A method and device for configuring a deep reservoir suspended substance inversion model considering sources

Abstract

This application provides a method and apparatus for configuring a suspended solids inversion model that takes into account the source of suspended solids in the Shenzhen-Dalian Reservoir. Based on the observation that there are differences in suspended solids inversion between water bodies with internal and external suspended solids in the Shenzhen-Dalian Reservoir scenario, it introduces K-means clustering analysis combined with Manhattan distance to configure differentiated suspended solids inversion models. This enables the model to achieve a deep-level adaptation and high-precision suspended solids inversion effect for different water bodies, thereby providing good data support for water quality monitoring in the Shenzhen-Dalian Reservoir.

Description

Technical Field

[0001] This application relates to the field of water quality monitoring, specifically to a method and apparatus for configuring a suspended solids inversion model for deep and large reservoirs that takes into account the source. Background Technology

[0002] Affected by multiple factors such as upstream water flow, rainfall, artificial scheduling, and seasonal changes, the source and composition of particulate matter in the Shenzhen-Dalian Reservoir are constantly changing, resulting in dynamic fluctuations in the optical properties of the water body. This often leads to complex and continuous dynamic changes in the water body type, which puts a burden on water quality monitoring.

[0003] In existing technologies, traditional single suspended matter inversion algorithms are mostly based on the single feature of reflectivity in the strong absorption band of pigment or non-pigment particles to build models. However, in the case of dynamic changes in the water body of deep and large reservoirs, it is difficult to take into account the spatiotemporal heterogeneity of particle composition and the resulting differences in the optical properties of the water body. This leads to deviations between the model assumptions and the actual water conditions, which can easily result in problems such as decreased inversion accuracy, amplified errors, or even complete failure.

[0004] Therefore, in order to accurately realize the dynamic inversion of suspended matter in deep and large reservoirs, it is urgent to adapt to the complex changes in water body types, break through the application limitations of single algorithms, carry out targeted classification inversion algorithm research, and construct suspended matter inversion models suitable for different water body types. Summary of the Invention

[0005] This application provides a method and apparatus for configuring a suspended solids inversion model that takes into account the source of suspended solids in the Shenzhen-Dalian Reservoir. Based on the observation that there are differences in suspended solids inversion between water bodies with internal and external suspended solids in the Shenzhen-Dalian Reservoir scenario, this application introduces K-means clustering analysis combined with Manhattan distance to configure differentiated suspended solids inversion models. This allows for deep adaptation and high-precision suspended solids inversion results for different water bodies, thereby providing good data support for water quality monitoring in the Shenzhen-Dalian Reservoir.

[0006] Firstly, this application provides a method for configuring a suspended solids inversion model for deep and large reservoirs that takes into account the source, the method including: For the target deep reservoir, determine the corresponding sensitive spectral characteristics in different water bodies with different sources of suspended solids; For water bodies with different sources of suspended solids, the corresponding sensitive spectral features are used as a reference, and K-means clustering analysis based on Manhattan distance is combined to obtain the results of differential measurement of water body spectral features. Based on the differentiated measurement results of the corresponding water spectral characteristics of water bodies with different sources of suspended matter, different suspended matter inversion models for deep and large reservoirs are configured. Among them, the suspended matter inversion model for deep and large reservoirs is used to perform suspended matter concentration inversion processing on the target suspended matter source water body based on the water spectral data input to the model and combined with the target sensitive spectral characteristics of the target suspended matter source water body in the appropriate band.

[0007] Secondly, this application provides a configuration device for a source-inverted model of suspended solids in deep reservoirs, the device comprising: The determination unit is used to determine the corresponding sensitive spectral characteristics in the spectral bands for different suspended solids sources involved in the target deep reservoir. The analysis unit is used to obtain the differential measurement results of the spectral characteristics of water bodies with different sources of suspended matter, using the corresponding sensitive spectral characteristics as a reference and combining K-means clustering analysis based on the Manhattan distance strategy. The configuration unit is used to configure different suspended solids inversion models for deep and large reservoirs based on the different measurement results of the corresponding water spectral characteristics of water bodies with different sources of suspended solids. The suspended solids inversion model for deep and large reservoirs is used to perform suspended solids concentration inversion processing on the target suspended solids source water body based on the water spectral data input to the model and the band of the target sensitive spectral characteristics of the target suspended solids source water body.

[0008] Thirdly, this application provides a processing device, including a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the method provided in the first aspect of this application when it invokes the computer program in the memory.

[0009] Fourthly, this application provides a computer-readable storage medium storing a plurality of instructions adapted for loading by a processor to execute the method provided in the first aspect of this application.

[0010] From the above, it can be concluded that this application has the following beneficial effects: To address the suspended solids inversion target of the Shenzhen Grand Reservoir, this application, based on the differences in suspended solids inversion between endogenous and exogenous suspended solids in the Shenzhen Grand Reservoir scenario, introduces K-means clustering analysis combined with Manhattan distance to configure differentiated suspended solids inversion models. This enables the deep adaptation and high accuracy of suspended solids inversion for different water bodies, thereby providing good data support for water quality monitoring in the Shenzhen Grand Reservoir. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A schematic flowchart illustrating a method for configuring a suspended solids inversion model for the Shenzhen Reservoir, which is relevant to this application. Figure 2 This is a schematic diagram of the overall design architecture of the solution in this application. Figure 3 A schematic diagram of a configuration device for the suspended solids inversion model of the Shenzhen Reservoir, which is considered in this application; Figure 4 This is a schematic diagram of one type of processing equipment used in this application. Detailed Implementation

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules is not necessarily limited to those explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices. The naming or numbering of steps appearing in this application does not imply that the steps in the method flow must be performed in the chronological / logical order indicated by the naming or numbering. The execution order of named or numbered process steps can be changed according to the desired technical purpose, as long as the same or similar technical effect is achieved.

