Water pollution event index screening method and device, electronic equipment and storage medium

Abstract

The application discloses a water pollution event index screening method and device, electronic equipment and a storage medium. The method comprises collecting historical monitoring data sets and real-time monitoring flow data of a water body and preprocessing the data; calculating initial entropy weights of each monitoring index, and screening out a candidate index set with high sensitivity according to a preset threshold; for pollutants in the candidate index set, fusing acute toxicity and biological enrichment risk thereof, constructing a toxicity coefficient and performing normalization processing, and dividing a toxicity grade based on the toxicity coefficient; measuring the correlation strength of the initial entropy weight ordering and the normalized toxicity weight ordering, combining a benchmark toxicity contribution threshold based on a risk threshold and a water body function correction factor calculated based on a water body national standard limit value, and calculating adaptive comprehensive weights of each candidate index; and verifying the decision coefficient of the comprehensive weight and the actual risk entropy to screen out final early warning perception indexes. The application is free of experience parameters, the results are reproducible, and is suitable for various water bodies.

Description

Technical Field

[0001] This invention belongs to the field of water pollution early warning and emergency response technology, specifically involving water pollution event indicator screening methods, devices, electronic equipment, and storage media. Background Technology

[0002] Sudden water pollution incidents are characterized by their sporadic occurrence, high risk, and wide-ranging impact, such as chemical leaks and hazardous chemical transportation accidents. Currently, there are over 40,000 known compounds globally. A limited number of early warning indicators cannot accurately reflect the risk, while a large number of indicators would result in excessively high early warning costs. Therefore, the selection of early warning indicators is crucial and also a major challenge in early warning work. Current mainstream early warning indicator selection technologies mainly revolve around two core approaches: "data volatility" and "pollution risk level," but neither can adequately meet the emergency early warning needs under complex pollution scenarios.

[0003] In existing technologies, traditional methods, such as the entropy weight method and the coefficient of variation method, rely solely on the fluctuation of pollutant concentration data for weighting, which easily misses highly toxic low-concentration pollutants (such as heavy metals like mercury and cadmium, or highly toxic organic compounds like phenol and cyanide, whose half-effect concentrations are low). (Often below 0.1 mg / L, but with small concentration fluctuations in the initial stage of a sudden leak), it is easily identified as a "low-sensitivity indicator." On the other hand, weights calculated based on historical data cannot reflect the trend of pollutant migration and transformation, making it difficult to respond to spatiotemporal changes in concentration. Some techniques that attempt to combine toxicity screening indicators, such as the method of classifying limits based on the "Surface Water Environmental Quality Standard" (GB3838-2002), suffer from the problems of strong subjectivity in toxicity quantification and lack of correlation with monitoring data characteristics. The former often relies on empirical coefficients to set toxicity weights or adopts simplified classifications without clear quantitative standards; the latter only focuses on the inherent risk of pollutants and does not combine real-time monitoring sensitivity (e.g., when the concentration of highly toxic pollutants is stable for a long time and far below the limit, the warning priority should be lower than that of "medium-low toxic pollutants with a sudden increase in concentration"), which can easily lead to "over-warning" or "unfocused warnings" and increase emergency response costs.

[0004] In summary, current technologies generally suffer from technical problems such as reliance on empirical parameters, inability of weights to reflect the pollution process, insufficient risk-sensitivity coordination, and separation of the relationship between data fluctuations and toxicity risks. Summary of the Invention

[0005] The present application aims at the problems of non-objective toxicity quantification, linear limitation of correlation description, poor scene adaptability and the like in existing water pollution early warning, and discloses a sudden water pollution event early warning perception index screening method and device based on entropy weight-toxicity fusion, an electronic device and a storage medium. The present application realizes the fully objective screening of water pollution early warning perception indexes by introducing a sliding window dynamic standardization, an entropy weight method preliminary screening, an acute toxicity-bioaccumulation risk two-dimensional toxicity quantification, and a weight fusion mechanism based on a Kendall tau coefficient, a benchmark toxicity contribution threshold and a water body function correction factor.

