A method and system for river health state assessment and early warning

Abstract

The application discloses a kind of river health state evaluation and early warning method and system, belong to ecological environment monitoring and automation control technical field, which comprises: obtaining the real-time monitoring data of river monitoring point;According to real-time monitoring data, the final water quality grade of current time is evaluated, and the diurnal comprehensive index of current time is calculated;Obtain the historical reference diurnal comprehensive index based on historical monitoring data, and calculate the dynamic early warning threshold according to the historical reference diurnal comprehensive index and the final water quality grade of current time;The diurnal comprehensive index of current time is compared with the dynamic early warning threshold, when the diurnal comprehensive index of current time is lower than dynamic early warning threshold, generate early warning information, and generate control instruction according to the comparison result, to adjust the operating parameter of front-end monitoring equipment.The present application provides a complete technical solution of precision, intelligentization and automation for the long-term supervision of "Sewage Zero Direct Discharge Area".

Description

Technical Field

[0001] This invention relates to the field of ecological environment monitoring and automated control technology, and in particular to a method and system for assessing and providing early warning of river health status. Background Technology

[0002] Establishing a long-term monitoring mechanism for "zero direct discharge zones" is one of the core tasks of current water environment governance. However, existing river water quality monitoring technologies face significant bottlenecks in practical applications, mainly in the following two aspects:

[0003] First, there is a conflict between monitoring frequency and operation and maintenance costs: Traditional methods, which mainly rely on manual sampling or low-frequency automatic detection (such as every 4 hours), are prone to missing sudden or intermittent instantaneous sewage discharge events due to the excessively long sampling intervals, resulting in regulatory blind spots. If high-frequency monitoring (such as every 5 minutes) is simply implemented to increase the detection probability, it will lead to a surge in the consumption of monitoring reagents and serious waste of equipment energy. At the same time, the transmission and storage of massive amounts of redundant data will also place excessive demands on communication bandwidth and infrastructure, resulting in a dilemma of "precision versus cost".

[0004] Secondly, there is the problem of alarm mechanisms "failing" in dynamic environments: most existing automatic monitoring systems adopt static, fixed threshold alarm strategies (such as triggering an alarm when water quality exceeds Class V standards), lacking the ability to dynamically adapt to changes in environmental background values. On the one hand, during dry seasons or when background values ​​fluctuate significantly, this mechanism is prone to misinterpreting normal sensor fluctuations as pollution events, leading to a persistently high false alarm rate. On the other hand, when water quality background values ​​remain at a poor level for an extended period, a fixed, lenient threshold may mask further deterioration trends, creating a risk of missed alarms. More importantly, this "post-event alarm" mode can only reflect whether water quality exceeds standards, failing to quantitatively assess the overall health status of the river and its changing trends. Furthermore, the system operates in isolation, unable to immediately link with front-end equipment upon detecting anomalies, thus missing the optimal window for pollution tracking and containment.

[0005] Therefore, there is an urgent need for a method for assessing and issuing early warnings of river health status. This method should integrate high-frequency monitoring data with environmental background parameters to achieve a shift from "fixed threshold alarms" to "health trend early warnings" through quantitative assessment of river water quality. This method should be able to accurately identify instantaneous abnormal fluctuations and potential deterioration trends in water quality, addressing the problems of balancing monitoring frequency and cost, and high false alarm / missed alarm rates in existing technologies. Furthermore, it should be able to dynamically adjust the operating logic of monitoring equipment based on the assessment results, providing precise and intelligent technical support for the long-term supervision of "zero direct discharge zones" for wastewater. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for assessing and warning the health status of rivers, in order to solve the problems in existing river monitoring, such as the difficulty in balancing operation and maintenance costs and capture probability due to the single sensing and identification dimension and rigid sampling strategy, the frequent false alarms and missed alarms of static thresholds in dynamic background environments, and the lack of predictive ability for water quality evolution trends.

[0007] To achieve the above objectives, this application adopts the following technical solution:

[0008] This application discloses a method for assessing and providing early warning of river health status, comprising the following steps:

[0009] Acquire real-time monitoring data from river monitoring points, including monitoring values ​​of various water quality indicators;

[0010] The final water quality level at the current moment is assessed based on the real-time monitoring data, and the daily comprehensive index at the current moment is calculated.

[0011] Obtain the historical baseline daily comprehensive index based on historical monitoring data, and calculate the dynamic early warning threshold based on the historical baseline daily comprehensive index and the final water quality level at the current moment;

[0012] The daily comprehensive index at the current moment is compared with the dynamic early warning threshold. When the daily comprehensive index at the current moment is lower than the dynamic early warning threshold, an early warning message is generated, and a control command is generated based on the comparison result to adjust the operating parameters of the front-end monitoring equipment.