[0015] The module division described in this application is a logical division. In practical applications, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual coupling, direct coupling, or communication connections may be through interfaces, and the indirect coupling or communication connections between modules may be electrical or other similar forms, none of which are limited in this application. Moreover, the modules or sub-modules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed across multiple circuit modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution in this application.

[0016] Before introducing the source-introduction model configuration method for suspended solids in deep reservoirs provided in this application, we will first introduce the background content involved in this application.

[0017] The method, apparatus, and computer-readable storage medium for configuring suspended solids inversion models for the Shenzhen-Dongming Reservoir, which take into account the source, provided in this application can be applied to processing equipment. Based on the differences in suspended solids inversion between endogenous and exogenous suspended solids in the Shenzhen-Dongming Reservoir scenario, this application introduces K-means clustering analysis combined with Manhattan distance to configure differentiated suspended solids inversion models. This enables the configuration of highly accurate suspended solids inversion models that are deeply adapted to different water bodies, thereby providing good data support for water quality monitoring in the Shenzhen-Dongming Reservoir.

[0018] The source-considered method for configuring the suspended solids inversion model of the Shenzhen-Dongming Reservoir mentioned in this application can be implemented by a source-considered device for configuring the suspended solids inversion model of the Shenzhen-Dongming Reservoir, or by different types of processing devices such as servers, physical hosts, or user equipment (UE) that integrate the source-considered device. The source-considered device can be implemented in hardware or software. The UE can be a smartphone, tablet, laptop, desktop computer, or personal digital assistant (PDA) or other terminal device. The processing devices can be configured in a cluster.

[0019] It is understandable that the proposed solution is usually based on existing data or data that has already been collected. Therefore, the processing equipment that implements the configuration method of the suspended solids inversion model of Shenzhen Reservoir (which is based on the source of this application) or that is equipped with the corresponding application service of the configuration method of the suspended solids inversion model of Shenzhen Reservoir (which is based on the source of this application) usually only needs to meet the required data processing capabilities. The specific equipment type and equipment deployment form are quite flexible.

[0020] If the direct acquisition of relevant data is involved, then further hardware and software adaptations are needed for the processing equipment to enable it to acquire data. For example, if real-time acquisition of water spectral data is required, the relevant remote sensing system can be incorporated into the equipment cluster of the processing equipment, or the processing equipment itself can be the control part of the relevant remote sensing system. Alternatively, a third-party call can be used to trigger the relevant remote sensing system outside the processing equipment to perform real-time water spectral data acquisition.

[0021] In addition, if there is a need to display the processing progress (including the processing results), the processing device itself can be configured with the required display screen (including touch screen) to display the specific content. Of course, the processing device can also display the specific content through an external display device or other devices with a display screen.

[0022] The following section introduces the configuration method for the source-introduction model of suspended solids in the Shenzhen Reservoir, which takes into account the source provided in this application.

[0023] First, refer to Figure 1 , Figure 1 This paper illustrates a flowchart of a method for configuring a suspended solids inversion model for the Shenzhen-Dalian Reservoir, taking into account the source information provided in this application. Specifically, the method for configuring a suspended solids inversion model for the Shenzhen-Dalian Reservoir, taking into account the source information provided in this application, may include the following steps S101 to S103: Step S101: For the target deep reservoir, determine the corresponding sensitive spectral characteristics in different water bodies with different sources of suspended matter. Understandably, this application addresses the challenge of continuously changing particulate matter sources and composition, as well as dynamic fluctuations in water optical properties, that often occur in large reservoirs due to various factors (such as upstream water inflow, rainfall, artificial scheduling, and seasonal changes). It starts with water spectral data to construct a better-performing suspended matter inversion model for large reservoirs.

[0024] In this context, this application focuses on the specific sensitive spectral characteristic bands of different suspended solids sources in the Shenzhen Reservoir, which reflect the suspended solids. Therefore, after determining the current Shenzhen Reservoir or the target Shenzhen Reservoir for which the model needs to be constructed, the corresponding sensitive spectral characteristics of different suspended solids sources in the target Shenzhen Reservoir can be determined.

[0025] As an exemplary embodiment, the water bodies with different sources of suspended solids involved in this application can be specifically divided into two types: water bodies with endogenous suspended solids and water bodies with exogenous suspended solids.

[0026] Among them, the exogenous suspended matter water bodies are mainly composed of inorganic particles formed by sediment transport caused by sudden changes in water flow state and hydraulic conditions. Their optical properties are closely related to the absorption process of non-pigment particles. In contrast, the endogenous suspended matter water bodies are mainly composed of algal particles. Their optical properties are directly regulated by the absorption of pigment particles. The optical properties of these two types of particles dominate the spectral response characteristics of the water bodies.