[0006] To achieve the above-mentioned purpose, a water pollution event index screening method is provided in a specific embodiment of the present application, which comprises:

[0007] Collecting historical monitoring data sets and real-time monitoring flow data of a water body and preprocessing the data;

[0008] Based on the preprocessed data, calculating the initial entropy weight of each monitoring index, and screening out a candidate index set with high sensitivity according to a preset threshold;

[0009] For the pollutants in the candidate index set, fusing the acute toxicity and the bioaccumulation risk thereof, constructing a toxicity coefficient and performing normalization processing, and dividing the toxicity grade based on the toxicity coefficient;

[0010] Measuring the correlation strength of the initial entropy weight ordering and the normalized toxicity weight ordering, combining a benchmark toxicity contribution threshold based on a risk threshold and a water body function correction factor calculated based on a water body national standard limit value, calculating the adaptive comprehensive weight of each candidate index, calculating the actual risk entropy based on the real-time monitoring data according to the adaptive comprehensive weight, and verifying the determination coefficient of the comprehensive weight and the actual risk entropy to screen out the final early warning perception index.

[0011] In one or more embodiments of the present application, the historical monitoring data set is , including samples, indexes, and the real-time monitoring flow data is ; the preprocessing of the data comprises adopting sliding window dynamic standardization to dynamically capture the instantaneous peak value of the sudden pollutants, converting the value of each index to interval to obtain the standardized data value .

[0012] In one or more embodiments of the present application, the screening step of the high-sensitivity candidate index set comprises:

[0013] Calculating the standardized ratio of each index based on the preprocessed data;

[0014] The entropy value is calculated using the entropy weight method based on the standardized ratio. Entropy The smaller the value, the higher the data dispersion of this indicator, and the greater its information contribution.

[0015] According to entropy value Calculate the initial entropy weight of this index A candidate indicator set is generated by using the sensitivity weights to characterize the indicator. The candidate index set Initial entropy weight of the corresponding index All are greater than the preset threshold.

[0016] In one or more embodiments of the present invention, the method further includes:

[0017] Based on half-effect concentration With bioconcentrated factors Constructing toxicity coefficients , , For the first The half-effect concentration of each pollutant on aquatic organisms is expressed in mg / L. For the first Bioconcentrating factors of pollutants;

[0018] Based on candidate indicator set Toxicity coefficient Normalization is performed to obtain the value in the interval. Standardized toxicity coefficient within the range;

[0019] Toxicity levels are determined based on standardized toxicity coefficients and preset standards.

[0020] In one or more embodiments of the present invention, the calculation step of the adaptive comprehensive weight of each candidate index includes:

[0021] Define the core parameters in the adaptive fusion and comprehensive weight calculation process, including at least the Kendalltau coefficient. Benchmark toxicity contribution threshold and water body function correction factors ;

[0022] The adaptive comprehensive weights are calculated based on the above core parameters, including:

[0023]

[0024] in, Contribute to entropy weight. The larger the size, the higher the proportion; Contributing to toxicity, The smaller the percentage, the higher the proportion; the two ensure that the correlation dominates the allocation. Ensure that the entropy weight contribution is not less than The weight of extremely toxic indicators is not suppressed; The weighting is adjusted according to the differences in national standard limits to suit the risk requirements of different water bodies.

[0025] In one or more embodiments of the present invention, the Kendalltau coefficient ,in: (The number of consistent pairs is the same as the number of logarithmic pairs where the entropy weight ranking of the two indicators is the same as the toxicity ranking). (This refers to the inconsistency of the number of pairs, i.e., the logarithm of the entropy-weighted ranking of the two indicators versus the logarithmic ranking). =Candidate Set The number of indicators;

[0026] The benchmark toxicity contribution threshold Among them, the benchmark risk entropy =0.1 , The maximum normalized toxicity coefficient for the candidate set;

[0027] The water body function correction factor ,in: Reference water body limits, Target water body limits.