[0013] Preferably, the step of assessing the final water quality level at the current moment based on the real-time monitoring data includes:

[0014] The original water quality level at the current moment is obtained by comparing the real-time monitoring data with the water quality standards;

[0015] Obtain the final water quality level at the previous moment, and calculate the change in water quality level between the original water quality level and the final water quality level at the previous moment;

[0016] The original water quality level, the final water quality level at the previous moment, and the change in level are weighted and combined to obtain the final water quality level at the current moment.

[0017] The filter weight coefficients of the weighted combination are dynamically adjusted based on the river flow velocity information.

[0018] Preferably, the calculation of the daily composite index at the current moment includes:

[0019] Calculate the real-time health index based on the real-time monitoring data.

[0020] The daily comprehensive index for the current moment is obtained by performing a time-decay weighted average of the real-time health index at each moment within a preset time period.

[0021] Preferably, the step of calculating the real-time health index based on the real-time monitoring data includes:

[0022] Each monitoring value is converted into a dimensionless single health sub-index, with different conversion rules used for indicators with different physical characteristics;

[0023] Each individual health sub-index is positively oriented, and the information entropy of each index is calculated based on the positively oriented values.

[0024] The dynamic weights of each individual indicator are determined based on the information entropy, and the individual health sub-indices are weighted and summed using the dynamic weights to obtain the real-time health index at the current moment.

[0025] Preferably, the calculation of the dynamic early warning threshold based on the historical baseline daily comprehensive index and the final water quality level at the current moment includes:

[0026] The tolerance coefficient for the water quality level is determined based on the final water quality level at the current moment, and the environmental correction coefficient is determined based on the environmental background parameters.

[0027] Calculate the product of the tolerance coefficient and the environmental correction coefficient and subtract it from 1. Multiply the historical baseline daily comprehensive index by the difference to obtain the dynamic early warning threshold.

[0028] As a preferred option, after generating the warning information, the following should also be included:

[0029] Extract the real-time health index from multiple monitoring points before the triggering of the early warning and calculate its fluctuation characteristics;

[0030] The confidence level of this triggering event is calculated based on the fluctuation characteristics and the historical benchmark daily composite index;

[0031] An alarm command is output when the confidence level is greater than or equal to a preset threshold; otherwise, the warning information is blocked.

[0032] Preferably, the step of generating control commands based on the comparison results to adjust the operating parameters of the front-end monitoring equipment includes:

[0033] When the real-time health index is in the excellent range and the fluctuation range is less than the preset threshold, the sampling frequency of the monitoring terminal is reduced; when a valid early warning event is detected or the rate of change of the real-time health index exceeds the preset threshold, the sampling frequency of the monitoring terminal is increased.

[0034] Preferably, the step of generating control commands based on the comparison results to adjust the operating parameters of the front-end monitoring equipment further includes at least one of the following:

[0035] When a valid early warning event is triggered, the image acquisition device associated with the current monitoring point is determined according to the preset device association relationship, and a control command is generated to drive the image acquisition device to acquire images of the area where the monitoring point is located, and the acquired image data is marked and stored.

[0036] When the daily comprehensive index at the current moment is detected to be lower than the dynamic warning threshold, and the confidence level is less than the preset threshold and the fluctuation characteristics exceed the preset threshold, a sensor self-cleaning command is generated, and a retest is performed after the cleaning is completed; or when it is determined from the image data fed back by the image acquisition device that there is no obvious abnormality on the water surface, a sensor self-cleaning command is generated, and a retest is performed after the cleaning is completed.

[0037] A system for assessing and providing early warning of river health status includes:

[0038] The monitoring terminal is set at the river monitoring point to collect real-time monitoring data, which includes the monitoring values ​​of various water quality indicators.

[0039] An image acquisition device is installed in the area where the monitoring point is located to acquire video images;

[0040] An edge computing gateway is communicatively connected to the monitoring terminal and the image acquisition device, respectively. The edge computing gateway includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method as described in any one of the above.

[0041] A computer-readable storage medium storing a computer program that, when executed by a computer, enables a method for assessing and providing early warning of river health status as described in any one of the preceding descriptions.

[0042] The present invention has the following beneficial effects:

[0043] 1. By adaptively adjusting the sampling frequency, while ensuring monitoring accuracy during dry and polluted periods, the average reagent consumption of the equipment can be reduced by more than 30%, and communication traffic can be saved by 40%, effectively solving the problem of high cost caused by high-frequency monitoring.

[0044] 2. By introducing a dynamic early warning threshold generation mechanism based on water quality levels and environmental background parameters, the sensitivity of early warnings adaptively adjusts to the actual risk status of the river, effectively eliminating false alarms caused by changes in the environmental background. Simultaneously, discrete state filtering technology smooths sensor noise, and a confidence assessment mechanism is introduced to perform secondary verification of suspected early warnings, distinguishing between actual pollution trends and transient disturbances. Furthermore, the newly added intelligent cleaning control mechanism can identify and eliminate abnormal data caused by probe clogging, biofouling, etc., and automatically retests after cleaning. The synergistic effect of these multiple mechanisms reduces the traditional false alarm rate from 15% in existing technologies to below 2%, significantly improving regulatory efficiency and user experience.