[0027] Of course, in reality, there may be other types of water bodies that contain suspended solids, and these can be classified accordingly. This application will use two types of water bodies, namely endogenous and exogenous suspended solids, to illustrate the subsequent treatment plan.

[0028] In addition, the bands of the sensitive spectral features involved here can be the bands of the particle size / level of the water body from which the suspended matter originates (i.e., determining the bands of the sensitive spectral features for each type of water body from which the suspended matter originates), or they can be determined from an overall perspective to help effectively distinguish the bands of the sensitive spectral features from different water bodies from which the suspended matter originates within the same band range.

[0029] For the latter, as an example, it can be specifically configured as the 650nm-840nm band where there are significant differences in the water reflectance of water bodies with different sources of suspended matter.

[0030] Behind this, the data regarding the spectral band of the sensitive spectral features involved here can be either readily available data (such as data directly configured manually) or it can involve real-time processing. Taking real-time processing as an example, it can include the following processing content: First, we can collect curves of total particulate matter absorption coefficient, algal pigment particulate matter absorption coefficient, and non-pigment particulate matter absorption coefficient in water bodies during different seasons. This allows us to analyze the absorption characteristics of pigment particulate matter, non-pigment particulate matter, and total pigment particulate matter. In this process, we have: 1.1) Water bodies with endogenous suspended solids / water bodies dominated by pigments: Pigment particulate matter absorption characteristics: A distinct absorption peak exists around 675 nm; Absorption coefficient of non-pigment particles: decreases exponentially from 400nm to 900nm.

[0031] 1.2) Water bodies with exogenous suspended solids / Water bodies dominated by non-pigmentary particulate matter: Pigment particulate matter absorption characteristics: exponential decay in the 400nm-900nm range; Absorption coefficient of non-pigment particles: decreases exponentially from 400nm to 900nm.

[0032] Based on the above analysis of the absorption characteristics of pigment particles, non-pigment particles, and total pigment particles, we further combine the water body's water reflectance spectral curve to dynamically identify the spectrally sensitive characteristic bands of the water body. In this process, we have: 2.1) Spectral characteristics of water bodies with endogenous suspended solids / dominating pigments: There is a reflection peak between 500nm and 600nm, and the reflection peak is located near 560nm; A distinct reflection peak exists near 700nm; A distinct reflection peak exists around 800nm.

[0033] 2.2) Spectral characteristics of water bodies with exogenous suspended solids / water bodies dominated by non-pigmentary particulate matter: A distinct reflection peak exists near 600nm; A distinct reflection peak exists around 800nm.

[0034] Based on the above analysis, this application can conclude that water bodies with different suspended solids concentrations exhibit significant differences in water reflectance in the 650nm-840nm wavelength band. This can serve as prior knowledge for water body classification, i.e., it can be used as the wavelength band of sensitive spectral features that will be the focus of subsequent attention.

[0035] Step S102: For water bodies with different sources of suspended solids, the corresponding sensitive spectral characteristics are used as a reference, and K-means clustering analysis based on Manhattan distance strategy is combined to obtain the results of differential measurement of water body spectral characteristics. It is understood that this application is based on the suspended sensitive characteristic band, that is, the band where the sensitive spectral characteristics are determined above. Considering that water bodies with high suspended matter concentration exhibit stronger reflectivity characteristics than those with low suspended matter concentration as they move from the red band to the near-infrared band (650nm-840nm), and since the signal-to-noise ratio of the spectral data acquired by most sensors is poor in the 650nm-840nm range, there are errors in the centroid and cluster distance allocation of each sample data.

[0036] To combat the effects of this noise, this application specifically introduces Manhattan distance as a distance metric for the K-means clustering algorithm. In the K-means clustering process, Manhattan distance can linearly sum the spectral deviations of each dimension, making the contribution of different band dimensions to the total distance more balanced, while maintaining a more dispersed distance distribution in high-dimensional spectral data, thereby efficiently distinguishing different water samples and improving classification accuracy.

[0037] The K-means clustering process combined with the Manhattan distance strategy described here can be understood as serving the analysis and processing of the results of differentiating water spectral characteristics. It can be interpreted as K-means clustering analysis based on the Manhattan distance strategy.

[0038] Below, from the perspective of quantitative formulas, we will provide a more specific and vivid explanation of the specific implementation schemes that can be adopted for K-means clustering analysis based on the Manhattan distance strategy.

[0039] Specifically, the water body spectral characteristic difference measurement results involved in this application may include the following processing: 2.1) For the initial water spectral sample data, the Manhattan distance is used as the distance metric in the clustering process. K-means clustering is performed to obtain the corresponding clustered water spectral sample data. The Manhattan distance is expressed as: , in, Manhattan distance (a non-negative value used in clustering to measure the similarity between samples / sample-cluster centers; the smaller the distance, the more similar they are). For the p-th hyperspectral sample (the p-th sample contains the remote sensing reflectance in the 650nm-840nm band), Let n be the center of the 0th cluster (composed of the mean / median of all hyperspectral samples within the cluster in each band, representing the "representative sample" of the cluster), where n is the total number of bands. Let p be the remote sensing reflectance of the p-th hyperspectral sample in the j-th band. The remote sensing reflectance of the center of the o-th cluster in the j-th band; As can be seen from this, the proposed solution also involves the acquisition and processing of water spectral sample data for clustering purposes. This water spectral sample data can be readily available data that can be extracted locally or online, manually entered data, or data collected in real time as the proposed solution is implemented.