[0028] In one or more embodiments of the present invention, the verification step of the determination coefficient of the comprehensive weight and the actual risk entropy includes:

[0029] Calculate the actual risk entropy based on real-time monitoring data. And verify the weights and coefficient of determination The condition is considered acceptable; the actual risk entropy calculation formula is as follows:

[0030]

[0031] in, To monitor concentration in real time, data is taken from real-time monitoring streams. ;

[0032] The coefficient of determination The calculation method is as follows:

[0033]

[0034] in, , is the weighted risk entropy prediction value. , where is the average actual risk entropy.

[0035] In another aspect of the present application, a water pollution event index screening device is provided, comprising a collection unit, a preliminary screening unit, a toxicity division unit, and a calculation and verification unit, wherein,

[0036] The collection unit is configured to collect historical monitoring data sets and real-time monitoring flow data of a water body and pre-process the data;

[0037] The preliminary screening unit is configured to calculate initial entropy weights of each monitoring index based on the pre-processed data, and screen out a candidate index set with high sensitivity according to a preset threshold;

[0038] The toxicity division unit is configured to fuse acute toxicity and biological enrichment risk of pollutants in the candidate index set, construct a toxicity coefficient and perform normalization processing, and divide toxicity grades based on the toxicity coefficient;

[0039] The calculation and verification unit is configured to measure the correlation strength of the initial entropy weight ranking and the normalized toxicity weight ranking, combine a benchmark toxicity contribution threshold based on a risk threshold and a water body function correction factor calculated based on a water body national standard limit value, calculate adaptive comprehensive weights of each candidate index, calculate actual risk entropy based on real-time monitoring data according to the adaptive comprehensive weights, and verify the determination coefficient of the comprehensive weights and the actual risk entropy, so as to screen out final early warning awareness indexes.

[0040] In another aspect of the present application, an electronic device is provided, comprising at least one processor, and a memory storing instructions which, when executed by the at least one processor, cause the at least one processor to perform a water pollution event index screening method.

[0041] In another aspect of the present application, a computer readable storage medium is provided, having a computer program stored thereon, which, when executed by a processor, performs the steps of a water pollution event index screening method.

[0042] Advantages: Compared with the prior art, the present application objectively evaluates data fluctuation sensitivity by entropy weight method, objectively quantifies pollutant toxicity by fusing EC50 and BCF of authoritative databases, and introduces parameters such as Kendall tau coefficient, benchmark toxicity contribution threshold, and water body function correction factor, without setting empirical parameters throughout the process, ensuring the reproducibility and objectivity of the screening results, and effectively avoiding the subjective defects of traditional methods;

[0043] This invention creatively integrates data sensitivity (entropy weight) with inherent toxicity risk (toxicity quantification) and uses the Kendalltau coefficient to characterize the nonlinear correlation strength, effectively solving the problems of traditional methods that miss highly toxic low-concentration pollutants or over-warn, thus ensuring the accuracy and relevance of warnings.

[0044] This invention uses a weighting rationality verification mechanism to ensure that the selected indicators can effectively reflect actual pollution risks. R² (≥0.7), reducing the false alarm rate. Simultaneously, this method can directly interface with IoT sensors, enabling rapid and accurate early warning of sudden water pollution events in complex aquatic environments;

[0045] This invention introduces a water body function correction factor and clarifies the limit calculation logic based on the "Surface Water Environmental Quality Standard" (GB3838-2002), "Drinking Water Source Water Quality Standard" (CJ3020-1993), and "Integrated Wastewater Discharge Standard" (GB8978-1996). This enables the method to adaptively adjust the index weights according to the functional categories and environmental standards of different water bodies, thereby achieving precise early warning. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 recorded in the present invention. 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 water pollution event indicator screening method according to an embodiment of the present invention;

[0048] Figure 2 This is a block diagram of a water pollution event indicator screening device according to one embodiment of the present invention;

[0049] Figure 3 This is a hardware structure diagram of a computing device according to an embodiment of the present invention. Detailed Implementation

[0050] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0051] As described in the background section, existing water pollution early warning systems suffer from problems such as subjective toxicity quantification, limited linear correlation characterization, and poor scenario adaptability.