[0045] 3. The deep integration and coordinated control of water quality monitoring and video surveillance have enabled "monitoring as source tracing". This has not only completely changed the passive situation of traditional manual investigation being time-consuming and difficult to obtain evidence, but also provided intuitive and reliable video evidence for pollution source tracing, law enforcement penalties and responsibility determination, which has greatly improved the efficiency and deterrent effect of supervision and law enforcement. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart of a method for assessing and providing early warning of river health status, provided in an embodiment of this application.

[0048] Figure 2 This is a flowchart illustrating the final water quality rating process provided in the embodiments of this application;

[0049] Figure 3 This is a flowchart illustrating the calculation of real-time health indexes provided in an embodiment of this application;

[0050] Figure 4 This is a flowchart illustrating the calculation of the dynamic early warning threshold provided in an embodiment of this application;

[0051] Figure 5 This is a flowchart illustrating the determination of early warning information provided in an embodiment of this application;

[0052] Figure 6 This is a schematic diagram of the structure of a system for assessing and warning the health status of a river channel, as provided in an embodiment of this application. Detailed Implementation

[0053] To make the technical solution of this application clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The terms "first," "second," etc., in the claims and specification of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate. This is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of units is not necessarily limited to those units, but may include other units not explicitly listed or inherent to these processes, methods, products, or apparatuses.

[0054] Example 1

[0055] like Figure 1 As shown, a method for assessing and providing early warning of river health status includes the following steps:

[0056] S110. Obtain real-time monitoring data from river monitoring points, including monitoring values ​​of various water quality parameters.

[0057] Monitoring terminals set up at river monitoring points collect monitoring data in real time, including the values ​​of various water quality parameters (such as pH, DO, and ammonia nitrogen), and transmit this data to an edge gateway. Upon receiving the data, the edge gateway first calculates the deviation rate between the current value of each indicator and the corresponding indicator in the previous frame. If the deviation rates of all indicators are below a set threshold (e.g., 2%) and the mandatory heartbeat time (e.g., 1 hour) has not been reached, the transmission is blocked to save communication bandwidth; otherwise, subsequent steps are executed. This edge preprocessing mechanism effectively reduces the transmission of redundant data, laying the foundation for balancing high-frequency monitoring and low power consumption.

[0058] S120. Determine the final water quality level at the current moment based on real-time monitoring data, and calculate the daily comprehensive index at the current moment.

[0059] This step is based on the real-time monitoring data obtained in step S110. On the one hand, it evaluates the final water quality level at the current moment through discrete state filtering to eliminate sensor noise and respond quickly to sudden pollution changes. On the other hand, it calculates the daily comprehensive index at the current moment based on the historical monitoring data of the day, providing an overall water quality assessment for the current period for the subsequent generation of dynamic thresholds.

[0060] Furthermore, such as Figure 2 As shown, the final water quality level at the current moment is assessed based on real-time monitoring data, including the following steps:

[0061] S210. Obtain the original water quality level at the current moment by comparing the real-time monitoring data with the water quality standards;

[0062] S220. Obtain the final water quality level at the previous moment and calculate the change in water quality level between the original water quality level and the final water quality level at the previous moment.

[0063] S230. The original water quality level, the final water quality level at the previous moment, and the change in level are weighted and combined to obtain the final water quality level at the current moment; wherein, the filter weight coefficient of the weighted combination is dynamically adjusted according to the river flow velocity information.

[0064] Specifically, the concentration values ​​of each monitoring indicator are first converted into the corresponding water quality category according to national water quality standards (such as GB 3838-2002) to obtain the original water quality level at the current time t. This level is typically represented by values ​​from Class I to Class V (worst), corresponding to values ​​of 1 to 5 (or higher), with higher values ​​indicating worse water quality. This initial level directly reflects the qualitative results of the current monitoring data according to national standards, but it is susceptible to noise from individual measurements.

[0065] Then obtain the final water quality level at the previous time t-1. This value is the stable output after filtering at the previous moment, and the level change is defined. It measures the magnitude of the change in the current original level relative to the final level at the previous moment, and is a key indicator for identifying sudden pollution events.

[0066] Then, a discrete-state filter is constructed, the output of which is... It is a weighted sum of the original water quality level at the current moment, the final water quality level at the previous moment, and the change in level. The weighting coefficient can adapt to the changes in the river hydrological characteristics, thereby achieving a "stable but not dull" evaluation effect.

[0067] In this operation, two dynamic factors are first introduced to characterize the impact of the river environment on the filtering requirements:

[0068] 1. Flow velocity influencing factors ,in, Let be the measured flow velocity at the current time t. This represents the historical maximum flow rate. Higher flow rates result in more thorough water mixing, potentially smoothing out sensor noise. Therefore, the weight of the current monitoring value should be appropriately reduced, while the weight of the final water quality grade from the previous moment should be increased.