[0040] In addition, it should be noted that the water spectral sample data can be long-term data, that is, water spectral data collected within a certain time range of the target deep reservoir.

[0041] The Manhattan distance, in a 2D plane, is the distance between two points on the vertical axis plus the distance on the horizontal axis, and can be specifically expressed by the above formula.

[0042] 2.1) Based on the clustered water body spectral sample data, further band division is performed within the bands containing the sensitive spectral features to obtain different sub-bands. The sensitivity scores of the two types of water bodies in these sub-bands are then calculated to select the target sensitive spectral feature bands from these sub-bands. The sensitivity score is expressed as: , in, The sensitivity score for the i-th subdivision band (the higher the score, the more sensitive the band is to distinguish water body types). =1,2, This represents the spectral mean of the k-th type of water body in the i-th band (corresponding to the previous example, the first type of water body here can be configured as either endogenous or exogenous suspended solids water body as needed). Let be the global spectral mean of all water body types in the i-th band. The standard deviation of the spectrum of all water samples in the i-th band (normalized to eliminate the influence of band dimensions). 2.2) Based on the clustered water body spectral sample data, calculate the classification index of the two types of water bodies at the target sensitive spectral characteristic bands. The classification index is expressed as: , in, As a classification indicator, The water reflectance of the characteristic reflection peak. The wavelength of the characteristic reflection peak. The water-free reflectance of the characteristic reflection valley. The wavelength of the characteristic reflection valley; 2.3) Based on the clustered water body spectral sample data, target sensitive spectral characteristic bands, and classification indicators, calculate the mean (central tendency) and standard deviation (dispersion) of the intra-class distance between the two water body classes. After verifying the mean and standard deviation of the intra-class distance, continue to calculate the intra-class distance quartiles (the core of the box plot, corresponding to the box plot method). The intra-class distance quartiles include the lower quartiles. (25th percentile after sorting by intra-class distance from smallest to largest), median (50th percentile), upper quartile (75th percentile) and interquartile range (Measures the degree of centralization of the sample distribution) - ; It is understandable that the distribution characteristics of these three aspects are existing indicators, so this application has not elaborated on them in detail.

[0043] The first two distribution features are specifically used to check and evaluate whether there are differences between the data of the two types of water bodies. Plotting curves shows whether the two types of curves overlap under positive and negative standard deviations. If they do not overlap, it means that the clustering effect is good.

[0044] After verification, the corresponding intra-class quartiles are calculated. The resulting intra-class quartiles are used to provide data for the next step (2.4).

[0045] 2.4) For each of the two water body types, based on the intra-class distance quartiles, determine the distance outlier boundary for a single water body type and the distinction threshold between the two water body types, which can be expressed as follows: , , in, This is the boundary value of the first single-type water body distance anomaly. This is the boundary value of the second type of water body distance anomaly. like The threshold for distinguishing between the two types of water bodies (or the minimum threshold for distinguishing them) is expressed as: , in, Thresholds for distinguishing between the two types of water bodies, If the distance distributions of the two types overlap, the threshold for distinguishing the two types of water bodies is expressed as: , If the classification index (i.e., the newly calculated index) of unknown samples at the target sensitive spectral characteristic bands needs to be processed... (greater than) If the water quality meets the criteria, it is classified as Class 2 water body; otherwise, it is classified as Class 1 water body.

[0046] As can be seen, the preceding series of processes serve the purpose of the processing step 2.4) here. The strategy for determining the distinction threshold in the processing step 2.4) here is the core of this application in distinguishing between the two types of water bodies. The resulting water body spectral feature difference measurement result, namely the distance between the two types of water bodies from the outlier boundary and its discrimination mechanism, serves as the key processing logic required for subsequent model configuration.

[0047] Step S103: Based on the differentiating measurement results of the corresponding water spectral characteristics of different suspended solids source water bodies, configure different suspended solids inversion models for deep and large reservoirs. The suspended solids inversion model for deep and large reservoirs is used to perform suspended solids concentration inversion processing on the target suspended solids source water body based on the water spectral data input to the model and the band of the target sensitive spectral characteristics of the target suspended solids source water body.

[0048] After obtaining the differentiated measurement results of the spectral characteristics of water bodies reflecting different sources of suspended solids through the previous step S102, different suspended solids inversion models for deep and large reservoirs can be configured to suit different sources of suspended solids.

[0049] At this point, it can also be combined with Figure 2 The diagram shown illustrates one aspect of the overall design architecture of the proposed solution, providing a more visual understanding of the overall design architecture (technical approach) of the solution mentioned above.

[0050] Furthermore, it can be understood that the model input of the suspended solids inversion model of Shenzhen Reservoir is the corresponding collected water spectral data. During the model operation, it combines the sensitive spectral characteristics of a certain source of suspended solids in a pre-adapted band (the band corresponds to the previous step S101 and is used to constrain the model processing range), and uses the water spectral feature differentiation measurement results determined in step S102 as the key features (corresponding to the case where the training samples are the water spectral feature differentiation measurement results as the key sample content) to carry out suspended solids concentration inversion processing.