[0052] In response to the above technical problems, such as Figure 1 As shown, this invention introduces a method for screening water pollution event indicators, including the following steps:

[0053] Step S1: Collect historical monitoring datasets and real-time monitoring flow data of the water body and preprocess the data;

[0054] Step S2: Based on the preprocessed data, calculate the initial entropy weight of each monitoring indicator, and select a set of highly sensitive candidate indicators according to a preset threshold.

[0055] Step S3: For pollutants in the candidate index set, integrate their acute toxicity and bioaccumulation risk, construct toxicity coefficients and normalize them, and classify toxicity levels based on the toxicity coefficients.

[0056] Step S4: Measure the correlation strength between the initial entropy weight ranking and the normalized toxicity weight ranking, combine the baseline toxicity contribution threshold based on the risk threshold and the water body function correction factor calculated based on the national water body standard limit, and calculate the adaptive comprehensive weight of each candidate indicator; calculate the actual risk entropy based on the adaptive comprehensive weight and real-time monitoring data, and verify the determination coefficient between the comprehensive weight and the actual risk entropy to select the final early warning perception indicators.

[0057] In a further embodiment, step S1 includes:

[0058] Data on various water quality indicators are collected from sources such as automatic water quality monitoring stations, manual sampling points, and historical databases. These include, but are not limited to, pH, dissolved oxygen, conductivity, turbidity, chemical oxygen demand (COD), ammonia nitrogen, total phosphorus, heavy metal ions (such as mercury, cadmium, and arsenic), and various organic pollutants (such as benzene compounds, phenol, and cyanide). Both types of data include historical monitoring datasets. Covering the wet / dry / normal water periods, a total of One sample, One indicator; real-time monitoring of streaming data Based on water quality sensor data, it reflects the dynamic changes in pollutant concentration;

[0059] The collected data undergoes preprocessing, including outlier removal, missing value imputation (e.g., using linear interpolation, regression interpolation, or K-nearest neighbor imputation), and data format standardization to ensure data quality. Traditional min-max standardization, based on global extrema, easily smooths local pollution peaks but fails to reflect abnormal fluctuations of sudden pollutants within local time windows. This invention employs sliding window dynamic standardization to dynamically capture the instantaneous peak values ​​of sudden pollutants, transforming the value of each indicator to... The interval, the formula is as follows:

[0060]

[0061] No. Each indicator within the time window The moving average over a period of time (e.g., 30 minutes) reflects short-term data trends, and is calculated using the following formula:

[0062]

[0063] The standard deviation of the corresponding window characterizes the degree of data dispersion. To prevent extremely small constants with a denominator of zero,

[0064]

[0065] These are standardized data values. Through this dynamic standardization, the values ​​of all indicators are compressed to between 0 and 1, better reflecting the real-time fluctuations and trends of water quality indicators, eliminating the influence of dimensions, and making subsequent calculations more comparable.

[0066] In a further embodiment, step S2 mainly employs the entropy weight method for initial screening of highly sensitive indicators, including:

[0067] Traditionally based on the standardized data from step 1 The "information utility value" of each indicator is calculated using the entropy weight method, and highly sensitive indicators (those with significant data fluctuations and large information contributions) are screened. The specific steps are as follows.

[0068] In a further embodiment, the calculation of the index weighting is included, the first The first indicator in the The standardized proportion of a sample reflects its contribution to the overall distribution of the indicator, and the formula is:

[0069]

[0070] In a further embodiment, the calculation includes an information entropy value, based on information theory principles, where entropy is... The smaller the value, the higher the dispersion of the indicator data (the more significant the fluctuation), and the greater the information contribution. The formula is:

[0071]

[0072] In a further embodiment, the calculation of entropy weights and initial entropy weights are included. The sensitivity weights of the characterization indicators are expressed by the following formula:

[0073]

[0074] reserve Indicators ( ),when At that time, a candidate indicator set is generated. .

