Rapid tracing method and system for tracing groundwater pollution sources

Abstract

The present invention relates to the technical field of water pollution detection, and in particular to a rapid tracing method and system for tracing groundwater pollution sources. The present invention constructs a preliminary prediction model of pollutant concentrations at different sampling points based on the pollutant concentration sequence of samples at different sampling points; based on the water flow velocity sequence, analyzes the abnormality of the water flow velocity at each sampling point, and obtains the necessity of weight correction at each sampling point; according to the pollutant type quantity sequence, analyzes the cumulative change in the number of pollutant types, and obtains the type constant at each sampling point; according to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model, predict the pollutant concentrations in different areas of the water area to be traced, and identify the pollution source. The corrected prediction model corrects the model prediction error caused by the confluence effect, thereby improving the accuracy and efficiency of pollution source tracing.

Description

Technical Field

[0001] The present invention relates to the technical field of water pollution detection, and in particular to a rapid tracing method and system for tracing the source of groundwater pollution. Background Art

[0002] There are many methods for tracing the source of groundwater pollution, which are generally divided into physical tracing, which uses hydrogeological models to analyze the flow direction and velocity of groundwater to determine the possible source of pollutants; chemical analysis, which uses specific pollutant components and proportions to identify the characteristics of different pollution sources (chemical fingerprint analysis); and microbial methods, which analyze the types of microorganisms and their community structure in groundwater to determine the characteristics of pollution sources. Among them, the main detection equipment includes flow meters, chromatographs, ultraviolet spectrometers, etc.

[0003] For general water pollution source tracing needs, chemical fingerprint analysis is usually used, that is, the source of pollution is determined based on parameters such as the proportion of pollutants in different sampling waters. The steps of chemical fingerprint analysis mainly include: setting sampling points at different geographical locations in the polluted area to collect groundwater samples; using high-precision instruments to analyze parameters such as the concentration of pollutants in water samples, and then obtaining the concentration coefficient of the area to determine the possibility of the source. Different water areas have different concentration coefficients. By modeling and analyzing the concentration values ​​of different sampling waters, the model prediction results are obtained, and the source of pollution is determined based on the model prediction results. In this process, due to the nonlinear migration characteristics of groundwater, that is, the nonlinear characteristics of groundwater flow (confluence effect), the concentration of pollutants does not decrease monotonically with distance, which makes it possible to establish a prediction model based on pollutant concentrations. There will be errors, which increases the difficulty of locating the pollution source. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a rapid tracing method and system for tracing the source of groundwater pollution.

[0005] According to a first aspect of an embodiment of the present invention, a rapid tracing method for tracing the source of groundwater pollution is provided, and the technical solution adopted is specifically as follows:

[0006] Take samples multiple times to obtain the water velocity sequence at different sampling points in the waters to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points;

[0007] According to the pollutant concentration series of samples at different sampling points, a preliminary prediction model of pollutant concentration at different sampling points is constructed based on the prediction algorithm model;

[0008] Based on the water flow velocity sequence, analyzing the abnormality of the water flow velocity at each sampling point, and obtaining the necessity of weight correction at each sampling point;

[0009] According to the pollutant type quantity sequence, analyzing the cumulative change in the number of pollutant types to obtain a constant number of types at each sampling point;

[0010] According to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model;

[0011] Based on the modified prediction model, the pollutant concentrations in different areas of the waters to be traced are predicted and the pollution sources are identified.

[0012] In some embodiments of the present invention, according to the pollutant concentration sequence of samples at different sampling points, based on the prediction algorithm model, a preliminary prediction model of pollutant concentration at different sampling points is constructed, including:

[0013] Based on the ARIMA prediction model, the pollutant concentration series and weight values ​​of samples at different sampling points are input to build a preliminary prediction model for pollutant concentrations at different sampling points;

[0014] The spatial coordinates of the sampling points are used as independent variables to output the predicted values ​​of pollutant concentrations at different sampling points.

[0015] In some embodiments of the present invention, based on the water flow velocity sequence, analyzing the abnormality of the water flow velocity at each sampling point to obtain the necessity of weight correction at each sampling point includes:

[0016] According to the water flow velocity sequence of each sampling point, the mean water flow velocity of each sampling point is obtained;

[0017] Obtain the median of the mean water velocity values ​​at all sampling points;

[0018] Analyzing the outliers of the water flow velocity at each sampling point according to the mean water flow velocity;

[0019] According to the mean value and the median value of the water flow velocity, combined with the outlier property, the abnormality of the water flow velocity at each sampling point is analyzed to obtain the necessity of weight correction for each sampling point.