[0051] Therefore, considering the complexity and high cost of obtaining inherent optical quantities (including total particulate matter absorption coefficient, algal pigment particulate matter absorption coefficient, and non-pigment particulate matter absorption coefficient), which are difficult to directly use for classifying suspended solids in reservoirs from different sources, the adaptive sensitive feature distance threshold obtained earlier, i.e., the water body spectral feature differentiation measurement result, is used as the core classification basis to achieve efficient and accurate classification of dynamic deep and large reservoir water bodies and their suspended solids inversion. This avoids the constraints of the inherent optical quantity acquisition problem on the classification work. By matching exclusive models for different types of water bodies, the differences in spectral characteristics of various water bodies are fully adapted, effectively improving the accuracy and applicability of suspended solids concentration inversion.

[0052] The core of the suspended solids inversion model configuration work for the deep reservoir is the model training work. The training samples can be water spectral sample data involved in the previous step S102, or they can be brand new samples, or they can be samples that contain water spectral sample data involved in the previous step S102. This is quite flexible.

[0053] The working status of the training samples after the model is put into actual use can be used as a label or as the data content of the training samples themselves to participate in the model training.

[0054] Furthermore, it should be noted that this application does not limit the configured suspended solids inversion model of the deep reservoir. It only uses the differential measurement results of the water spectral characteristics of the source water body of the suspended solids as the only key feature. In the subsequent application of the model in this application, it can be used in conjunction with other model working logic (corresponding to the configuration of the same model) or other inversion models.

[0055] As a practical application example, the measured suspended solids concentration and sensitivity characteristics analysis of water bodies dominated by algal pigments and those dominated by non-pigmented particulate matter include: 1) Water bodies with endogenous suspended solids: The concentration of suspended solids ranged from 1.9 mg / L to 35.6 mg / L, with an average of 8.2 mg / L. A distinct reflection peak exists in the 680nm-705nm remote sensing reflectance range, and the height of the reflection peak is significantly positively correlated with the concentration of suspended matter (organic suspended matter). 2) Water bodies with exogenous suspended solids: The suspended solids concentration ranged from 7.5 mg / L to 69.5 mg / L, with an average of 16.4 mg / L. There are obvious reflection peaks in the remote sensing reflectance at 580nm-610nm and 780-800nm, and the height of the reflection peaks is significantly positively correlated with the concentration of suspended matter (inorganic suspended matter).

[0056] Meanwhile, the configuration of the suspended solids inversion model for the Shenzhen Reservoir involves different model versions at different stages. This application can also perform power function regression, linear regression, and univariate quadratic / quadratic polynomial regression based on different model factors, and conduct comparative analysis through K-fold verification to select the best model for final use, so as to deploy the model version with the best performance in practical applications.

[0057] Correspondingly, the following set of comparative analysis examples can be provided: Table 1 - Examples of Comparative Analysis As can be easily seen from the table, IND is a general term for algorithm factors or model factors, while Rrs in the table is remote sensing reflectance, and Rrs(x) specifically refers to the remote sensing reflectance of the x-band.

[0058] Furthermore, the model configuration process may also involve an evaluation phase. In this regard, during the configuration of the suspended solids inversion model for the Shenzhen Reservoir, this application may also employ evaluation methods including the root mean square error (RMSE), mean absolute percentage error (MAPE), and coefficient of determination (R²). 2 The model performance is evaluated using evaluation metrics such as , and if the requirements are met, the model configuration is completed.

[0059] These evaluation metrics are, understandably, common and mature metrics used in model configuration. Specifically, they include: RMSE represents the sample standard deviation of the difference (residual) between predicted and observed values. It is used to measure the degree of deviation between predicted and observed values. The range of RMSE values ​​is... The smaller the value, the smaller the prediction error and the stronger the predictive ability of the model. MAPE represents the average percentage error between the predicted and actual values. It is an important indicator for measuring the accuracy of predictions. The range of MAPE values ​​is... A smaller MAPE value indicates a more accurate model. Generally, a MAPE of less than 25% is considered a relatively good model.

[0060] R 2 Rm represents the ratio of the sum of the squared differences between predicted and observed values ​​to the variance of the observed values, used to indicate the goodness of fit of the model. 2 The value range of R is 0~1. 2 A value of 0 indicates a poor model fit. 2 A value of 1 indicates that the model has no errors.

[0061] The corresponding calculation formula is as follows: , , , in, These are measured values ​​of suspended solids. is the predicted value of suspended matter, and n is the number of measured suspended matter samples.

[0062] After the model configuration is completed, and if it is easy to understand, it can be put into further practical application. In this regard, the method of this application may also include: Identify the source water body of suspended solids in the target deep reservoir; The suspended solids inversion model of the target deep reservoir corresponding to the source water body of the suspended solids is called to perform suspended solids inversion processing to obtain the suspended solids concentration inversion results of the target deep reservoir.

[0063] It can be noted that, in the case of configuring suspended solids inversion models for deep reservoirs with different water body types according to this application, in the model application work, the water body type of the current target deep reservoir can be determined manually or automatically, that is, the source water body of the suspended solids, and the corresponding target deep reservoir suspended solids inversion model can be called to perform specific suspended solids inversion processing.

[0064] Of course, in specific applications, this application can also further modify the previously configured suspended solids inversion model of deep reservoirs. Based on different suspended solids inversion models of deep reservoirs, a fusion model can be obtained. The front part of the fusion model can perform matching processing on the water spectral data input to the model to match the source water of suspended solids. Based on this, the corresponding matching target suspended solids inversion model of deep reservoirs can be called within the fusion model for further inversion processing.

