfunction from the initial time series data, the second time series data can be obtained. The second time series data may contain data collected by the monitoring equipment during the period when the equipment was out of the water. In order to improve the accuracy of data collection, the data collected during the period when the monitoring equipment was out of the water can be removed from the second time series data to obtain the third time series data.

[0059] Among them, the data collected during the period when the monitoring equipment is out of water can be removed using conductivity as the criterion. For example, the conductivity value when the monitoring equipment is fully immersed in water can be used as the water threshold. If the conductivity of the monitoring equipment is lower than the water threshold, it is considered that the monitoring equipment is in an out-of-water state, and the data collected during the period when the monitoring equipment is out of water can be removed and not included in the statistics. If the conductivity of the monitoring equipment is higher than or equal to the water threshold, it is considered that the monitoring equipment is in a water state, and the data collected by the monitoring equipment can be retained.

[0060] Once the third time-series data is obtained, it can be processed beyond its range to obtain the fourth time-series data. If the third time-series data is higher than the upper limit of the monitoring device's range, the upper limit of the monitoring device can be used as the fourth time-series data. If the third time-series data is lower than the detection limit (i.e., the lower limit of the range) of the monitoring device, the detection limit of the monitoring device can be used as the fourth time-series data.

[0061] Once the fourth time-series data is obtained, it can be denoised to obtain the fifth time-series data, thereby reducing the interference of single-point noise on anomaly identification and improving the accuracy of anomaly identification. The denoising methods may include Savitzky-Golay (SG) filtering, Gaussian filtering, mean filtering, median filtering, etc. The embodiments of this disclosure do not limit the specific methods used for denoising.

[0062] Once the fifth time-series data is obtained, it can be resampled to obtain the sixth time-series data. The time resampling process unifies the fifth time-series data collected by different monitoring devices to the same sampling frequency and merges the fifth time-series data from different dimensions according to timestamps to obtain the sixth time-series data. It should be understood that the embodiments of this disclosure do not limit the specific value of the sampling frequency and can be set according to the actual application scenario.

[0063] For example, for monitoring devices 1 and 2 set at different monitoring points, with monitoring device 1 sampling at a frequency of M times / hour and monitoring device 2 sampling at a frequency of N times / hour, the fifth time series data collected by different monitoring devices can be unified to a 1-hour frequency. For any hour t, the average of the M fifth time series data obtained by monitoring device 1 in hour t can be used as the sixth time series data after time resampling, and the average of the N fifth time series data obtained by monitoring device 2 in hour t can be used as the sixth time series data after time resampling.

[0064] Once the sixth time-series data is obtained, missing data can be filled in to obtain the first time-series data. The missing data filling method can include forward filling and backward filling. For example, for a gap caused by a data transmission interruption, forward filling can use the nearest non-missing value preceding the missing value to fill the current gap. For example, for a sixth time-series data set: [1, missing, missing, 6, missing], the first time-series data after forward filling is: [1, 1,1, 6, 6]. Similarly, backward filling can use the nearest non-missing value following the missing value to fill the current gap. For example, for a sixth time-series data set: [1, missing, missing, 6, missing,2], the first time-series data after forward filling is: [1, 6, 6, 6, 2,2].

[0065] By processing the initial time-series data, higher quality and more accurate first-time-series data can be obtained, thereby improving the data quality of the sample set. This, in turn, improves the effect of subsequent training of the initial image recognition model, resulting in a more robust target image recognition model, and ultimately improving the accuracy and reliability of anomaly identification and tracing results.

[0066] In one possible implementation, due to the significant differences in the physicochemical properties of different water environment monitoring indicators, their background concentrations, variation ranges, and differences from potential external water that may intrude into different target water environment monitoring areas (such as rivers, underground waterways, and water supply networks) vary considerably. This difference directly determines that a uniform water environment monitoring indicator cannot be used to monitor various different water environment monitoring areas. To improve the accuracy and applicability of water environment monitoring, water environment monitoring indicators suitable for each target water environment monitoring area can be selected.

[0067] For example, in step S11, at least one water environment monitoring indicator for evaluating the target water environment monitoring area can be selected from a variety of preset initial water environment monitoring indicators; for each selected water environment monitoring indicator, the first time series data of the water environment monitoring indicator at different monitoring points in the target water environment monitoring area are obtained respectively.

[0068] Optionally, based on subjective experience, at least one water environment monitoring indicator can be directly selected from a variety of preset initial water environment monitoring indicators to evaluate the target water environment monitoring area.

[0069] Optionally, based on the water body type of the target water environment monitoring area (e.g., river water, groundwater, tap water, rainwater, etc.), at least one water environment monitoring indicator can be selected from a variety of preset initial water environment monitoring indicators to evaluate the target water environment monitoring area.

[0070] Optionally, a neural network model for screening water environment monitoring indicators can be pre-trained. Using the trained neural network model, at least one water environment monitoring indicator for evaluating the target water environment monitoring area can be selected from a variety of preset initial water environment monitoring indicators.