[0020] In some embodiments of the present invention, the necessity of weight correction of each sampling point is obtained, and then the following steps are further included:

[0021] According to the necessity of weight correction, judging whether there is a confluence effect at the sampling point, and obtaining a confluence effect area;

[0022] The weight value of the confluence effect area participating in the preliminary prediction model is modified.

[0023] In some embodiments of the present invention, judging whether there is a confluence effect at the sampling point according to the necessity of weight correction and obtaining a confluence effect area includes:

[0024] Setting a threshold of necessity;

[0025] Determining whether the necessity of weight correction is greater than or equal to the necessity threshold;

[0026] If yes, there is a confluence effect at the sampling point corresponding to the necessity of weight correction, that is, the sampling point corresponding to the necessity of weight correction is a confluence effect area;

[0027] If not, there is no confluence effect at the sampling point corresponding to the necessity of weight correction, that is, the sampling point corresponding to the necessity of weight correction is a non-confluence effect area.

[0028] In some embodiments of the present invention, according to the pollutant type quantity sequence, analyzing the cumulative change in the number of pollutant types to obtain the type constant at each sampling point includes:

[0029] According to the pollutant type quantity sequence, the mean value of the pollutant type quantity is obtained;

[0030] According to the pollutant type quantity sequence, the difference between the pollutant type quantities of samples obtained at adjacent times is calculated, and all samples obtained at adjacent times are traversed to obtain the cumulative change in the pollutant type quantities;

[0031] According to the accumulated change amount and in combination with the mean value of the number of pollutant types, the constant amount of the type at each sampling point is obtained.

[0032] In some embodiments of the present invention, according to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model, including:

[0033] According to the necessity of weight correction and the type constant, the proximity degree of the pollution source is obtained by setting the parameter index weight;

[0034] According to the pollutant concentration sequence, a mean pollutant concentration is obtained;

[0035] According to the proximity of the pollution source and the mean value of the pollutant concentration, the weight value of the preliminary prediction model is modified to obtain a modified prediction model.

[0036] In some embodiments of the present invention, multiple sampling is performed to obtain a water velocity sequence at different sampling points in the water area to be traced, as well as a pollutant concentration sequence and a pollutant type quantity sequence of samples at different sampling points, including:

[0037] According to the area and shape of the water area to be traced, the water area to be traced is divided into evenly distributed sampling points manually;

[0038] Use a flow meter to measure the real-time water flow velocity at each sampling point to obtain a water flow velocity sequence;

[0039] Use a sampler to collect water samples from each sampling point at different times;

[0040] The water samples are measured to obtain a pollutant concentration sequence and a pollutant type quantity sequence of the water samples.

[0041] According to a second aspect of an embodiment of the present invention, a rapid tracing system for tracing a groundwater pollution source is provided, comprising: a memory and a processor, wherein:

[0042] The memory is used to store program codes;

[0043] The processor is used to read the program code stored in the memory and execute the method described in the first aspect of the embodiment of the present invention.

[0044] In some embodiments of the present invention, the processor comprises:

[0045] The data acquisition module is used for multiple sampling to obtain the water velocity sequence at different sampling points in the water area to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points;

[0046] A preliminary prediction model building module is used to build a preliminary prediction model of pollutant concentrations at different sampling points based on the pollutant concentration sequence of samples at different sampling points and the prediction algorithm model;

[0047] The revised prediction model construction module is first used to analyze the abnormality of the water flow velocity at each sampling point based on the water flow velocity sequence to obtain the necessity of weight correction at each sampling point; then used to analyze the cumulative change in the number of pollutant types according to the pollutant type number sequence to obtain the type constant at each sampling point; finally used to correct the weight value of the preliminary prediction model according to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, to obtain a revised prediction model;

[0048] The pollution source identification module is used to predict the pollutant concentrations in different areas of the water area to be traced and identify the pollution sources based on the modified prediction model.