[0075] In a further embodiment, step S3 is primarily used for objective toxicity quantification, based on the candidate set. By integrating the two dimensions of "acute toxicity and bioaccumulation risk," the inherent toxicity of pollutants can be quantified to avoid missing highly toxic substances. The specific steps are as follows:

[0076] In a further embodiment, a toxicity factor is defined using a half-effect concentration. With bioconcentrated factors Constructing toxicity coefficients , The smaller the value, the lower the acute toxicity. The larger the size, the higher the risk of bioaccumulation, as shown in the formula:

[0077]

[0078] in, For the first The half-effect concentration (mg / L) of each pollutant on aquatic organisms is derived from the EPAECOTOX database; For the first The bioconcentration factor (dimensionless) of each pollutant is derived from the EPAEPISuite model.

[0079] In a further embodiment, this includes normalizing the toxicity coefficient: to ensure that the toxicity coefficient is of the same order of magnitude as the entropy weight (both are within the range of...). The interval is normalized using the following formula:

[0080]

[0081] The maximum toxicity coefficient within the candidate set is used, ensuring that the maximum value after normalization is 1 and the minimum value is ≥0.

[0082] In a further embodiment, toxicity level classification is included: based on normalization results, and in conjunction with the "Surface Water Environmental Quality Standard" (GB3838-2002) and the "Technical Guidelines for Environmental Risk Assessment" (HJ2.3-2018), toxicity levels are classified to provide a basis for subsequent safety net mechanisms: Extremely High Toxicity: (e.g., mercury, cadmium); highly toxic: (e.g., phenol, cyanide); poisoning: (e.g., ammonia nitrogen, total arsenic); low toxicity: (such as total phosphorus and COD).

[0083] In a further embodiment, step S4 includes adaptive fusion and comprehensive weight calculation. In this step, the Kendalltau coefficient (non-linear correlation measure), the baseline toxicity contribution threshold, and the water body function correction factor (scenario adaptation) are used to achieve a fully objective fusion of entropy weight and toxicity. The water body function correction factor is specifically calculated based on national standard limits, and the parameter values ​​are determined as follows:

[0084] In a further embodiment, the core factors are defined: the physical meaning and mathematical definition of each core parameter in the adaptive fusion and comprehensive weight calculation process are defined, as shown in the table below:

[0085] Table 1. Definition of core parameters for adaptive fusion and comprehensive weight calculation process.

[0086]

[0087] In a further embodiment, the Kendalltau coefficient is calculated. The steps include processing the candidate set All internal indicators are calculated according to the following steps. Entropy weights Sort to obtain the ranking sequence (like ,but =1, =2); for normalized toxicity coefficient Sort to obtain the ranking sequence ; Statistical analysis of all indicators Consistent against Inconsistency ,like In order to be consistent with ;like For inconsistency Substitute into the formula to calculate. :

[0088]

[0089] Range of values , When the value approaches 1, the proportion of "highly sensitive and highly toxic" indicators is high (data fluctuations are consistent with toxicity). When the value approaches -1, the proportion of "high sensitivity - low toxicity" / "low sensitivity - high toxicity" indicators is high (data fluctuations deviate from toxicity).

[0090] In a further embodiment, the baseline toxicity contribution threshold is calculated by substituting each indicator in the candidate set into the formula:

[0091]

[0092] In a further embodiment, calculating the water body function correction factor includes: determining the GB3838-2002 category corresponding to the target water body based on the actual application scenario, and obtaining the "benchmark water body limit" of the target indicator from GB3838-2002. Class III water) and target water body limit ( (corresponding category), substitute into the formula:

[0093]

[0094] like (If the requirements for the target water body are more stringent), then (Weight amplification); if ,but (No magnification); if (If the target water body requirements are lenient), then (Weight reduction, applicable only to non-sensitive water bodies). If the target water body is an "emergency backup reservoir for drinking water sources," the "Drinking Water Source Quality Standard" (CJ3020-1993) must be additionally referenced. When the limit in CJ3020-1993 is stricter than the corresponding category in GB3838-2002, the limit in CJ3020-1993 shall be used as the standard. Prioritize ensuring drinking water safety; for characteristic pollutants not explicitly specified in GB3838-2002 (such as specific industrial wastewater pollutants). The Class I discharge standard of the "Integrated Wastewater Discharge Standard" (GB8978-1996) shall be adopted. The risk control threshold of the target water body is selected to ensure the complete objectivity of the factor calculation.