[0071] Optionally, based on the industrial pollution discharge characteristics of the target water environment monitoring area, at least one initial water environment monitoring indicator that can represent the pollution discharge characteristics of the target water environment monitoring area can be selected from a variety of preset initial water environment monitoring indicators and used as the water environment monitoring indicator for evaluating the target water environment monitoring area.

[0072] Optionally, the environmental disturbance resistance of each initial water environment monitoring indicator can be determined based on the water body type of the target water environment monitoring area. The environmental disturbance resistance indicates the ability of the initial water environment monitoring indicator to resist environmental interference in the target water environment monitoring area. The discrimination of each initial water environment monitoring indicator can be determined based on different historical second anomaly types existing in the target water environment monitoring area. The discrimination indicates the strength of specificity exhibited by different initial water environment monitoring indicators when identifying different second anomaly types. A score for each initial water environment monitoring indicator is determined based on its environmental disturbance resistance and discrimination. Based on the score, at least one water environment monitoring indicator for evaluating the target water environment monitoring area is selected from multiple initial water environment monitoring indicators.

[0073] For example, based on the water body type of the target water environment monitoring area, initial time-series data corresponding to various initial water environment monitoring indicators for that water body type can be obtained. Specifically, the initial time-series data of the initial water environment monitoring indicators corresponding to that water body type can be obtained by monitoring equipment deployed at different monitoring points within the target water environment monitoring area; or, the initial time-series data of the initial water environment monitoring indicators corresponding to that water body type can be read from a preset indicator database.

[0074] Then, for each initial water environment monitoring indicator's initial time-series data, the reciprocal of the coefficient of variation (the ratio of the standard deviation to the mean of the time-series data) of the initial time-series data of that initial water environment monitoring indicator can be used as the environmental disturbance resistance of that initial water environment monitoring indicator. Alternatively, when the initial water environment monitoring indicator is a water quality indicator, for each water quality indicator, the environmental disturbance resistance of that water quality indicator can be determined based on the correlation results between the initial time-series data of that water quality indicator at different lag periods and the initial time-series data of the flow rate without lag. The lag period describes the duration by which the change in the first time-series data of the water quality indicator caused by environmental disturbance in the target water environment monitoring area occurs later than the change in the first time-series data of the flow rate indicator caused by environmental disturbance in the target water environment monitoring area. The embodiments of this disclosure do not limit the method for determining environmental disturbance resistance.

[0075] Among them, the resistance to environmental disturbance can be used to indicate the ability of initial water environment monitoring indicators to resist environmental disturbances in the target water environment monitoring area. For example, the larger the value of the resistance to environmental disturbance, the less susceptible the initial water environment monitoring indicators are to severe disturbances from instantaneous hydrological and meteorological conditions; the smaller the value of the resistance to environmental disturbance, the more susceptible the initial water environment monitoring indicators are to severe disturbances from instantaneous hydrological and meteorological conditions.

[0076] Determining the resistance to environmental disturbance for each initial water environment monitoring indicator helps to select water environment monitoring indicators with strong resistance to environmental interference from a variety of initial water environment monitoring indicators, thereby avoiding or reducing data distortion caused by environmental fluctuations.

[0077] In parallel, the discriminative power of each initial water environment monitoring indicator can be determined based on the historical second anomaly type existing in the target water environment monitoring area. For example, for each initial water environment monitoring indicator, the probability distribution of the historical time series data of that initial water environment monitoring indicator when the historical second anomaly type occurred, and the probability distribution of the initial time series data of that initial water environment monitoring indicator in the current target water environment monitoring area are obtained, and the similarity between these two probability distributions is determined as the discriminative power of that initial water environment monitoring indicator.

[0078] The discriminant value can be used to indicate the strength of specificity exhibited by different initial water environment monitoring indicators in identifying the second anomaly type. For example, for the same historical second anomaly type, if the discriminant value of a certain initial water environment monitoring indicator is larger, it indicates that the initial water environment monitoring indicator is more likely to identify the historical second anomaly type. If the discriminant value of a certain initial water environment monitoring indicator is smaller, it indicates that the initial water environment monitoring indicator is less likely to identify the historical second anomaly type. Here, the second anomaly type refers to the type of abnormal event, such as combined sewer overflow, external water intrusion, backflow, etc.

[0079] Next, the score of each initial water environment monitoring indicator can be determined based on the weighted average of the resistance to environmental disturbance and the discrimination of each initial water environment monitoring indicator; and the initial water environment monitoring indicators with scores greater than a preset threshold are selected as the water environment monitoring indicators for evaluating the target water environment monitoring area.

[0080] This approach allows for the introduction of a multi-dimensional quantitative evaluation mechanism, taking into account the specificity and resistance to environmental interference exhibited by water environment monitoring indicators when identifying different types of historical anomalies. It overcomes the subjectivity in the selection of water environment monitoring indicators and objectively and scientifically screens the optimal water environment monitoring indicators or combinations, which is conducive to improving the efficiency and accuracy of water environment monitoring.

[0081] For each selected water environment monitoring indicator, the first time series data of that water environment monitoring indicator at different monitoring points within the target water environment monitoring area can be obtained.