[0069] 2. Volatility Influencing Factors ,in, The standard deviation of the final water quality grade over the past 24 hours. This is the preset upper limit of the historical standard deviation. This factor reflects the degree of recent water quality fluctuations. The greater the fluctuation, the more unstable the water quality is, and the more necessary it is to enhance the ability to respond to sudden changes. Therefore, the weight of the grade change should be increased.

[0070] At the same time, set the basic weight vector. These correspond to the basic weights of the current item (the original water quality level at the current moment), the historical item (the final water quality level at the previous moment), and the abrupt change item (the change in level), respectively. The basic weights are dynamically adjusted using adjustment matrix A, which is designed as follows:

[0071] .

[0072] The actual weight vector is then... .

[0073] in, , , These are the weights of the current item, historical items, and mutation items, respectively, and they satisfy the following conditions: + + This design enables the following: when the flow rate increases, the weight of the current item decreases and the weight of the historical item increases; when historical fluctuations increase, the weight of the abrupt change item increases and the weight of the current item decreases accordingly, thereby adaptively adjusting the filtering characteristics.

[0074] Finally, the final water quality level at the current moment. The calculation formula is as follows:

[0075] .

[0076] This discrete-state filter combines smoothness and sensitivity: in routine monitoring, the final water quality level of the previous moment dominates, effectively suppressing random noise from the sensor; when sudden pollution causes abrupt changes in the level, the weight of the level change is increased, enabling the filtering result to quickly track actual water quality changes. This method overcomes the misjudgment and loopholes caused by traditional single thresholds or simple averaging, providing accurate and stable water quality level input for subsequent early warning.

[0077] After completing the water quality rating, this step also calculates the daily comprehensive index for the current moment based on the real-time health index at each time of the day.

[0078] Furthermore, such as Figure 3 As shown, the real-time health index is calculated based on real-time monitoring data, including the following steps:

[0079] S310. Convert each monitoring value into a dimensionless single health sub-index, with different conversion rules used for indices with different physical characteristics.

[0080] S320. Perform positive transformation on each individual health sub-index, and calculate the information entropy of each index based on the positive transformation values.

[0081] S330. Determine the dynamic weight of each individual indicator based on information entropy, and use the dynamic weight to perform a weighted summation of the individual health sub-indices to obtain the real-time health index at the current moment.

[0082] The conversion is performed based on the characteristics of the monitored indicator. Specifically, for very small indicators (ammonia nitrogen, the lower the better), the conversion formula is as follows:

[0083] .

[0084] in, Indicates the monitored value. and These represent the lower limit for Class V (worst) water quality and the upper limit for Class I (worst) water quality, respectively. When < When, take 1; when > When the index is set to 0, the sub-index is ensured to be within the range of [0,1].

[0085] For very large indicators (such as dissolved oxygen, where higher is better), the conversion formula is as follows:

[0086] .

[0087] when > When, take 1; when < Take 0 at that time.

[0088] For moderate indicators (such as pH, the middle is optimal), first calculate the deviation. Then, normalization is performed (assuming a maximum allowable deviation of 0.5), as shown in the following formula:

[0089] .

[0090] when When <0, take 0.

[0091] The above transformation converts heterogeneous monitoring data from multiple water quality sensors into a single health sub-index within the range [0,1], based on the optimization direction of each indicator (extremely large, extremely small, or moderate). The higher the value, the healthier the water quality. This standardization process eliminates dimensional differences between different indicators, laying a mathematical foundation for subsequent multi-indicator fusion calculations.

[0092] After converting each monitoring value into its corresponding individual health sub-index, first convert all of them into positive values. That is, subtract the fractional exponent from 1: =1- , where i represents time and j represents the index. The positively oriented values. The larger the value, the worse the water quality at that moment, which helps to reflect the fluctuation characteristics of the indicator in subsequent entropy calculations.

[0093] Then calculate the entropy value of the j-th indicator within a preset historical time period (e.g., the past 24 hours):

[0094] .

[0095] Where n is the number of samples, when =0 is defined Entropy It reflects the dispersion of indicator data. The smaller the entropy value, the greater the data fluctuation. The greater the amount of information the indicator provides in the current period, and the stronger its indicative role in water quality changes.

[0096] Then calculate the difference coefficient based on the entropy value. The larger the difference coefficient, the more important the indicator, thus yielding the dynamic weights of each indicator:

[0097] .

[0098] Where m represents the total number of indicators. This weight is objectively generated entirely based on the fluctuation characteristics of the data itself, avoiding the subjective bias of expert scoring or fixed weights in traditional methods.

[0099] Finally, the individual health scores are weighted and summed using dynamic weights, and then mapped to a 0-5 scale to obtain the real-time health index for the current moment.

[0100] .

[0101] This index condenses multi-dimensional water quality information into a single quantitative indicator. The higher the value, the better the water quality: 5 points corresponds to Class I water, 4 points corresponds to Class II water, and so on, with 0 points corresponding to Class V or worse water.