[0095] In a further embodiment, calculating the adaptive synthesis weights includes:

[0096] Integrating the above factors, the comprehensive weight is calculated according to the logic of "association allocation, minimum toxicity, and scenario adaptation", and the formula is as follows:

[0097]

[0098] Contribute to entropy weight. The larger the size, the higher the proportion; Contributing to toxicity, The smaller the percentage, the higher the proportion; the two ensure that the correlation dominates the allocation. Ensure that the entropy weight contribution is not less than The weight of extremely toxic indicators is not suppressed; The weighting is adjusted according to the differences in national standard limits to suit the risk requirements of different water bodies.

[0099] In a further embodiment, the rationality of the weights needs to be verified. To ensure that the overall weights are consistent with the actual pollution risk, the actual risk entropy is calculated based on real-time monitoring data. And verify the weights and coefficient of determination ( To be considered qualified, the weights can effectively reflect the actual risks, where the formula for calculating the actual risk entropy is:

[0100]

[0101] in, To monitor concentration in real time, data was taken from... ;

[0102] The coefficient of determination is calculated as follows:

[0103]

[0104] in, , is the weighted risk entropy prediction value. , where is the average actual risk entropy. When the coefficient of determination When the time (or other preset qualified threshold) is reached, it indicates that the calculated comprehensive weight can effectively reflect the actual pollution risk, and this indicator is selected as the final early warning perception indicator.

[0105] Based on the real-time monitoring data of the selected early warning indicators, combined with preset early warning models and thresholds, water pollution early warning information is generated. For example, when the real-time monitoring value of any early warning indicator exceeds its health risk standard, environmental quality standard, or specific early warning threshold, different levels of early warning (such as blue, yellow, orange, and red warnings) are triggered. The early warning information may include the type of pollution, possible sources, affected areas, and recommended emergency measures, and is sent to relevant managers and departments through audible and visual alarms, SMS, and app push notifications.

[0106] like Figure 2As shown, an apparatus for implementing a water pollution incident indicator screening method according to a specific embodiment of the present invention is described. The apparatus includes: a data acquisition unit 201, a preliminary screening unit 202, a toxicity classification unit 203, and a calculation and verification unit 204.

[0107] The acquisition unit 201 is used to acquire historical monitoring datasets and real-time monitoring flow data of the water body and to preprocess the data;

[0108] The preliminary screening unit 202 is used to calculate the initial entropy weight of each monitoring indicator based on the preprocessed data, and to screen out a set of highly sensitive candidate indicators according to a preset threshold.

[0109] The toxicity classification unit 203 is used to construct a toxicity coefficient and perform normalization processing on pollutants in the candidate index set by integrating their acute toxicity and bioaccumulation risk, and classify the toxicity level based on the toxicity coefficient.

[0110] The calculation and verification unit 204 is used to measure the correlation strength between the initial entropy weight ranking and the normalized toxicity weight ranking, and to calculate the adaptive comprehensive weight of each candidate indicator by combining the benchmark toxicity contribution threshold based on the risk threshold and the water body function correction factor calculated based on the national standard limit for water bodies; based on the adaptive comprehensive weight, the actual risk entropy is calculated based on real-time monitoring data, and the determination coefficient between the comprehensive weight and the actual risk entropy is verified, so as to select the final early warning perception indicators.

[0111] Figure 3 The hardware structure diagram of the electronic device 30 for the water pollution incident indicator screening method is shown according to an embodiment of this specification. Figure 3 As shown, the electronic device 30 may include at least one processor 301, a memory 302 (e.g., non-volatile memory), a RAM 303, and a communication interface 304, and the at least one processor 301, memory 302, RAM 303, and communication interface 304 are connected together via a bus 305. At least one processor 301 executes at least one computer-readable instruction stored or encoded in the memory 302.

[0112] It should be understood that the computer-executable instructions stored in memory 302, when executed, cause at least one processor 301 to perform the above-described combinations in the various embodiments of this specification. Figure 2 The description includes various operations and functions.