[0082] In step S11, the first time-series data of water environment monitoring indicators at different monitoring points within the target water environment monitoring area are obtained. In step S12, a sample set is constructed based on the topological relationship between different monitoring points and the first time-series data.

[0083] The topological relationships include hydraulic flow direction information and spatial connectivity.

[0084] In one possible implementation, step S12 may include: determining the upstream and downstream positional relationship between different monitoring points based on the hydraulic flow direction information and spatial connection relationship of different monitoring points within the target water environment monitoring area; using the acquisition time arranged chronologically in the first time series data as the vertical axis of the sample heat map, and using the different monitoring points within the target water environment monitoring area arranged according to the upstream and downstream positional relationship as the horizontal axis of the sample heat map; for any monitoring point and any acquisition time, determining the pixel value of the sample heat map in the sample set at the coordinate position defined by the horizontal axis corresponding to the monitoring point and the vertical axis corresponding to the acquisition time, based on the monitoring value of the water environment monitoring index of the monitoring point at the acquisition time.

[0085] For example, based on the processed first time-series data and the upstream and downstream spatial relationships of different monitoring points, a two-dimensional heatmap can be constructed with time as the vertical axis and upstream and downstream monitoring points as the horizontal axis. Then, the monitoring values ​​of the first time-series data can be mapped to color gradients to obtain a sample heatmap.

[0086] In this way, massive amounts of first-time series data from multiple monitoring points over a long period of time can be transformed into sample heat maps with upstream and downstream location relationships, thereby transforming curve processing into big data processing for image recognition, which facilitates the processing and display of water environment business data.

[0087] In one possible implementation, to make the trained target image recognition model more accurate, the sample heat map can be updated regularly based on the first time-series data collected in real time and actual needs (the fastest update frequency is synchronized with the data collection frequency). This ensures that the sample heat map can reflect the dynamic changes of water environment monitoring indicators in real time, providing the latest visual data support for the training of subsequent image recognition models.

[0088] Figure 2 This diagram illustrates the effect of a sample heatmap provided in an embodiment of this disclosure. Figure 2 As shown, the vertical axis of the sample heatmap represents time, and the horizontal axis represents the upstream and downstream monitoring points arranged according to their upstream and downstream location within the target water environment monitoring area. The pixel values ​​of the sample heatmap (different pixel values ​​in...) Figure 2 (The colors can be different) represent the monitoring values ​​of water environment monitoring indicators.

[0089] Among them, the sample heat map can be constructed based on the first time series data and the upstream and downstream spatial location relationship of different monitoring points, so as to integrate the time dimension, spatial dimension and data characteristics (monitoring values) of water environment monitoring indicators, thereby better reflecting the spatiotemporal distribution characteristics of water environment monitoring indicators.

[0090] In the vertical axis of the sample heatmap, the data can be arranged in chronological order of the collection times of the first time series data. The unit of collection time can be hours, days, weeks, months, etc. The time granularity between two adjacent collection times can be flexibly adjusted according to the actual application scenario. The embodiments of this disclosure do not impose any restrictions on this. Similarly, in the horizontal axis of the sample heatmap, the monitoring points can be arranged in the order of upstream and downstream hierarchy (for example, if the target water environment monitoring area is a drainage pipe network, then from the upstream source monitoring point to the downstream outlet monitoring point) to clarify the spatial relationship between each monitoring point.

[0091] In one possible implementation, to supervise the training process of the subsequent image recognition model, label information can be added to the sample heatmap. This label information may include at least one of the following: abnormal region location, first anomaly type, and coordinates of the anomaly source point. The abnormal region location may correspond to spatiotemporal coordinates in the sample heatmap, i.e., a specific time period or a specific upstream / downstream monitoring point. The first anomaly type may include water quality anomalies or water quantity anomalies (e.g., abnormal flow rate, abnormal liquid level, etc.), and the degree of anomaly corresponding to the water quality or water quantity anomaly (e.g., mild anomaly, severe anomaly). It may also include anomaly type events, such as combined sewer overflows, backflow, siltation, inflow and infiltration, etc. The anomaly source point is used to indicate the source point of the anomaly type when multiple abnormal region locations belong to the same type of anomaly.

[0092] In one possible implementation, threshold ranges can be preset for water environment monitoring indicators to indicate the degree of anomaly. If the first time series data falls within the normal lower limit to normal upper limit range, this type of first time series data can be considered normal data, corresponding to blue pixel color values; if the first time series data falls within the range of slightly abnormal lower limit to normal lower limit, or normal upper limit to slightly abnormal upper limit, this type of first time series data can be considered slightly abnormal data, corresponding to yellow pixel color values; if the first time series data falls within the range of severely abnormal lower limit to slightly abnormal lower limit, or slightly abnormal upper limit to severely abnormal upper limit, this type of first time series data can be considered severely abnormal data, corresponding to red pixel color values. Figure 2 As shown, the left side shows the effect of fewer abnormal data, and the right side shows the effect of more abnormal data.