[0049] Compared with the prior art, the rapid tracing method and system for tracing the source of groundwater pollution provided by the present invention first locates several sampling points of the water area to be traced, obtains the real-time water flow velocity through the flow meter, and obtains samples by multiple samplings, and analyzes the pollutant concentration sequence and the pollutant type quantity sequence in the sample; then, according to the pollutant concentration sequence of the samples at different sampling points, a preliminary prediction model of the pollutant concentration at different sampling points is constructed based on the prediction algorithm model; then, based on the water flow velocity sequence, the abnormality of the water flow velocity of each sampling point is analyzed to determine the necessity of weight correction; and for the part with high weight correction necessity, the type constant is analyzed according to the pollutant type quantity sequence; finally, according to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model. The corrected prediction model corrects the model prediction error caused by the confluence effect, and achieves the purpose of accurate and rapid tracing. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 A basic flow chart of a rapid tracing method for groundwater pollution source tracking provided by an embodiment of the present invention;

[0052] Figure 2 A schematic diagram of the basic composition of a rapid tracing system for tracing the source of groundwater pollution provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation method, structure, characteristics and effects of the rapid tracing method and system for tracing the source of groundwater pollution proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. Terms such as "comprises", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such article or device. In the absence of further restrictions, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the article or device including the element.

[0055] A specific scheme of a rapid tracing method for tracing the source of groundwater pollution provided by the present invention is described in detail below in conjunction with the accompanying drawings.

[0056] See also Figure 1 , which shows the basic process of a rapid tracing method for tracing the source of groundwater pollution provided by an embodiment of the present invention.

[0057] like Figure 1 As shown, a rapid tracing method for tracing the source of groundwater pollution provided by an embodiment of the present invention specifically includes:

[0058] S100: Take samples multiple times to obtain the water velocity sequence at different sampling points in the water area to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points.

[0059] Take samples multiple times to obtain the water velocity sequence at different sampling points in the waters to be traced, as well as the pollutant concentration sequence and pollutant type and quantity sequence of samples at different sampling points. Specifically including:

[0060] Locate sampling points: According to the area and shape of the waters to be traced, the waters to be traced are divided into evenly distributed sampling points. Specifically, every 5% of the area of ​​the waters to be traced can be used as a sampling area, and a sampling point can be located. Sampling area, set Each sampling point is marked with a fixed spatial coordinate to ensure that subsequent sampling is consistent, that is, each sampling point contains information about the spatial coordinates and the sample area number.

[0061] Flow velocity monitoring: Use a flow meter to measure the real-time water velocity at each sampling point, obtain a water velocity sequence, and record the change in water velocity to assist in nonlinear migration analysis. The flow meter can be a Doppler flow meter, model SonTek FlowTracker, with a flow velocity range of 0.001–4 m / s and an accuracy of ±1%.

[0062] Water sampling: Use a sampler to collect water samples from each sampling point at different times. Commonly used portable water samplers are ISCO 6712, with a sampling volume of about 500 mL. To avoid cross contamination, use a 500 mL high-density polyethylene (HDPE) disposable sampling bottle.

[0063] Water quality analysis: Use a portable water quality analyzer (such as Hach DR3900) to conduct on-site analysis and measurement of water samples to determine the pollutant concentration (mg / L) and the number of pollutant types (based on characteristic ions or chemical fingerprints, which can be further confirmed in combination with laboratory analysis) to obtain the pollutant concentration sequence and pollutant type quantity sequence of the water samples.

[0064] Data recording and processing: The sample area serial number and spatial coordinates of the sampling point, the water velocity sequence (a sequence of water velocity corresponding to the sample acquisition time), the pollutant concentration sequence and the pollutant type quantity sequence are recorded in a spreadsheet for subsequent model construction.

[0065] S200: According to the pollutant concentration sequence of samples at different sampling points and based on the prediction algorithm model, a preliminary prediction model of pollutant concentrations at different sampling points is constructed.

[0066] In order to achieve rapid tracing of groundwater pollution sources, a preliminary prediction model is first established based on the above sample parameters. In this embodiment, according to the pollutant concentration sequence of samples at different sampling points, a preliminary prediction model of pollutant concentrations at different sampling points is constructed based on the prediction algorithm model. Specifically, based on the ARIMA prediction model, the pollutant concentration sequence of samples at different sampling points and their weight values ​​(the initial value of each weight value is 1) are input to construct a preliminary prediction model of pollutant concentrations at different sampling points. The preliminary prediction model consists of the spatial coordinates of the sampling points. And the pollutant concentration corresponding to the sampling point and weight value It should be noted that the input value is the pollutant concentration sequence of samples at different sampling points, and the actual calculation is the mean of the pollutant concentration. The mean of the pollutant concentration is obtained by calculating the pollutant concentration sequence data. Using the mean of the pollutant concentration to participate in the calculation can avoid the instability of a single data and better reflect the overall characteristics of the sampling point; the spatial coordinates of the sampling point are used as independent variables to output the predicted values ​​of pollutant concentrations at different sampling points. The preliminary prediction model constructed based on the ARIMA prediction model predicts the pollutant concentration in the undetected area, finds the area with the maximum pollutant concentration, and obtains the pollution source, that is, the sampling point area corresponding to the maximum value of the output pollutant concentration prediction value is the predicted suspected pollution source. The form of the preliminary prediction model is:

[0067]

[0068] In the formula, Indicates The predicted value of pollutant concentration corresponding to each sampling point; , , Respectively represent Sampling points, sampling points and The mean pollutant concentration corresponding to the sampling points is Other than sampling points The mean pollutant concentration of all samples collected at the sampling points; , , Respectively represent Sampling points, sampling points and The weight value of the mean pollutant concentration corresponding to each sampling point in the preliminary prediction model, that is, except for the Other than sampling points The weight value of the mean pollutant concentration corresponding to each sampling point in the preliminary prediction model.

[0069] The ARIMA prediction model is an existing technology, and the specific steps are not repeated here. Based on the ARIMA prediction model, the pollutant concentrations of samples at different sampling points are input. and weight value , with the spatial coordinates of the sampling points As an independent variable, a preliminary prediction model is obtained. The preliminary prediction model retains the weight value coefficient to facilitate the subsequent correction of the weight value. The initial value of each weight value in the preliminary prediction model is 1. Due to the existence of the confluence effect, the weight values ​​of different samples need to be changed to avoid the prediction result error caused by the confluence effect.

[0070] S300: Based on the water flow velocity sequence, analyze the abnormality of the water flow velocity at each sampling point to obtain the necessity of weight correction at each sampling point.

[0071] The confluence effect refers to the phenomenon that different undercurrents converge to one area in the groundwater environment. In this confluence area, the water flow velocity is relatively turbulent, which is the confluence effect. In addition, the direction of water flow in the confluence area is relatively fixed, and the velocity is highest at the confluence. Because of the existence of the confluence effect, the concentration of pollutants in the confluence area may be diluted, which in turn causes deviations in the prediction results of the model constructed based on the concentration law. At this time, it is necessary to analyze the impact of the environment and correct the weight value. For example, some areas with low concentrations do not necessarily have a small weight when constructing the model results, and thus it is necessary to eliminate the prediction deviation caused by the confluence effect. Combined with the above analysis, in areas with high water flow velocities, it is necessary to pay further attention to the corresponding relevant pollution parameters to avoid prediction deviations caused by diluting the relevant concentration weights due to high water flow velocities.

[0072] Based on the above analysis, in this embodiment, based on the water flow velocity sequence, the abnormality of the water flow velocity of each sampling point is analyzed to obtain the necessity of weight correction of each sampling point. Specifically, it includes: according to the water flow velocity sequence of each sampling point, the mean water flow velocity of each sampling point is obtained. Here, the water flow velocity of the sampling point is represented by the mean water flow velocity, which can avoid the instability of single data and better reflect the overall characteristics of the sampling point; the median of the mean water flow velocity of all sampling points is obtained. The median is not affected by extreme values ​​and can provide a stable estimate of the general level of data. In the case of skewed data distribution or the presence of outliers, the median is usually more representative of the central trend of the data than the mean. Therefore, the median is used here to represent the general level of water flow velocity at all sampling points; according to the mean water flow velocity, the outlier of the water flow velocity of each sampling point is analyzed, wherein the outlier analysis adopts the existing technology, such as the Grubbs test statistical method, etc., and the specific steps are not repeated; according to the mean and median water flow velocity, combined with the outlier, the abnormality of the water flow velocity of each sampling point is analyzed to obtain the necessity of weight correction of each sampling point. Among them, the quantitative formula for the necessity of weight correction of sampling points is:

[0073]

[0074] In the formula, Indicates Necessity of weight correction for each sampling point; Indicates The mean water velocity at each sampling point; represents the median value of the mean water velocity at all sampling points; Indicates The outliers of water velocity at each sampling point; represents the linear normalization function.

[0075] By analyzing the difference between the mean water velocity of the area where the sampling point is located and the median of the mean water velocity of the areas where all sampling points are located , combined with the outliers of the sample flow rate values , determine the abnormality of the water flow velocity in the area where the sampling point is located. The farther away from the median of the mean water flow velocity, the more abnormal the water flow velocity in the area where the sampling point is located is compared with the overall water flow velocity, because the distribution range of its water flow velocity is farther away from the normal value, that is, the possibility of confluence effect in the area where the sampling point is located is higher, and the higher the degree of dilution of the corresponding concentration value, the higher the necessity of correcting the weight value of the prediction model, that is, the greater the necessity of correcting the weight value corresponding to the sampling point; and, the higher the outlier of the water flow velocity of the sampling point, that is, the higher the difference between the water flow velocity in the area where the sampling point is located and other water flow velocities, the more abnormal it is, indicating that the possibility of confluence in the area where the sampling point is located is higher, and the higher the degree of dilution of the corresponding concentration value, the higher the necessity of correcting the weight value of the prediction model, that is, the greater the necessity of correcting the weight value corresponding to the sampling point.