[0102] After completing the real-time health index calculation, the daily comprehensive index for the current moment is calculated based on the real-time health indices at various times throughout the day using a time-decay weighted average method. Specifically, for the set of monitoring times within the day... Calculate the real-time health index at each time point. The weighted average, with weights determined by the time decay coefficient. The value is determined, where t is the current time and k is a preset attenuation factor (which can be set according to the water body's self-purification capacity and management needs, for example, 0.1 to 0.3). Data closer to the current time is assigned a higher weight, thus reflecting the principle that "recent water quality status is more important for current evaluation." Daily Comprehensive Index The calculation formula is as follows:

[0103] .

[0104] This index condenses the water quality evolution information of the day into a single quantitative value, which can be used to assess the overall water quality level of the day, and can also be used as a historical benchmark to compare with the daily index at subsequent times, providing benchmark data for the generation of dynamic early warning thresholds.

[0105] This step achieves a dual, precise characterization of river water quality by combining discrete-state filtering with time-decayed weighted averaging. Firstly, it dynamically weights and combines the original water quality level, the final water quality level at the previous moment, and the change in level, adaptively adjusting the filtering weights based on river flow velocity. This effectively smooths sensor noise while maintaining a rapid response to sudden pollution events, resolving the contradiction of traditional water quality assessments being either "stable but not sensitive" or "sensitive but not stable." Secondly, it calculates a time-decayed weighted daily comprehensive index based on the real-time health index of the day, ensuring that data closer to the current moment contributes more, accurately reflecting the recent trend of water quality evolution. This step provides a stable and reliable level input and a sensitive comprehensive benchmark for the subsequent generation of dynamic early warning thresholds, laying the technical foundation for "precise assessment and trend early warning."

[0106] S130. Obtain the historical baseline daily comprehensive index based on historical monitoring data, and calculate the dynamic early warning threshold based on the historical baseline daily comprehensive index and the final water quality level at the current moment.

[0107] In this step, the first step is to obtain the historical baseline daily composite index based on historical monitoring data. This index is preferably the composite index of the previous day. This serves as the baseline reference for the day's water quality warning. Subsequently, the final water quality level at the current moment is assessed based on step S120. By combining environmental background parameters, an adaptive dynamic early warning threshold suitable for the current river channel condition is calculated.

[0108] Furthermore, such as Figure 4 As shown, the dynamic early warning threshold is calculated based on the historical baseline daily composite index and the current final water quality level, including the following steps:

[0109] S410. Determine the water quality tolerance coefficient based on the final water quality level at the current moment, and determine the environmental correction coefficient based on the environmental background parameters.

[0110] S420. Calculate the product of the level tolerance coefficient and the environmental correction coefficient and subtract it from 1. Multiply the historical baseline daily comprehensive index by the difference to obtain the dynamic early warning threshold.

[0111] Level Tolerance Coefficient This health index, used to characterize the permissible fluctuation range under current water quality conditions, is negatively correlated with the current water quality level: the better the water quality (the lower the final water quality level value), the stronger the river's ecological resilience, and the larger the permissible range of normal fluctuations. Larger values ​​indicate poorer water quality (higher final water quality rating), meaning weaker river carrying capacity and a greater need to be highly sensitive to any further deterioration. The value is relatively small. Specifically, it can be determined using a piecewise function or a continuous function, for example:

[0112] Class I / II water (corresponding to) =1 or 2): =0.2-0.3;

[0113] Class III / IV water (corresponding to) =3 or 4): =0.1-0.2;

[0114] Class V / Inferior Class V water (corresponding to =5 or above): =0.05.

[0115] Environmental correction factor This threshold is used to reflect the impact of different hydrological seasonal characteristics on early warning sensitivity and is determined based on environmental background parameters (such as dry season, wet season, and normal water season). For example, during the dry season, river flow is low and self-purification capacity is weak, requiring a tightening of the early warning threshold to strengthen monitoring. =1.2; During the high-water season, water quality fluctuates significantly due to rainwater erosion. Therefore, the threshold can be appropriately relaxed to avoid false alarms. =0.8; during the normal water period, take =1.0. Environmental background parameters can be automatically identified through preset calendar rules or real-time hydrological data.

[0116] Finally, dynamic early warning threshold The calculation formula is as follows:

[0117] .

[0118] in This is the historical baseline comprehensive index. The physical meaning of this formula is: using the historical baseline comprehensive index as a baseline, the warning sensitivity is dynamically adjusted based on the current water quality level and environmental conditions. When water quality deteriorates ( (reduced) or during the dry season ( When (increases), (1− × A decrease in the health index leads to a lower dynamic warning threshold, making the system more sensitive to the decline in the health index and able to promptly detect trends of water quality deterioration. Conversely, when the water quality is good or during the high-water season, the threshold increases, allowing for normal fluctuations within a certain range and avoiding false alarms caused by environmental interference.