[0093] For example, observed values ​​of water environment monitoring indicators can be mapped to pixel color values ​​in a sample heatmap, and a color gradient can be used to represent the degree of anomaly of the water environment monitoring indicators. The intensity of the color is positively correlated with the degree of data anomaly; the greater the deviation of the data from the normal range, the darker the color. This allows the sample heatmap to more intuitively present the spatiotemporal distribution characteristics and anomaly transmission trends of water environment monitoring indicators.

[0094] For example, suppose x represents a water environment monitoring indicator, L low L represents the lower limit of normal. high This represents the normal upper limit if x∈[L] low , L high If the water environment monitoring index x is lower than the normal lower limit L, the deviation can be set to 0; low (i.e., x < L) low ), which can be set to deviation D=L low -x; If the water environment monitoring index x is higher than the normal lower limit L high (i.e., x > L) high ), which can be set as deviation D = xL high . That is:

[0095] (1)

[0096] in, `clamp(a, b, c)` represents the preset maximum deviation, and `clamp(a, b, c)` is the clamping function used to restrict `a` to the interval [b, c]. If `a < b`, then `n = b`; if `a > c`, then `n = c`; if `b ≤ a ≤ c`, then `n = a`. Therefore, `n = clamp(a, b, c)` means: make `n` equal to the value of `a` after it is restricted to between `b` and `c`. For example: `clamp(5, 1, 10) = 5` (not out of range); `clamp(0, 1, 10) = 1` (below the lower limit, take 1); `clamp(15, 1, 10) = 10` (above the upper limit, take 10).

[0097] Then, we can let n = max(0, min(1, n)). If n = 0, the water environment monitoring index is in the normal range and has the lightest color. The larger n is, the more the water environment monitoring index deviates from the normal range and the darker its corresponding color is. When n = 1 or above, the maximum deviation can be reached and the darkest color is fixed.

[0098] In step S12, a sample set is constructed. In step S13, an initial image recognition model is trained based on the sample set to obtain a trained target image recognition model.

[0099] In one possible implementation, the initial image recognition model includes at least a first recognition sub-model and a second recognition sub-model. The first recognition sub-model is used to capture the abnormal transmission characteristics of the monitoring values ​​of the water environment monitoring indicators in the time dimension from the sample heat map. The abnormal transmission characteristics include the time of abnormal occurrence and the abnormal transmission speed. The second recognition sub-model is used to capture the spatial location characteristics between different monitoring points from the sample heat map. The spatial location characteristics include the upstream and downstream positional relationships between different monitoring points.

[0100] The input to the initial image recognition model can be a sample heatmap, and the output of the initial image recognition model can include anomaly identification output and source tracing output. The anomaly identification output can be used to indicate whether there is an anomaly in the sample heatmap, as well as the time range and first anomaly type of the anomaly. The source tracing output can be used to indicate the coordinates of the anomaly source point and the anomaly propagation path (e.g., the point sequence of the anomaly cross-regional propagation path).

[0101] The first recognition sub-model may include a temporal convolutional network (TCN) layer; the second recognition sub-model may include a network layer with a spatial attention mechanism (Squeeze-and-Excitation Networks, SE-Net).

[0102] In addition, the initial image recognition model may also include other network layers. For example, the initial image recognition model may sequentially include a two-dimensional convolutional network layer, a network layer with a spatial attention mechanism, a stacked deep convolutional network layer, a temporal convolutional network layer, a global pooling network layer, and a fully connected feature fusion network layer.

[0103] It should be understood that the embodiments of this disclosure are merely examples. In the case where the initial image recognition model includes a first recognition sub-model and a second recognition sub-model, network layers can be added or reduced according to the actual application scenario. The embodiments of this disclosure do not limit the specific structure of the initial image recognition model.

[0104] By designing a neural network model that integrates the first and second identification sub-models, it is beneficial to adapt to the spatiotemporal distribution characteristics of the sample heat map and improve the accuracy of spatiotemporal distribution feature identification. The first identification sub-model can capture the abnormal transmission pattern of the monitoring values ​​of water environment monitoring indicators in the time dimension, including the time of abnormal occurrence and the speed of abnormal transmission. The first identification sub-model can highlight the spatial correlation characteristics of upstream and downstream monitoring points and improve the image recognition model's ability to identify weak anomalies and cross-regional anomalies.

[0105] In one possible implementation, training an initial image recognition model based on the sample set to obtain a trained target image recognition model includes: dividing the sample set into a training set, a validation set, and a test set according to a preset ratio; training the initial image recognition model based on the training set and the validation set to obtain a first image recognition model; evaluating the performance of the first image recognition model using the test set to obtain an evaluation result; and using the first image recognition model as the target image recognition model if the evaluation result reaches a preset threshold.

[0106] For example, the constructed sample set can be divided into a training set (e.g., 70%), a validation set (e.g., 20%), and a test set (e.g., 10%) according to a preset ratio (e.g., 7:2:1). The training set is input into the initial image recognition model, and the cross-entropy loss function is used as the optimization objective. The Adaptive Moment Estimation (Adam) optimizer is used to adjust the parameters of the initial image recognition model, and the loss value of the cross-entropy loss function is minimized through backpropagation. The performance of the initial image recognition model is verified in real time using the validation set, and the hyperparameters such as the learning rate, convolutional kernel size, and number of network layers of the initial image recognition model are adjusted to avoid overfitting or underfitting of the initial image recognition model until the first image recognition model is obtained.