[0076] In some embodiments of the present invention, the necessity of weight correction of each sampling point is obtained, and then the following steps are further included:

[0077] According to the necessity of weight correction, it is judged whether there is a confluence effect at the sampling point, and the confluence effect area is obtained. According to the necessity of weight correction, it is judged whether there is a confluence effect at the sampling point, and the confluence effect area is obtained. The specific steps are: setting the necessity threshold, which can be set to 0.7; judging whether the necessity of weight correction is greater than or equal to the necessity threshold; if so, , then there is a confluence effect at the sampling point corresponding to the necessity of weight correction, that is, the sampling point corresponding to the necessity of weight correction is the confluence effect area; if not, that is , then there is no confluence effect at the sampling point corresponding to the necessity of weight correction, that is, the sampling point corresponding to the necessity of weight correction is a non-confluence effect area. The weight value of the confluence effect area participating in the preliminary prediction model is corrected.

[0078] The weight value of the confluence effect area participating in the preliminary prediction model is corrected, and the specific correction method is step S400 and step S500.

[0079] S400: Analyze the cumulative change in the number of pollutant types according to the pollutant type number sequence to obtain the type constant at each sampling point.

[0080] When the confluence effect appears in an area, the water quality within the area is not easy to accumulate corresponding pollutants due to the fast water flow speed, and the corresponding pollutant concentration is low; except for the pollutant concentration parameter, the parameter value of the number of pollutant types is less affected by the confluence effect, that is, the number of pollutant types is not easy to change significantly; the specific reason is: when the water flow speed is accelerated, the pollutants are quickly diffused and diluted, but the dilution effect is balanced for different types of pollutants, so although the concentration of a single pollutant is reduced, the total number of types will not decrease significantly. At this time, the higher the constant amount of types, the more severe the confluence effect.

[0081] Based on the above analysis, this embodiment analyzes the cumulative change in the number of pollutant types according to the pollutant type number sequence to obtain the type constant at each sampling point. Specifically, it includes: obtaining the mean number of pollutant types according to the pollutant type number sequence; calculating the absolute value of the difference between the number of pollutant types of samples obtained at adjacent times according to the pollutant type number sequence, traversing all samples obtained at adjacent times, and obtaining the cumulative change in the number of pollutant types; according to the cumulative change, combined with the mean number of pollutant types, obtaining the type constant at each sampling point. Among them, the quantitative formula of the type constant of the sampling point is:

[0082]

[0083] In the formula, Indicates The constant quantity of the types of sampling points; Indicates The mean number of pollutant types corresponding to each sampling point; Indicates The number of pollutant types in the samples collected at different times from the sampling points; Indicates The absolute value of the difference in the number of pollutant types in samples collected at adjacent times from the sampling points; Indicates The total number of samples collected at different times from each sampling point; represents the linear normalization function; Represents a correction parameter, which can be 0.1 and is used to prevent the denominator from being zero.

[0084] The absolute value of the difference in the number of pollutant types in samples collected at adjacent times at the sampling point The smaller it is, the more constant the number of pollutant types corresponding to the sampling point is in the time dimension, so the constant number of types at the sampling point is greater; It means that the reciprocal sum of the absolute value of the difference in the number of pollutant types of samples collected at all adjacent times of the sampling point is obtained to obtain the cumulative change in the number of pollutant types. The larger the value, the larger the corresponding type constant. The larger the value is, the larger the constant value of the corresponding sampling point is. Near the pollution source, even if there is a confluence effect, pollutants are continuously generated, and the corresponding number of pollutant types exists at a high number at different times.

[0085] In some other embodiments of the present invention, for the confluence effect area, that is, the pollutant type quantity sequence corresponding to the samples in the confluence effect area is obtained, and according to the pollutant type quantity sequence, the cumulative change in the number of pollutant types is analyzed to obtain the type constant of each sampling point. The specific operation is the same as the above, and will not be repeated here. The main difference between this embodiment and the above embodiment is that only the type horizontal quantitative analysis is performed on the sampling points in the confluence effect area. For the non-confluence effect area, since there is no confluence effect or the confluence effect is small, the corresponding weight value in the model does not need to be corrected, so it is not necessary to perform the type constant analysis on it, which can reduce the amount of data analysis and improve the analysis efficiency.