[0119] This dynamic early warning threshold generation method overcomes the shortcomings of traditional fixed threshold alarms, which are difficult to adapt to different water quality background and environmental conditions. It realizes the adaptive adjustment of early warning sensitivity according to the river risk level, and provides a scientific basis for subsequent accurate early warning.

[0120] S140. Compare the current daily comprehensive index with the dynamic early warning threshold. When the current daily comprehensive index is lower than the dynamic early warning threshold, generate an early warning message and generate a control command based on the comparison result to adjust the operating parameters of the front-end monitoring equipment.

[0121] This step first involves setting the current daily composite index. The dynamic early warning threshold calculated in step S130 Comparison. When < If a suspected warning signal is triggered, a warning message will be generated. This comparison result serves as the basis for subsequent refined processing and closed-loop control.

[0122] Furthermore, such as Figure 5 As shown, after generating the early warning information, the following steps are also included to distinguish between real pollution events and transient disturbances:

[0123] S510. Extract the real-time health index of multiple monitoring points before the triggering of the early warning and calculate its fluctuation characteristics;

[0124] S520. Calculate the confidence level of this triggering event based on the fluctuation characteristics and the historical benchmark daily composite index;

[0125] S530: When the confidence level is greater than or equal to the preset threshold, an alarm command is output; otherwise, the warning information is blocked.

[0126] After the system generates an early warning message, it extracts the real-time health index from multiple consecutive monitoring points (e.g., 5 times) prior to the trigger time. Calculate the standard deviation of these data. As a fluctuation characteristic, the standard deviation reflects the dispersion of water quality data in the short period before the trigger, and is an important indicator for determining whether the data is stable.

[0127] The confidence level is calculated based on this fluctuation characteristic. The formula used to quantify the probability that a current warning is a genuine pollution event is as follows:

[0128] .

[0129] The principle behind this formula is: if the data fluctuation before the trigger is small ( If the data is small, then the current downward trend is more likely to be true contamination, with a confidence level close to 1; if the data fluctuates drastically (…), ... If the noise level is high, it may be caused by the dryness of the tea aroma or momentary interference, which has a low confidence level.

[0130] The calculated confidence level Compare with a preset threshold (e.g., 0.8): If If the value is ≥0.8, it is considered a valid early warning event, a formal alarm command is output, management personnel are notified, and subsequent linkage control is triggered; if If the value is less than 0.8, it is determined to be transient disturbance noise (such as passing ships, bubble interference, etc.), and the warning information is intercepted and only logged, thereby effectively reducing the false alarm rate of the system.

[0131] After completing the early warning determination, the system generates corresponding control commands based on the comparison results (including real-time health index status, valid early warning events, etc.) to dynamically adjust the operating parameters of the front-end monitoring equipment, thereby achieving intelligent closed-loop control of "energy saving in peacetime and tracking in wartime".

[0132] The specific control method is as follows:

[0133] 1. Adaptive sampling frequency adjustment. The system adjusts the sampling frequency based on real-time health indices. and its rate of change The sampling frequency of the monitoring terminal is dynamically adjusted. The adjustment rule is: when the real-time health index is continuously (e.g., 12 times, 1 hour each time) within the preset excellent range (e.g., corresponding to Class I / II water), ≥4) and the fluctuation range is less than the preset threshold (e.g., standard deviation). When the value is <0.1, it indicates that the water quality is stable and excellent. The system automatically reduces the sampling frequency (e.g., from 5 minutes / time to 10 minutes / time) and enters a low-power cruise mode to save reagent consumption, power, and communication bandwidth. When a valid early warning event is detected (i.e., an alarm that passes the confidence assessment) or the rate of change of the real-time health index exceeds a preset threshold (e.g., ...), the system will automatically switch to a low-power cruise mode. When the sampling frequency is >0.5 / hour, it indicates that the water quality may be deteriorating or has been polluted. The system immediately increases the sampling frequency (e.g., adjusts it to 1 minute / time) and enters high-frequency tracking mode to capture the pollution diffusion process with high temporal resolution, providing accurate data support for emergency response.

[0134] 2. Visual Linkage Evidence Collection. When a valid early warning event is triggered, the system determines the image acquisition device (such as a high-definition camera) associated with the current monitoring point and its preset position information based on preset device associations. It then generates and sends PTZ (pan-tilt-zoom) control commands to drive the device to automatically turn to the monitoring point area for zoom capture and video recording. The system also adds "critical event" markers to the collected image data to prevent cyclic overwriting. This mechanism achieves "monitoring as source tracing," providing direct evidence for subsequent law enforcement and pollution source tracing.

[0135] 3. Intelligent cleaning control.

[0136] The system also features sensor self-diagnosis and cleaning functions to eliminate abnormal data caused by probe clogging, biofouling, etc. When the current daily comprehensive index is detected to be lower than the dynamic warning threshold or the confidence level is lower than the preset low threshold (e.g., ...), the system will automatically detect abnormal data. Furthermore, the fluctuation characteristics of real-time health indices If the data exceeds the preset threshold, it indicates that the data is abnormal but may be caused by sensor clogging; or, based on the image data fed back by the image acquisition device, the image recognition algorithm determines that there are no obvious signs of pollution such as oil stains or discoloration on the water surface (i.e., the data is abnormal but the visual appearance is normal), indicating that the sensor may be physically obstructed or damaged.