[0107] Then, the performance of the trained first image recognition model can be evaluated using a test set to obtain the evaluation results. If the anomaly recognition accuracy and source tracing accuracy of the first image recognition model do not reach the preset threshold (e.g., 90%), the sample size in the sample set can be increased (e.g., supplementing samples of special anomaly scenarios), or the structure of the image recognition model can be adjusted, and the training process can be repeated until the evaluation result of the first image recognition model reaches the preset threshold. The first image recognition model can then be used as the target image recognition model.

[0108] In this way, the parameters of the initial image recognition model can be optimized by training with a sample set. The trained target image recognition model can be adapted to different target water environment monitoring areas and different abnormal scenarios. It can accurately identify weak anomalies and cross-regional anomalies, reduce missed reports and false reports, and improve recognition accuracy compared with threshold judgment methods in related technologies.

[0109] After obtaining the trained target image recognition model in step S13, in step S14, based on the trained target image recognition model, the real-time updated heat map to be tested within the target water environment monitoring area can be identified, and based on the identification results of the heat map to be tested, anomaly identification, source tracing and operation and maintenance can be realized.

[0110] In one possible implementation, the identification result may include anomaly identification results and source tracing results for the target water environment monitoring area. The anomaly identification results and the source tracing results can be used to trigger early warning information for the target water environment monitoring area. The anomaly identification results include a first anomaly type and the location of the anomaly area. The source tracing results include the location of the anomaly source and the confidence level of the location of the anomaly source.

[0111] For example, a heat map of the target water environment monitoring area can be acquired in real time. This heat map has the same size and resolution as the sample heat map. The method for acquiring the heat map can be referenced from the sample heat map, and will not be elaborated here.

[0112] The real-time acquired heat map to be tested can be input into the target image recognition model for detection. The target image recognition model can extract the spatiotemporal distribution features in the heat map to be tested through the first recognition sub-model and the second recognition sub-model. These spatiotemporal distribution features may include the color features of the abnormal area (e.g., the depth of color), spatial location features (e.g., the upstream and downstream monitoring points of the abnormal transmission), and time evolution features (e.g., the time of the abnormality and the speed of the abnormal transmission).

[0113] The target image recognition model can determine whether there are water anomalies in the heat map under test based on the extracted spatiotemporal distribution features and output the recognition results, which may include anomaly recognition results and source tracing results.

[0114] The anomaly identification results may include at least one of a first anomaly type and an anomaly area location. The first anomaly type may include water quality anomaly or water quantity anomaly, the degree of anomaly corresponding to the water quality anomaly or water quantity anomaly (e.g., mild anomaly, severe anomaly), and the type of anomaly event, such as rainwater and sewage mixing, backflow, inflow and infiltration, etc.; the anomaly area location may correspond to the spatiotemporal coordinates on the heat map to be measured, that is, a certain time period and a certain upstream and downstream monitoring point;

[0115] Furthermore, the target image recognition model can also deduce the transmission path of anomalies based on the spatiotemporal distribution characteristics of abnormal areas and the correlation between upstream and downstream monitoring points, thus locating the source of the anomaly and forming a closed-loop management system. For example, if an anomaly first appears at an upstream monitoring point, and similar anomalies gradually spread downstream, the source of the anomaly is determined to be a certain upstream point. The target image recognition model can output the source point of the anomaly and its corresponding confidence level (e.g., 95%) as the source tracing result, which can provide a reference for subsequent handling.

[0116] Once the anomaly identification and source tracing results are obtained, an early warning result for the target water environment monitoring area can be triggered. This early warning result can be used to quickly locate the source of the anomaly and its transmission path, providing accurate guidance for water environment emergency response, shortening the response time, and improving governance efficiency.

[0117] In this way, early warning and source tracing can be integrated, combining anomaly identification with upstream and downstream source tracing. This not only triggers early warnings and alerts in real time, but also quickly locates the source of anomalies and their transmission paths, providing precise guidance for water environment emergency response, shortening response time, and improving governance efficiency.

[0118] In one possible implementation, after obtaining the trained target image recognition model, the target image recognition model can be further optimized to obtain a better target image recognition model. For example, if the recognition result of the target image recognition model meets preset conditions, new samples are determined based on the recognition result; the target image recognition model is then trained based on the new samples to obtain the optimized target image recognition model.

[0119] Among them, the preset condition can be the confirmation information of user feedback. That is, after the trained target image recognition model outputs the recognition result, the user provides feedback on the accuracy of the recognition result. If the recognition result is confirmed by the user, it means that the recognition result meets the preset condition. The heat map to be tested corresponding to the recognition result can be used as a new sample, and the target image recognition model can continue to be trained based on the new sample to obtain the optimized target image recognition model.