[0113] In the embodiments of this specification, electronic device 30 may include, but is not limited to: personal computer, server computer, workstation, desktop computer, laptop computer, notebook computer, mobile computing device, smartphone, tablet computer, cellular phone, personal digital assistant (PDA), handheld system, messaging device, wearable computing device, consumer electronic device, etc.

[0114] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0115] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will 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 program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A system that specifies functions in one or more boxes.

[0116] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction set implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0117] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0118] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0119] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for screening water pollution event indicators, characterized in that, The method includes: Historical monitoring datasets and real-time monitoring flow data of the water body were collected and preprocessed; the historical monitoring dataset is... ,include One sample, The indicators, the real-time monitoring stream data are The data preprocessing includes using a sliding window for dynamic normalization to dynamically capture the instantaneous peak values ​​of sudden pollutants and converting the values ​​of each indicator to... The interval is obtained after standardization of data values ; Based on the preprocessed data, the initial entropy weight of each monitoring indicator is calculated, and a set of highly sensitive candidate indicators is selected according to a preset threshold; the selection steps for the set of highly sensitive candidate indicators include: The standardized ratios of each indicator are calculated based on the preprocessed data. ; The entropy value is calculated using the entropy weight method based on the standardized ratio. Entropy The smaller the value, the higher the data dispersion of this indicator, and the greater its information contribution. According to entropy value Calculate the initial entropy weight of this index A candidate indicator set is generated by using the sensitivity weights to characterize the indicator. The candidate index set Initial entropy weight of the corresponding index All are greater than the preset threshold; For pollutants in the candidate index set, their acute toxicity and bioaccumulation risk are integrated to construct a toxicity coefficient, which is then normalized. Toxicity levels are then classified based on the toxicity coefficient. The method further includes: Based on half-effect concentration With bioconcentrated factors Constructing toxicity coefficients , , For the first The half-effect concentration of each pollutant on aquatic organisms is expressed in mg / L. For the first Bioconcentrating factors of pollutants; Based on candidate indicator set Toxicity coefficient Normalization is performed to obtain the value in the interval. Standardized toxicity coefficient within the range; Toxicity levels are classified based on standardized toxicity coefficients and preset standards; The correlation strength between the initial entropy weight ranking and the normalized toxicity weight ranking is measured. Combined with the baseline toxicity contribution threshold based on the risk threshold and the water body function correction factor calculated based on the national water body standard limit, the adaptive comprehensive weight of each candidate indicator is calculated. Based on the adaptive comprehensive weight, the actual risk entropy is calculated based on real-time monitoring data, and the determination coefficient between the comprehensive weight and the actual risk entropy is verified to select the final early warning perception indicators. The calculation steps for the adaptive comprehensive weights of each candidate index include: Define the core parameters in the adaptive fusion and comprehensive weight calculation process, including at least the Kendalltau coefficient. Benchmark toxicity contribution threshold and water body function correction factors ; The adaptive comprehensive weights are calculated based on the above core parameters, including: in, Contribute to entropy weight. The larger the size, the higher the proportion; Contributing to toxicity, The smaller the percentage, the higher the proportion; the two ensure that the correlation dominates the allocation. Ensure that the entropy weight contribution is not less than The weight of extremely toxic indicators is not suppressed; The Kendalltau coefficient is adjusted by amplifying / reducing the weighting based on differences in national standard limits to suit the risk requirements of different water bodies; ,in: (The number of consistent pairs is the same as the number of logarithmic pairs where the entropy weight ranking of the two indicators is the same as the toxicity ranking). The inconsistency is the logarithm of the entropy weight ranking of the two indicators, which is the opposite of the toxicity ranking. =Candidate Set The number of indicators; The benchmark toxicity contribution threshold Among them, the benchmark risk entropy =0.1, The maximum normalized toxicity coefficient for the candidate set; The water body function correction factor ,in: Reference water body limits, Target water body limits.