[0137] The cleaning command triggers the monitoring terminal to activate the wipers, air blower, or ultrasonic cleaning device. After cleaning, a retest is immediately performed. If the retest result is normal, the alarm is canceled; if it is still abnormal, the alarm process is re-entered. This mechanism effectively avoids false alarms caused by improper sensor maintenance and improves the long-term reliability of the system.

[0138] Through the above multi-level comparison, evaluation and control, this step achieves high-confidence screening of early warning information and dynamic allocation of detection resources on demand.

[0139] This embodiment constructs a complete closed loop of "data acquisition-analysis and evaluation-feedback control," which significantly reduces operation and maintenance costs while ensuring monitoring accuracy. It also improves the accuracy of early warning and enables automatic evidence collection of pollution incidents, providing a precise, intelligent, and automated complete technical solution for the long-term supervision of "zero direct discharge zones for wastewater."

[0140] Example 2

[0141] like Figure 6 As shown, a system for assessing and issuing early warnings of river health status is used to implement the method described in Example 1. The system includes:

[0142] The monitoring terminal is set up at the river monitoring point to collect real-time monitoring data, which includes the monitoring values ​​of various water quality indicators.

[0143] Image acquisition equipment is installed in the area where the monitoring point is located to acquire video images;

[0144] An edge computing gateway is communicatively connected to a monitoring terminal and an image acquisition device. The edge computing gateway includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described above.

[0145] Specifically, the monitoring terminal has a communication interface that can receive downlink control commands from the edge computing gateway and dynamically adjust the sampling frequency according to the commands. At the same time, the monitoring terminal can integrate a self-cleaning device (such as a wiper, air blower or ultrasonic cleaner) to automatically perform sensor cleaning operation when a cleaning command is received, and perform retesting after cleaning is completed.

[0146] Image acquisition equipment is typically a high-definition camera with pan-tilt-zoom (PTZ) capability, which can adjust the shooting angle and focal length according to instructions.

[0147] As the local control core, the edge computing gateway is responsible for real-time processing of monitoring data, dynamic early warning generation, early warning judgment, and generating control commands based on the judgment results to send to the monitoring terminal (such as adjusting the sampling frequency and starting cleaning) and image acquisition equipment (such as rotating to a specified preset position and capturing video), thereby realizing adaptive scheduling of monitoring resources and collaborative linkage of multiple devices.

[0148] This system is not an isolated monitoring device, but a complete automated control system. The edge computing gateway, as the core control unit, establishes a dynamic mapping relationship between sampling frequency and real-time health indices and their rate of change by executing the method described in Example 1. It can adaptively adjust the monitoring terminal's operating mode according to water quality conditions: automatically switching to a low-power cruise mode when water quality is excellent and stable, significantly reducing the sampling frequency to save reagent consumption and communication bandwidth; and immediately switching to a high-frequency tracking mode when a pollution event or sudden change in water quality is detected, ensuring data accuracy during critical periods. Simultaneously, the system achieves automated evidence retention for "monitoring as tracing" through the coordinated control of image acquisition devices. Through this collaborative work, the system significantly reduces operation and maintenance costs while ensuring monitoring accuracy, effectively solving the cost problem of high-frequency monitoring and achieving the intelligent regulatory goal of "energy saving and consumption reduction in peacetime, and rapid tracking in wartime."

[0149] Example 3

[0150] A computer-readable storage medium storing a computer program that, when executed by a computer, enables a method for assessing and providing early warning of river health status as described above.

[0151] For example, a computer program can be divided into one or more modules / units, one or more modules / units are stored in memory and executed by a processor, and data I / O interface transmission is completed by input interface and output interface to complete the present invention. One or more modules / units can be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution process of the computer program in a computer device.

[0152] Computer devices can be desktop computers, laptops, handheld computers, and cloud servers, etc. Computer devices may include, but are not limited to, memory and processors. Those skilled in the art will understand that this embodiment is merely an example of a computer device and does not constitute a limitation on the computer device. It may include more or fewer components, or combine certain components, or different components. For example, a computer device may also include an input device, network access device, bus, etc.

[0153] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0154] Memory can be an internal storage unit of a computer device, such as a hard drive or RAM. Memory can also be an external storage device of a computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory can include both internal and external storage units of a computer device. Memory is used to store computer programs and other programs and data required by the computer device. Memory can also be used for temporary storage on output devices. The aforementioned storage media include various media that can store program code, such as USB flash drives, portable hard drives, Read-Only Memory (ROM), Random Access Memory (RAM), discs, or optical discs.