[0120] In real-world scenarios, newly added samples that meet preset conditions can be collected at fixed intervals (such as daily, weekly, monthly, etc.) and input into the target image recognition model for incremental training. The parameters of the target image recognition model can be fine-tuned to obtain a target image recognition model that can adapt to newly added samples and has better recognition accuracy, thereby improving the accuracy and adaptability of anomaly identification and source tracing.

[0121] In one possible implementation, the method further includes: acquiring influencing factors of anomalies in the target water environment monitoring area, the influencing factors including at least one of rainfall events and hydraulic facility operation events; and performing correlation analysis between the source tracing results and the influencing factors to determine the cause information of the anomalies.

[0122] The methods for association analysis may include association rule mining algorithms (Apriori), FP-growth (Frequent Pattern Growth Algorithm), statistical correlation analysis, grey relational analysis, etc., and the embodiments disclosed herein do not impose specific limitations on them.

[0123] For example, correlation analysis can be performed on the source tracing results and each influencing factor to obtain the correlation coefficient of each influencing factor. Influencing factors with correlation coefficients exceeding the preset correlation threshold are then summarized. The summarized results can be used as anomaly cause information, which can be used to generate anomaly source tracing and diagnosis reports.

[0124] For example, the correlation coefficient of rainfall events can be obtained by identifying, merging, classifying, and performing correlation analysis. First, a rainfall event can be defined as starting when the rainfall intensity exceeds a minimum rainfall intensity threshold, and ending when the intensity falls below the threshold but is extended for a certain period (e.g., the duration of the rainfall's impact). The cumulative rainfall during the event is calculated and classified according to relevant water resources department standards (e.g., light rain, moderate rain, heavy rain, torrential rain, extremely heavy rain, exceptionally heavy rain), yielding rainfall data. Then, correlation analysis can be performed between the time-series data of the anomaly sources indicated by the source tracing results and the rainfall data to output the correlation coefficient of the rainfall events.

[0125] For example, data fusion can be performed on operational events of hydraulic structures (such as pumping stations, gates, and diversion facilities) to obtain operational data of hydraulic structures. Correlation analysis can be performed on the operational data of hydraulic structures and the time-series data of the abnormal sources indicated by the source tracing results to output the correlation coefficients of the operational events of hydraulic structures.

[0126] It should be understood that since video data, remote sensing data, and aerial survey data in the target water environment monitoring area often carry information closely related to influencing factors, correlation analysis can be used to determine the correlation coefficients of video data, remote sensing data, and aerial survey data, which will not be elaborated here.

[0127] By correlating the source tracing results with various influencing factors, the causes of anomalies can be identified quickly and accurately.

[0128] In summary, the water environment monitoring method of this disclosure realizes fully automated operation from data acquisition, preprocessing, heat map generation to anomaly identification, early warning and source tracing. The update frequency is synchronized with the data acquisition frequency, which can meet the needs of real-time water environment monitoring, greatly reduce manual intervention and improve the intelligence level of water environment monitoring.

[0129] In addition, this disclosure also provides a water environment monitoring system, electronic equipment, computer-readable storage medium, and program, all of which can be used to implement any of the water environment monitoring methods provided in this disclosure. The corresponding technical solutions and descriptions are described in the relevant section of the method and will not be repeated here.

[0130] Figure 3 A block diagram of the water environment monitoring system provided in this disclosure embodiment is shown, such as... Figure 3 As shown, the system includes:

[0131] The acquisition module 41 is used to acquire the first time series data of water environment monitoring indicators at different monitoring points within the target water environment monitoring area, wherein the water environment monitoring indicators include at least one of water quality indicators, flow rate and liquid level.

[0132] The construction module 42 is used to construct a sample set based on the topological relationship between different monitoring points and the first time series data. The sample set includes sample heat maps corresponding to the water environment monitoring indicators. The sample heat maps are used to reflect the spatiotemporal distribution characteristics of the monitoring values ​​of the corresponding water environment monitoring indicators.

[0133] Training module 43 trains an initial image recognition model based on the sample set to obtain a trained target image recognition model;

[0134] The identification module 44 is used to determine the identification result of the heat map to be measured based on the real-time updated heat map to be measured within the target water environment monitoring area and the target image recognition model. The identification result includes the anomaly identification result and the source tracing result of the target water environment monitoring area. The anomaly identification result and the source tracing result are used to trigger the early warning information of the target water environment monitoring area. The anomaly identification result includes a first anomaly type and an anomaly area location. The source tracing result includes the anomaly source location and the confidence level of the anomaly source location.

[0135] In one possible implementation, the acquisition module 41 is used to: select at least one water environment monitoring indicator for evaluating the target water environment monitoring area from a variety of preset initial water environment monitoring indicators; and for each selected water environment monitoring indicator, acquire the first time series data of the water environment monitoring indicator at different monitoring points within the target water environment monitoring area.