[0086] S500: According to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model.

[0087] On the basis of the constant amount of species and high abnormality of water flow velocity (necessity of weight correction), the characteristics of the area where the sampling point is located can be characterized as follows: even if the pollutant concentration in the area becomes lower due to the corresponding confluence effect, the types of pollutants are still rich and continue to flow at different times. Therefore, the area with lower concentration of such pollutants has a higher weight value.

[0088] Based on the above analysis, in this embodiment, according to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model. Further including:

[0089] First, according to the necessity of weight correction and the type constant, the parameter index weight is set to obtain the degree of proximity to the pollution source. Specifically, the larger the type constant value, the higher the necessity of weight correction, the more likely it is to be an area close to the pollution source, and the corresponding weight value corresponding to the pollutant concentration in the area should be enlarged. Therefore, the calculation formula for the degree of proximity to the pollution source is:

[0090]

[0091] In the formula, Indicates The proximity of each sampling point to the pollution source; Represents the parameter index weight. Here, the weight of the constant type is increased, so ; Indicates Necessity of weight correction for each sampling point; Indicates The type constant of the sampling points.

[0092] The larger the constant value of the species, the more necessary the weight correction is, and the more likely it is that the area is close to the pollution source. The higher the value.

[0093] Then, based on the pollutant concentration sequence, the mean pollutant concentration is obtained.

[0094] Finally, according to the proximity of the pollution source and the mean value of the pollutant concentration, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model. Specifically, according to the proximity of the pollution source and the mean value of the pollutant concentration, the weight value of the preliminary prediction model is corrected to obtain the model correction weight. The calculation formula of the model correction weight is:

[0095]

[0096] In the formula, Indicates Model correction weight corresponding to each sampling point; Indicates The weight value corresponding to each sampling point; Indicates The proximity of each sampling point to the pollution source; Indicates The mean pollutant concentration of all samples collected at the sampling points; represents the linear normalization function.

[0097] The size of the model correction weight is inversely proportional to the pollutant concentration value, that is, on the basis of satisfying the proximity of the pollution source, if the pollutant concentration value is still small, then it can be determined that the pollutant concentration is diluted due to the confluence effect, but in fact the area is near a pollution source with a higher pollutant concentration. Because the corresponding indicators of other parameters are quantified higher, it is necessary to increase the weight of the pollutant concentration corresponding to the sampling point, that is, it is necessary to increase the model correction weight corresponding to the sampling point.

[0098] After obtaining the model correction weight, the preliminary prediction model is corrected to obtain a corrected prediction model. The predicted pollutant concentration result output by the corrected prediction model is:

[0099]

[0100] In the formula, Indicates The predicted value of the corrected pollutant concentration corresponding to each sampling point; , , Respectively represent Sampling points, sampling points and The mean pollutant concentration corresponding to the sampling points is Other than sampling points The mean pollutant concentration of all samples collected at the sampling points; , , Respectively represent Sampling points, sampling points and The weight value of the mean pollutant concentration corresponding to each sampling point in the preliminary prediction model, that is, except for the Other than sampling points The model correction weight of the mean pollutant concentration corresponding to each sampling point in the corrected prediction model.

[0101] S600: Based on the modified prediction model, the pollutant concentrations in different areas of the waters to be traced are predicted and the pollution sources are identified.

[0102] By modifying the prediction model to predict the pollutant concentration values ​​in different areas of the waters to be traced, the areas corresponding to the highest pollutant concentration values ​​are marked as suspected pollution sources, thereby achieving the purpose of rapid and efficient tracing.

[0103] After obtaining the suspected pollution source through the modified prediction model, the specific operations to further determine the pollution source are as follows:

[0104] Comparison between predicted concentration values ​​and measured values: Compare the predicted pollutant concentration values ​​output by the modified prediction model with the actually monitored pollutant concentration values ​​to identify abnormal areas or time points (i.e. areas or time points where there are significant deviations between the predicted values ​​and the measured values), and eliminate abnormal areas and time points.

[0105] Spatial distribution analysis: Map the predicted pollutant concentration values ​​to different areas and observe the spatial distribution characteristics. Areas with higher pollutant concentration values ​​are close to potential pollution sources.

[0106] Time series analysis: Combined with the time series characteristics of the ARIMA model, the trend of pollutant concentration changes is analyzed. Areas with higher stability in pollutant concentration values ​​are close to potential pollution sources. This is mainly because although the pollutant concentration at the pollution source may not be the highest, it has a continuous characteristic in the time series.