[0155] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for assessing and issuing early warnings of river health status, characterized in that, Includes the following steps: Acquire real-time monitoring data from river monitoring points, including monitoring values ​​of various water quality indicators; The final water quality level at the current moment is assessed based on the real-time monitoring data, and the daily comprehensive index at the current moment is calculated. Obtain the historical baseline daily comprehensive index based on historical monitoring data, and calculate the dynamic early warning threshold based on the historical baseline daily comprehensive index and the final water quality level at the current moment; The daily comprehensive index at the current moment is compared with the dynamic early warning threshold. When the daily comprehensive index at the current moment is lower than the dynamic early warning threshold, an early warning message is generated, and a control command is generated based on the comparison result to adjust the operating parameters of the front-end monitoring equipment.

2. The method for assessing and issuing early warning of river health status according to claim 1, characterized in that, The step of assessing the final water quality level at the current moment based on the real-time monitoring data includes: The original water quality level at the current moment is obtained by comparing the real-time monitoring data with the water quality standards; Obtain the final water quality level at the previous moment, and calculate the change in water quality level between the original water quality level and the final water quality level at the previous moment; The original water quality level, the final water quality level at the previous moment, and the change in level are weighted and combined to obtain the final water quality level at the current moment. The filter weight coefficients of the weighted combination are dynamically adjusted based on the river flow velocity information.

3. The method for assessing and issuing early warning of river health status according to claim 1, characterized in that, The calculation of the daily composite index at the current moment includes: Calculate the real-time health index based on the real-time monitoring data. The daily comprehensive index for the current moment is obtained by performing a time-decay weighted average of the real-time health index at each moment within a preset time period.

4. The method for assessing and issuing early warning of river health status according to claim 3, characterized in that, The calculation of the real-time health index based on the real-time monitoring data includes: Each monitoring value is converted into a dimensionless single health sub-index, with different conversion rules used for indicators with different physical characteristics; Each individual health sub-index is positively oriented, and the information entropy of each index is calculated based on the positively oriented values. The dynamic weights of each individual indicator are determined based on the information entropy, and the individual health sub-indices are weighted and summed using the dynamic weights to obtain the real-time health index at the current moment.

5. The method for assessing and issuing early warning of river health status according to claim 1, characterized in that, The calculation of the dynamic early warning threshold based on the historical baseline daily comprehensive index and the final water quality level at the current moment includes: The tolerance coefficient for the water quality level is determined based on the final water quality level at the current moment, and the environmental correction coefficient is determined based on the environmental background parameters. Calculate the product of the tolerance coefficient and the environmental correction coefficient and subtract it from 1. Multiply the historical baseline daily comprehensive index by the difference to obtain the dynamic early warning threshold.

6. The method for assessing and issuing early warning of river health status according to claim 1, characterized in that, After generating the warning information, it also includes: Extract the real-time health index from multiple monitoring points before the triggering of the early warning and calculate its fluctuation characteristics; The confidence level of this triggering event is calculated based on the fluctuation characteristics and the historical benchmark daily composite index; An alarm command is output when the confidence level is greater than or equal to a preset threshold; otherwise, the warning information is blocked.

7. The method for assessing and issuing early warning of river health status according to claim 1, characterized in that, The step of generating control commands based on the comparison results to adjust the operating parameters of the front-end monitoring equipment includes: When the real-time health index is in the excellent range and the fluctuation range is less than the preset threshold, the sampling frequency of the monitoring terminal is reduced; when a valid early warning event is detected or the rate of change of the real-time health index exceeds the preset threshold, the sampling frequency of the monitoring terminal is increased.

8. The method for assessing and issuing early warning of river health status according to claim 1, characterized in that, The step of generating control commands based on the comparison results to adjust the operating parameters of the front-end monitoring equipment also includes at least one of the following: When a valid early warning event is triggered, the image acquisition device associated with the current monitoring point is determined according to the preset device association relationship, and a control command is generated to drive the image acquisition device to acquire images of the area where the monitoring point is located, and the acquired image data is marked and stored. When the daily comprehensive index at the current moment is detected to be lower than the dynamic warning threshold, and the confidence level is less than the preset threshold and the fluctuation characteristics exceed the preset threshold, a sensor self-cleaning command is generated, and a retest is performed after the cleaning is completed; or when it is determined from the image data fed back by the image acquisition device that there is no obvious abnormality on the water surface, a sensor self-cleaning command is generated, and a retest is performed after the cleaning is completed.

9. A system for assessing and issuing early warnings of river health status, characterized in that, include: The monitoring terminal is set at the river monitoring point to collect real-time monitoring data, which includes the monitoring values ​​of various water quality indicators. An image acquisition device is installed in the area where the monitoring point is located to acquire video images; An edge computing gateway is communicatively connected to the monitoring terminal and the image acquisition device, respectively. The edge computing gateway includes a memory and a processor. The memory stores a computer program, and the processor is used to execute the computer program to implement the method as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that, The computer program enables the computer to implement a method for assessing and warning the health status of a river as described in any one of claims 1 to 8 when executed.

Images