[0136] In one possible implementation, at least one water environment monitoring indicator for evaluating the target water environment monitoring area is selected from a plurality of preset initial water environment monitoring indicators. This includes: determining the environmental disturbance resistance of each initial water environment monitoring indicator based on the water body type of the target water environment monitoring area, wherein the environmental disturbance resistance indicates the ability of the initial water environment monitoring indicator to resist environmental disturbances in the target water environment monitoring area; determining the discrimination of each initial water environment monitoring indicator based on different historical second anomaly types existing in the target water environment monitoring area, wherein the discrimination indicates the strength of specificity exhibited by different initial water environment monitoring indicators in identifying different second anomaly types; determining a score for each initial water environment monitoring indicator based on the environmental disturbance resistance and the discrimination; and selecting at least one water environment monitoring indicator for evaluating the target water environment monitoring area from a plurality of initial water environment monitoring indicators based on the score.

[0137] In one possible implementation, the construction module 42 is used to: determine the upstream and downstream positional relationship between different monitoring points based on the hydraulic flow direction information and spatial connection relationship of different monitoring points within the target water environment monitoring area; use different collection times arranged chronologically in the first time series data as the vertical coordinate of the sample heat map, and use different monitoring points arranged according to the upstream and downstream positional relationship within the target water environment monitoring area as the horizontal coordinate of the sample heat map; for any monitoring point and any collection time, determine the pixel value of the sample heat map in the sample set at the coordinate position defined by the horizontal coordinate corresponding to the monitoring point and the vertical coordinate corresponding to the collection time, based on the monitoring value of the water environment monitoring index of the monitoring point at the collection time.

[0138] In one possible implementation, the initial image recognition model includes at least a first recognition sub-model and a second recognition sub-model. The first recognition sub-model is used to capture the abnormal transmission characteristics of the monitoring values ​​of the water environment monitoring indicators in the time dimension from the sample heat map. The abnormal transmission characteristics include the time of abnormal occurrence and the abnormal transmission speed. The second recognition sub-model is used to capture the spatial location characteristics between different monitoring points from the sample heat map. The spatial location characteristics include the upstream and downstream positional relationships between different monitoring points.

[0139] In one possible implementation, the training module 43 is further configured to: determine new samples based on the recognition results when the recognition results meet preset conditions; and train the target image recognition model based on the new samples to obtain an optimized target image recognition model.

[0140] In one possible implementation, the identification module 44 is further configured to: acquire the influencing factors of the anomaly in the target water environment monitoring area, the influencing factors including at least one of rainfall events and hydraulic facility operation events; and perform correlation analysis between the source tracing results and the influencing factors to determine the anomaly cause information.

[0141] In some embodiments, the system provided in this disclosure may have functions or include modules that can be used to execute the methods described in the above method embodiments. The specific implementation of these methods can be referred to the description in the above method embodiments, and for the sake of brevity, they will not be repeated here.

[0142] This disclosure also proposes a computer-readable storage medium storing computer program instructions that, when executed by a processor, implement the above-described method. The computer-readable storage medium can be volatile or non-volatile.

[0143] This disclosure also proposes an electronic device, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to invoke the instructions stored in the memory to execute the above-described method.

[0144] This disclosure also provides a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code, wherein when the computer-readable code is run in a processor of an electronic device, the processor in the electronic device performs the above-described method.

[0145] Electronic devices can be provided as terminals, servers, or other forms of devices.

[0146] Figure 4 A block diagram of an electronic device 1900 according to an embodiment of the present disclosure is shown. For example, the electronic device 1900 may be provided as a server or a terminal device. (Refer to...) Figure 4 The electronic device 1900 includes a processing component 1922, which further includes one or more processors, and memory resources represented by memory 1932 for storing instructions, such as application programs, that can be executed by the processing component 1922. The application programs stored in memory 1932 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 1922 is configured to execute instructions to perform the methods described above.

[0147] Electronic device 1900 may also include a power supply component 1926 configured to perform power management of electronic device 1900, a wired or wireless network interface 1950 configured to connect electronic device 1900 to a network, and an input / output interface 1958. Electronic device 1900 can operate on an operating system stored in memory 1932, such as a server operating system, a graphical user interface-based operating system, or a multi-user, multi-process computer operating system (Unix). TM Linux is a free and open-source Unix-like operating system. TM ), an open-source Unix-like operating system (FreeBSD) TM (or similar.)

[0148] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions that can be executed by a processing component 1922 of an electronic device 1900 to perform the above-described method.

[0149] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0150] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example, (but not limited to) electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0151] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0152] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0153] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0154] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0155] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0156] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0157] The computer program product can be implemented specifically through hardware, software, or a combination thereof. In one alternative embodiment, the computer program product is specifically embodied in a computer storage medium; in another alternative embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0158] The description of the various embodiments above tends to emphasize the differences between the various embodiments. The similarities or similarities between them can be referred to, and for the sake of brevity, they will not be repeated here.