[0107] Based on the same inventive concept as the above method, this embodiment also provides a rapid tracing system for tracing the source of groundwater pollution.

[0108] See also Figure 2 , which shows the basic composition of a rapid tracing system for tracing the source of groundwater pollution provided by an embodiment of the present invention.

[0109] like Figure 2 As shown, a rapid tracing system for tracing the source of groundwater pollution includes: a memory 10 and a processor 20, wherein:

[0110] A memory 10, used for storing program codes;

[0111] The processor 20 is used to read the program code stored in the memory 10 and perform multiple samplings to obtain the water flow velocity sequence of different sampling points in the water area to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points; based on the pollutant concentration sequence of samples at different sampling points, a preliminary prediction model of pollutant concentrations at different sampling points is constructed based on the prediction algorithm model; based on the water flow velocity sequence, the abnormality of the water flow velocity of each sampling point is analyzed to obtain the necessity of weight correction for each sampling point; based on the pollutant type quantity sequence, the cumulative change in the number of pollutant types is analyzed to obtain the type constant of each sampling point; based on the necessity of weight correction and the type constant, the weight value of the preliminary prediction model is corrected in combination with the pollutant concentration sequence to obtain a corrected prediction model; based on the corrected prediction model, the pollutant concentrations in different areas of the water area to be traced are predicted to identify the pollution source.

[0112] Furthermore, the processor 20 includes a data acquisition module 21, a preliminary prediction model building module 22, a revised prediction model building module 23 and a pollution source identification module 24. Among them:

[0113] The data acquisition module 21 is used for multiple sampling to obtain the water velocity sequence at different sampling points in the water area to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points;

[0114] A preliminary prediction model building module 22 is used to build a preliminary prediction model of pollutant concentrations at different sampling points based on the pollutant concentration sequence of samples at different sampling points and the prediction algorithm model;

[0115] The revised prediction model construction module 23 is first used to analyze the abnormality of the water flow velocity at each sampling point based on the water flow velocity sequence to obtain the necessity of weight correction at each sampling point; then used to analyze the cumulative change in the number of pollutant types according to the pollutant type number sequence to obtain the type constant at each sampling point; finally used to correct the weight value of the preliminary prediction model according to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, to obtain a revised prediction model;

[0116] The pollution source identification module 24 is used to predict the pollutant concentrations in different areas of the water area to be traced based on the modified prediction model and identify the pollution sources.

[0117] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0118] Various embodiments in this specification are described in a progressive manner, and the same or similar parts between various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A rapid tracing method for groundwater pollution source tracking, characterized in that: The method comprises: Take samples multiple times to obtain the water velocity sequence at different sampling points in the waters to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points; According to the pollutant concentration series of samples at different sampling points, a preliminary prediction model of pollutant concentration at different sampling points is constructed based on the prediction algorithm model; Based on the water flow velocity sequence, analyzing the abnormality of the water flow velocity at each sampling point, and obtaining the necessity of weight correction at each sampling point; According to the pollutant type quantity sequence, analyzing the cumulative change in the number of pollutant types to obtain a constant number of types at each sampling point; According to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model; Based on the modified prediction model, the pollutant concentrations in different areas of the waters to be traced are predicted to identify the pollution sources; Based on the water flow velocity sequence, the abnormality of the water flow velocity at each sampling point is analyzed to obtain the necessity of weight correction at each sampling point, including: According to the water flow velocity sequence of each sampling point, the mean water flow velocity of each sampling point is obtained; Obtain the median of the mean water velocity values ​​at all sampling points; Analyzing the outliers of the water flow velocity at each sampling point according to the mean water flow velocity; According to the mean value and the median value of the water flow velocity, combined with the outlier property, the abnormality of the water flow velocity at each sampling point is analyzed to obtain the necessity of weight correction for each sampling point.

2. The rapid tracing method for groundwater pollution source tracking according to claim 1 is characterized in that: According to the pollutant concentration sequence of samples at different sampling points, based on the prediction algorithm model, a preliminary prediction model for pollutant concentrations at different sampling points is constructed, including: Based on the ARIMA prediction model, the pollutant concentration series and weight values ​​of samples at different sampling points are input to build a preliminary prediction model for pollutant concentrations at different sampling points; The spatial coordinates of the sampling points are used as independent variables to output the predicted values ​​of pollutant concentrations at different sampling points.