[0065] Understandably, the ability to automatically identify dynamic water body types directly helps to further ensure the effectiveness of the proposed solution for accurately retrieving suspended solids from dynamic water bodies.

[0066] After obtaining the suspended solids concentration inversion results of the current target deep reservoir, it is easy to understand that, corresponding to the flexible and ever-changing real-time application needs in actual situations, local storage, off-site storage, result display, result forwarding, output completion prompts, or further data analysis and processing can be flexibly configured.

[0067] In conclusion, regarding the above solutions, this application, focusing on the suspended solids inversion target of the Shenzhen Grand Reservoir, addresses the differences in suspended solids inversion between endogenous and exogenous suspended solids in the Shenzhen Grand Reservoir scenario. It introduces K-means clustering analysis combined with Manhattan distance to configure differentiated suspended solids inversion models. This allows for deep adaptation and high-precision suspended solids inversion results for different water bodies, thereby providing excellent data support for water quality monitoring in the Shenzhen Grand Reservoir.

[0068] The above is an introduction to the source-independent suspended solids inversion model configuration method for the Shenzhen-Dalian Reservoir provided in this application. To facilitate better implementation of the source-independent suspended solids inversion model configuration method for the Shenzhen-Dalian Reservoir provided in this application, this application also provides a source-independent suspended solids inversion model configuration device from the perspective of functional modules.

[0069] See Figure 3 , Figure 3This is a schematic diagram of a configuration device for the suspended solids inversion model of the Shenzhen Reservoir, which is based on the source of this application. In this application, the configuration device 300 for the suspended solids inversion model of the Shenzhen Reservoir, which is based on the source of this application, may specifically include the following structure: The determination unit 301 is used to determine the corresponding sensitive spectral characteristics in the wavebands for different suspended solids sources in the target deep reservoir. Analysis unit 302 is used to obtain the differential measurement results of water body spectral characteristics by taking the corresponding sensitive spectral band as a reference and combining K-means clustering analysis based on Manhattan distance strategy for water bodies with different sources of suspended matter. Configuration unit 303 is used to continue to configure different suspended solids inversion models for deep and large reservoirs based on the different measurement results of the corresponding water spectral characteristics of water bodies with different sources of suspended solids. The suspended solids inversion model for deep and large reservoirs is used to perform suspended solids concentration inversion processing on the target suspended solids source water body based on the water spectral data input to the model and in combination with the waveband of the target sensitive spectral characteristics of the target suspended solids source water body.

[0070] As an exemplary embodiment, water bodies with different sources of suspended solids are specifically divided into two types: water bodies with endogenous suspended solids and water bodies with exogenous suspended solids.

[0071] As another exemplary embodiment, the wavelength range of the sensitive spectral features is specifically configured as the 650nm-840nm band, where there are significant differences in the water reflectance of water bodies from different sources of suspended matter.

[0072] As another exemplary embodiment, the results of the water body spectral characteristic difference measurement include the following processing: For the initial water spectral sample data, the Manhattan distance is used as the distance metric in the clustering process. K-means clustering is performed to obtain the corresponding clustered water spectral sample data. The Manhattan distance is expressed as: , in, For Manhattan distance, For the p-th hyperspectral sample, Let n be the center of the o-th cluster, and n be the total number of bands. Let p be the remote sensing reflectance of the p-th hyperspectral sample in the j-th band. The remote sensing reflectance of the center of the o-th cluster in the j-th band; Based on clustered water body spectral sample data, further band division is performed within the bands containing sensitive spectral features to obtain different sub-bands. Sensitivity scores for the two types of water bodies within these sub-bands are then calculated to select target sensitive spectral feature bands from these sub-bands. The sensitivity score is expressed as: , in, The sensitivity score for the i-th subdivision band is... = 1,2, Let be the spectral mean of the k-th type of water body in the i-th band. Let be the global spectral mean of all water body types in the i-th band. Let be the spectral standard deviation of all water samples in the i-th band; Based on the clustered water body spectral sample data, the classification index of the two types of water bodies at the target sensitive spectral characteristic bands is calculated. The classification index is expressed as: , in, As a classification indicator, The water reflectance of the characteristic reflection peak. The wavelength of the characteristic reflection peak. The water-free reflectance of the characteristic reflection valley. The wavelength of the characteristic reflection valley; Based on clustered water body spectral sample data, target sensitive spectral characteristic bands, and classification indices, the mean and standard deviation of the intra-class distance between the two water body classes are calculated. After verification by the mean and standard deviation of the intra-class distance, the intra-class distance quartiles are calculated, including the lower quartiles. , median Upper quartiles and interquartile range , - ; For each of the two water body types, based on the intra-class distance quartiles, the distance outlier boundary for a single water body type and the distinction threshold between the two water body types are determined, respectively, as follows: , , in, This is the boundary value of the first single-type water body distance anomaly. This is the boundary value of the second type of water body distance anomaly. like The threshold for distinguishing between the two types of water bodies is expressed as: , in, Thresholds for distinguishing between the two types of water bodies, If the distance distributions of the two types overlap, the threshold for distinguishing the two types of water bodies is expressed as: , If the classification index of the unknown sample to be processed is greater than the target sensitive spectral characteristic band, If the water quality meets the criteria, it is classified as Class 2 water body; otherwise, it is classified as Class 1 water body.