[0159] Those skilled in the art will understand that, in the above-described method of the specific implementation, the order in which each step is written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0160] If the technical solution of this application involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the individual through pop-up information or by asking the individual to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

[0161] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A water environment monitoring method, characterized in that, The method includes: Acquire first time-series data of water environment monitoring indicators at different monitoring points within the target water environment monitoring area, wherein the water environment monitoring indicators include at least one of water quality indicators, flow rate, and liquid level; Based on the topological relationship between different monitoring points and the first time series data, a sample set is constructed. The sample set includes sample heat maps corresponding to the water environment monitoring indicators. The sample heat maps are used to reflect the spatiotemporal distribution characteristics of the monitoring values ​​of the corresponding water environment monitoring indicators. The initial image recognition model is trained based on the sample set to obtain the trained target image recognition model. Based on the real-time updated heat map of the target water environment monitoring area and the target image recognition model, the recognition result of the heat map is determined. The recognition result includes the anomaly recognition result and the source tracing result of the target water environment monitoring area. The anomaly recognition result and the source tracing result are used to trigger the early warning information of the target water environment monitoring area. The anomaly recognition result includes a first anomaly type and the location of the anomaly area. The source tracing result includes the location of the anomaly source point and the confidence level of the anomaly source point.

2. The method according to claim 1, characterized in that, Obtain first-time-series data of water environment monitoring indicators at different monitoring points within the target water environment monitoring area, including: From a variety of preset initial water environment monitoring indicators, at least one water environment monitoring indicator is selected to evaluate the target water environment monitoring area; For each selected water environment monitoring indicator, the first time-series data of the water environment monitoring indicator at different monitoring points within the target water environment monitoring area are obtained.

3. The method according to claim 2, characterized in that, From a variety of preset initial water environment monitoring indicators, at least one water environment monitoring indicator is selected for evaluating the target water environment monitoring area, including: Based on the water body type of the target water environment monitoring area, the environmental disturbance resistance of each initial water environment monitoring indicator is determined. The environmental disturbance resistance is used to indicate the ability of the initial water environment monitoring indicator to resist environmental interference in the target water environment monitoring area. Based on the different historical second anomaly types existing in the target water environment monitoring area, the discrimination degree of each initial water environment monitoring indicator is determined. The discrimination degree is used to indicate the strength of the specificity of different initial water environment monitoring indicators in identifying different second anomaly types. The score for each initial water environment monitoring indicator is determined based on its resistance to environmental disturbance and its discriminative power. Based on the scores, at least one water environment monitoring indicator is selected from a variety of initial water environment monitoring indicators to evaluate the target water environment monitoring area.

4. The method according to claim 1, characterized in that, Based on the topological relationships between different monitoring points and the first time-series data, a sample set is constructed, including: Based on the hydraulic flow direction information and spatial connection relationship of different monitoring points within the target water environment monitoring area, the upstream and downstream positional relationship between different monitoring points is determined; The different collection times arranged chronologically in the first time series data are used as the vertical axis of the sample heat map, and the different monitoring points arranged according to the upstream and downstream positional relationship within the target water environment monitoring area are used as the horizontal axis of the sample heat map. For any monitoring point and any collection time, the pixel value of the sample heat map in the sample set is determined based on the monitoring value of the water environment monitoring index at that collection time. The pixel value is determined at the coordinate position defined by the horizontal axis corresponding to the monitoring point and the vertical axis corresponding to the collection time.

5. The method according to any one of claims 1 to 4, characterized in that, The initial image recognition model includes at least a first recognition sub-model and a second recognition sub-model. The first recognition sub-model is used to capture the abnormal transmission characteristics of the monitoring values ​​of the water environment monitoring indicators in the time dimension from the sample heat map. The abnormal transmission characteristics include the time of abnormal occurrence and the abnormal transmission speed. The second recognition sub-model is used to capture the spatial location characteristics between different monitoring points from the sample heat map. The spatial location characteristics include the upstream and downstream positional relationships between different monitoring points.

6. The method according to any one of claims 1 to 4, characterized in that, The method further includes: If the recognition result meets the preset conditions, new samples are determined based on the recognition result; The target image recognition model is trained based on the newly added samples to obtain the optimized target image recognition model.

7. The method according to any one of claims 1 to 4, characterized in that, The method further includes: The influencing factors of anomalies in the target water environment monitoring area are obtained, and the influencing factors include at least one of rainfall events and hydraulic facility operation events; The source tracing results are correlated with the influencing factors to determine the cause information of the anomaly.

8. A water environment monitoring system, characterized in that, The system includes: The acquisition module is used to acquire the first time-series data of water environment monitoring indicators at different monitoring points within the target water environment monitoring area, wherein the water environment monitoring indicators include at least one of water quality indicators, flow rate and liquid level; The construction module is used to construct a sample set based on the topological relationship between different monitoring points and the first time series data. The sample set includes sample heat maps corresponding to the water environment monitoring indicators. The sample heat maps are used to reflect the spatiotemporal distribution characteristics of the monitoring values ​​of the corresponding water environment monitoring indicators. The training module trains an initial image recognition model based on the sample set to obtain a trained target image recognition model. The identification module is used to determine the identification result of the heat map to be measured based on the real-time updated heat map of the target water environment monitoring area and the target image recognition model. The identification result includes the anomaly identification result and the source tracing result of the target water environment monitoring area. The anomaly identification result and the source tracing result are used to trigger the early warning information of the target water environment monitoring area. The anomaly identification result includes a first anomaly type and the location of the anomaly area. The source tracing result includes the location of the anomaly source point and the confidence level of the anomaly source point.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.