3. The rapid tracing method for groundwater pollution source tracking according to claim 1 is characterized in that: Obtain the necessity of weight correction for each sampling point, followed by: According to the necessity of weight correction, judging whether there is a confluence effect at the sampling point, and obtaining a confluence effect area; The weight value of the confluence effect area participating in the preliminary prediction model is modified.

4. The rapid tracing method for groundwater pollution source tracking according to claim 3 is characterized in that: According to the necessity of weight correction, whether there is a confluence effect at the sampling point is judged, and a confluence effect area is obtained, including: Setting a threshold of necessity; Determining whether the necessity of weight correction is greater than or equal to the necessity threshold; If yes, there is a confluence effect at the sampling point corresponding to the necessity of weight correction, that is, the sampling point corresponding to the necessity of weight correction is a confluence effect area; If not, there is no confluence effect at the sampling point corresponding to the necessity of weight correction, that is, the sampling point corresponding to the necessity of weight correction is a non-confluence effect area.

5. The rapid tracing method for groundwater pollution source tracking according to claim 1 or 3 is characterized in that: According to the pollutant type quantity sequence, the cumulative change in the number of pollutant types is analyzed to obtain the type constant at each sampling point, including: According to the pollutant type quantity sequence, the mean value of the pollutant type quantity is obtained; According to the pollutant type quantity sequence, the difference between the pollutant type quantities of samples obtained at adjacent times is calculated, and all samples obtained at adjacent times are traversed to obtain the cumulative change in the pollutant type quantities; According to the accumulated change amount and in combination with the mean value of the number of pollutant types, the constant amount of the type at each sampling point is obtained.

6. The rapid tracing method for groundwater pollution source tracking according to claim 1 is characterized in that: According to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, the weight value of the preliminary prediction model is corrected to obtain a corrected prediction model, including: According to the necessity of weight correction and the type constant, the proximity degree of the pollution source is obtained by setting the parameter index weight; According to the pollutant concentration sequence, a mean pollutant concentration is obtained; According to the proximity of the pollution source and the mean value of the pollutant concentration, the weight value of the preliminary prediction model is modified to obtain a modified prediction model.

7. The rapid tracing method for groundwater pollution source tracking according to claim 1 is characterized in that: Take samples multiple times to obtain the water velocity sequence at different sampling points in the waters to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points, including: According to the area and shape of the water area to be traced, the water area to be traced is divided into evenly distributed sampling points manually; Use a flow meter to measure the real-time water flow velocity at each sampling point to obtain a water flow velocity sequence; Use a sampler to collect water samples from each sampling point at different times; The water samples are measured to obtain a pollutant concentration sequence and a pollutant type quantity sequence of the water samples.

8. A rapid tracing system for tracing the source of groundwater pollution, characterized in that: The system comprises: a memory and a processor, wherein: The memory is used to store program codes; The processor is configured to read the program code stored in the memory and execute the method according to any one of claims 1 to 7.

9. The rapid tracing system for tracing the source of groundwater pollution according to claim 8 is characterized in that: The processor comprises: The data acquisition module is used for multiple sampling to obtain the water velocity sequence at different sampling points in the water area to be traced, as well as the pollutant concentration sequence and pollutant type quantity sequence of samples at different sampling points; A preliminary prediction model building module is used to build a preliminary prediction model of pollutant concentrations at different sampling points based on the pollutant concentration sequence of samples at different sampling points and the prediction algorithm model; The revised prediction model construction module is first used to analyze the abnormality of the water flow velocity at each sampling point based on the water flow velocity sequence to obtain the necessity of weight correction at each sampling point; then used to analyze the cumulative change in the number of pollutant types according to the pollutant type number sequence to obtain the type constant at each sampling point; finally used to correct the weight value of the preliminary prediction model according to the necessity of weight correction and the type constant, combined with the pollutant concentration sequence, to obtain a revised prediction model; A pollution source identification module, used to predict the pollutant concentrations in different areas of the waters to be traced based on the modified prediction model, and identify the pollution sources; Based on the water flow velocity sequence, the abnormality of the water flow velocity at each sampling point is analyzed to obtain the necessity of weight correction at each sampling point, including: According to the water flow velocity sequence of each sampling point, the mean water flow velocity of each sampling point is obtained; Obtain the median of the mean water velocity values ​​at all sampling points; Analyzing the outliers of the water flow velocity at each sampling point according to the mean water flow velocity; According to the mean value and the median value of the water flow velocity, combined with the outlier property, the abnormality of the water flow velocity at each sampling point is analyzed to obtain the necessity of weight correction for each sampling point.

Images