[0073] As another exemplary embodiment, during the configuration of the suspended matter inversion model for the Shenzhen Reservoir, power function regression, linear regression, and univariate quadratic regression were performed based on different model factors, and comparative analysis was conducted using the K-fold verification method to select the best model for final use.

[0074] As another exemplary embodiment, during the configuration of the suspended matter inversion model for the Shenzhen-Dalian Reservoir, evaluation indicators including root mean square error, mean absolute percentage error, and coefficient of determination are used to evaluate the model performance. If the requirements are met, the model configuration is completed.

[0075] As another exemplary embodiment, the device further includes an application unit 304 for: Identify the source water body of suspended solids in the target deep reservoir; The suspended solids inversion model of the target deep reservoir corresponding to the source water body of the suspended solids is called to perform suspended solids inversion processing to obtain the suspended solids concentration inversion results of the target deep reservoir.

[0076] This application also provides a processing device from a hardware architecture perspective. As mentioned earlier, in practice, a processing device may exist as a device cluster. In this case, each device in the device cluster can also be referred to as a processing device. See [reference needed]. Figure 4 , Figure 4 This diagram illustrates a structural schematic of the processing device of this application. Specifically, the processing device may include a processor 401, a memory 402, and an input / output device 403. The processor 401 executes the computer program stored in the memory 402 to implement, for example... Figure 1 The corresponding embodiments take into account the steps of the deep reservoir suspended matter inversion model configuration method considering the source; or, when processor 401 executes the computer program stored in memory 402, it implements as follows: Figure 3 Corresponding to the functions of each unit in the embodiment, the memory 402 is used to store the functions executed by the processor 401 as described above. Figure 1 The computer program required for configuring the inversion model of suspended matter in deep reservoirs that takes into account the source in the corresponding embodiment.

[0077] For example, a computer program may be divided into one or more modules / units, one or more of which are stored in memory 402 and executed by processor 401 to complete this application. One or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a computer device.

[0078] The processing device may include, but is not limited to, processor 401, memory 402, and input / output device 403. Those skilled in the art will understand that the illustrations are merely examples of the processing device and do not constitute a limitation on the processing device. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, the processing device may also include network access devices, buses, etc., and processor 401, memory 402, input / output device 403, etc., are connected via a bus.

[0079] Processor 401 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the processing device, connecting various parts of the device through various interfaces and lines.

[0080] The memory 402 can be used to store computer programs and / or modules. The processor 401 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 402 and by calling data stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function, etc.; the data storage area may store data created according to the use of the processing device, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device or other volatile solid-state storage device.

[0081] When processor 401 executes a computer program stored in memory 402, it can specifically perform the following functions: For the target deep reservoir, determine the corresponding sensitive spectral characteristics in different water bodies with different sources of suspended solids; For water bodies with different sources of suspended solids, the corresponding sensitive spectral features are used as a reference, and K-means clustering analysis based on Manhattan distance is combined to obtain the results of differential measurement of water body spectral features. Based on the differentiated measurement results of the corresponding water spectral characteristics of water bodies with different sources of suspended matter, different suspended matter inversion models for deep and large reservoirs are configured. Among them, the suspended matter inversion model for deep and large reservoirs is used to perform suspended matter concentration inversion processing on the target suspended matter source water body based on the water spectral data input to the model and combined with the target sensitive spectral characteristics of the target suspended matter source water body in the appropriate band.

[0082] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above description of the configuration device, treatment equipment, and corresponding units of the suspended solids inversion model for the Shenzhen Reservoir, taking into account the source, can be found in the following reference: Figure 1 The specific details of the configuration method for the suspended solids inversion model of the deep reservoir, which takes into account the source, in the corresponding embodiment will not be repeated here.

[0083] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0084] Therefore, this application provides a computer-readable storage medium storing a plurality of instructions that can be loaded by a processor to execute the present application. Figure 1 The steps of the method for configuring the suspended solids inversion model of the deep reservoir that takes into account the source in the corresponding embodiment can be referred to as follows: Figure 1 The description of the configuration method for the suspended solids inversion model of the deep reservoir, which takes into account the source, in the corresponding embodiment will not be repeated here.

[0085] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0086] Because of the instructions stored in the computer-readable storage medium, the present application can be executed as described above. Figure 1 The steps of the method for configuring the suspended solids inversion model of the deep reservoir, which takes into account the source, in the corresponding embodiment can thus achieve the results of this application. Figure 1 The beneficial effects that can be achieved by the configuration method of the suspended solids inversion model of the deep reservoir that takes into account the source in the corresponding embodiment are detailed in the previous description and will not be repeated here.

[0087] The foregoing has provided a detailed description of the configuration method, apparatus, processing equipment, and computer-readable storage medium for the suspended solids inversion model of the Shenzhen Reservoir, taking into account the source. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are merely for the purpose of helping to understand the core ideas of this application; furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for configuring a suspended solids inversion model for deep and large reservoirs that takes into account the source, characterized in that, The method includes: For the target deep reservoir, determine the corresponding sensitive spectral characteristics in different water bodies with different sources of suspended solids; For water bodies with different sources of suspended matter, the corresponding sensitive spectral features are used as a reference, and K-means clustering analysis based on Manhattan distance is combined to obtain the results of differential measurement of water body spectral features. Based on the differentiating measurement results of the spectral characteristics of the water bodies with different sources of suspended matter, different suspended matter inversion models for deep and large reservoirs are configured. The suspended matter inversion model for deep and large reservoirs is used to perform suspended matter concentration inversion processing on the target suspended matter source water body based on the water body spectral data input to the model and in combination with the waveband of the target sensitive spectral characteristics of the target suspended matter source water body.