2. The method for screening water pollution event indicators according to claim 1, characterized in that, The verification steps for the determination coefficients of the comprehensive weights and the actual risk entropy include: Calculate the actual risk entropy based on real-time monitoring data. And verify the weights and coefficient of determination The condition is considered acceptable; the actual risk entropy calculation formula is as follows: in, To monitor concentration in real time, data is taken from real-time monitoring streams. ; The coefficient of determination The calculation method is as follows: in, , is the weighted risk entropy prediction value. , where is the average actual risk entropy.

3. A water pollution incident indicator screening device, characterized in that, It includes a data collection unit, a preliminary screening unit, a toxicity classification unit, and a calculation and verification unit. The acquisition unit is used to collect historical monitoring datasets and real-time monitoring flow data of the water body and to preprocess the data; the historical monitoring dataset is... ,include One sample, The indicators, the real-time monitoring stream data are The data preprocessing includes using a sliding window for dynamic normalization to dynamically capture the instantaneous peak values ​​of sudden pollutants and converting the values ​​of each indicator to... The interval is obtained after standardization of data values ; The preliminary screening unit is used to calculate the initial entropy weight of each monitoring indicator based on the preprocessed data, and to screen out a set of highly sensitive candidate indicators according to a preset threshold; the screening steps for the set of highly sensitive candidate indicators include: The standardized ratios of each indicator are calculated based on the preprocessed data. ; The entropy value is calculated using the entropy weight method based on the standardized ratio. Entropy The smaller the value, the higher the data dispersion of this indicator, and the greater its information contribution. According to entropy value Calculate the initial entropy weight of this index A candidate indicator set is generated by using the sensitivity weights to characterize the indicator. The candidate index set Initial entropy weight of the corresponding index All are greater than the preset threshold; The toxicity classification unit is used to construct toxicity coefficients and normalize them for pollutants in the candidate indicator set by integrating their acute toxicity and bioaccumulation risk, and to classify toxicity levels based on the toxicity coefficients. It also includes: Based on half-effect concentration With bioconcentrated factors Constructing toxicity coefficients , , For the first The half-effect concentration of each pollutant on aquatic organisms is expressed in mg / L. For the first Bioconcentrating factors of pollutants; Based on candidate indicator set Toxicity coefficient Normalization is performed to obtain the value in the interval. Standardized toxicity coefficient within the range; Toxicity levels are classified based on standardized toxicity coefficients and preset standards; The calculation and verification unit is used to measure the correlation strength between the initial entropy weight ranking and the normalized toxicity weight ranking, and to calculate the adaptive comprehensive weight of each candidate indicator by combining the baseline toxicity contribution threshold based on the risk threshold and the water body function correction factor calculated based on the national standard limit for water bodies; based on the adaptive comprehensive weight, the actual risk entropy is calculated based on real-time monitoring data, and the determination coefficient between the comprehensive weight and the actual risk entropy is verified to select the final early warning perception indicators; the calculation steps of the adaptive comprehensive weight of each candidate indicator include: Define the core parameters in the adaptive fusion and comprehensive weight calculation process, including at least the Kendalltau coefficient. Benchmark toxicity contribution threshold and water body function correction factors ; The adaptive comprehensive weights are calculated based on the above core parameters, including: in, Contribute to entropy weight. The larger the size, the higher the proportion; Contributing to toxicity, The smaller the percentage, the higher the proportion; the two ensure that the correlation dominates the allocation. Ensure that the entropy weight contribution is not less than The weight of extremely toxic indicators is not suppressed; The Kendalltau coefficient is adjusted by amplifying / reducing the weighting based on differences in national standard limits to suit the risk requirements of different water bodies; ,in: (The number of consistent pairs is the same as the number of logarithmic pairs where the entropy weight ranking of the two indicators is the same as the toxicity ranking). The inconsistency is the logarithm of the entropy weight ranking of the two indicators, which is the opposite of the toxicity ranking. =Candidate Set The number of indicators; The benchmark toxicity contribution threshold Among them, the benchmark risk entropy =0.1, The maximum normalized toxicity coefficient for the candidate set; The water body function correction factor ,in: Reference water body limits, Target water body limits.

4. An electronic device, characterized in that, include: The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the water pollution event indicator screening method as described in any one of claims 1-2.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the water pollution incident indicator screening method as described in any one of claims 1-2.

Images