2. The method according to claim 1, characterized in that, The water bodies with different sources of suspended solids are specifically divided into two types: water bodies with endogenous suspended solids and water bodies with exogenous suspended solids.

3. The method according to claim 2, characterized in that, Specifically, the wavelength range of the sensitive spectral features is configured as the 650nm-840nm band, where there are significant differences in the water reflectance of the different suspended matter sources.

4. The method according to claim 3, characterized in that, The results of the differential measurement of water body spectral characteristics include the following processing: For the initial water spectral sample data, the Manhattan distance is used as the distance metric in the clustering process. K-means clustering is performed to obtain the corresponding clustered water spectral sample data. The Manhattan distance is expressed as: ， in, The Manhattan distance, For the p-th hyperspectral sample, Let n be the center of the o-th cluster, and n be the total number of bands. Let p be the remote sensing reflectance of the p-th hyperspectral sample in the j-th band. The remote sensing reflectance of the o-th cluster center in the j-th band; Based on the clustered water body spectral sample data, the bands within the bands containing the sensitive spectral features are further subdivided to obtain different sub-bands. Sensitivity scores for the two types of water bodies are then calculated within these sub-bands to select target sensitive spectral feature bands from the different sub-bands. The sensitivity score is expressed as: ， in, The sensitivity score for the i-th subdivided band, =1,2, Let be the spectral mean of the k-th water body in the i-th band. The global spectral mean for all water body types in the i-th band is given. Let be the spectral standard deviation of all water samples in the i-th band; Based on the clustered water body spectral sample data, the classification index of the two types of water bodies at the target sensitive spectral characteristic band is calculated, and the classification index is expressed as: ， in, For the classification index, The water reflectance of the characteristic reflection peak. The wavelength of the characteristic reflection peak. The water-free reflectance of the characteristic reflection valley. The wavelength of the characteristic reflection valley; Based on the clustered water body spectral sample data, the target sensitive spectral characteristic bands, and the classification index, the mean and standard deviation of the intra-class distance between the two water body types are calculated. After verification by the mean and standard deviation of the intra-class distance, the intra-class distance quartiles are calculated, including the lower quartiles. , median Upper quartiles and interquartile range , - ; For the two types of water bodies respectively, based on the intra-class distance quartiles, the distance outlier boundary for a single water body and the distinction threshold between the two water bodies are determined, respectively expressed as: ， ， in, This is the boundary value of the first single-type water body distance anomaly. This is the boundary value of the second type of water body distance anomaly. like The threshold for distinguishing between the two types of water bodies is expressed as: ， in, The threshold for distinguishing between the two types of water bodies, If the two distance distributions overlap, the threshold for distinguishing the two water bodies is expressed as follows: ， If the classification index of the unknown sample to be processed at the target sensitive spectral characteristic band is greater than 1 If the water quality meets the criteria, it is classified as Class 2 water body; otherwise, it is classified as Class 1 water body.

5. The method according to claim 1, characterized in that, In the process of configuring the suspended solids inversion model for the Shenzhen Reservoir, power function regression, linear regression, and univariate quadratic regression were also performed based on different model factors. The K-fold validation method was used to conduct comparative analysis in order to select the best model for final selection.

6. The method according to claim 1, characterized in that, During the configuration of the suspended solids inversion model for the Shenzhen-Dalian Reservoir, evaluation indicators including root mean square error, mean absolute percentage error, and coefficient of determination are used to assess the model performance. If the requirements are met, the model configuration is completed.

7. The method according to claim 1, characterized in that, The method further includes: Identify the matching source water body for suspended solids in the target deep reservoir; The suspended matter inversion model of the target deep reservoir corresponding to the source water body of the matched suspended matter is called to perform suspended matter inversion processing to obtain the suspended matter concentration inversion result of the target deep reservoir.

8. A configuration device for a suspended solids inversion model of deep reservoirs that takes into account the source, characterized in that, The device includes: The determination unit is used to determine the corresponding sensitive spectral characteristics in the spectral bands for different suspended solids sources involved in the target deep reservoir. The analysis unit is used to obtain the differential measurement results of the spectral characteristics of water bodies for different sources of suspended matter, taking the corresponding sensitive spectral characteristics as a reference and combining K-means clustering analysis based on the Manhattan distance strategy. The configuration unit is used to configure different suspended solids inversion models for deep and large reservoirs based on the different spectral feature differentiation measurement results of the water bodies with different suspended solids sources. The suspended solids inversion model for deep and large reservoirs is used to perform suspended solids concentration inversion processing on the target suspended solids source water body based on the water body spectral data input to the model and in combination with the waveband of the target sensitive spectral feature of the target suspended solids source water body.

9. A processing device, characterized in that, The method includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the method as described in any one of claims 1 to 7 when it invokes the computer program in the memory.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions adapted for loading by a processor to perform the method of any one of claims 1 to 7.

Images