Water environment monitoring system and method

Abstract

The invention provides a water environment monitoring system and method, and the method comprises the steps: preliminarily screening an abnormal region with an abnormal water environment from all local monitoring regions through the information difference degree of each piece of water spectral data, and extracting the water spectral features of each sensor node in the abnormal region; performing spectrum fusion on all the water body spectrum characteristics according to the topological incidence relation among the sensor nodes in the abnormal area, and further determining the fusion state quantity of water quality pollution in the abnormal area; interference characteristics of the spectrum sensor in the external environment in the abnormal area are obtained, correlation analysis is further carried out on the interference characteristics of the external environment and the fusion state quantity of the water quality pollution, and then the abnormal credibility of the water quality pollution in the abnormal area is determined according to an analysis result. And whether water quality abnormity exists in the water environment monitoring area is judged through the abnormity credibility. By adopting the scheme of the invention, multi-level anomaly analysis of the water environment monitoring area can be realized, so that the reliability of a monitoring result is improved.

Description

Technical Field

[0001] The present application relates to the field of environmental monitoring technology, and more specifically, to a water environment monitoring system and method. Background Art

[0002] Environmental monitoring has laid a technical foundation for responding to environmental risks, achieving green development goals and fulfilling international environmental protection obligations. Smart sensors play a key role in water environment monitoring. Their high sensitivity and high precision enable them to collect key parameters such as water temperature, dissolved oxygen, turbidity, pH value and spectral characteristics in real time, realizing dynamic monitoring and rapid early warning of water quality. By integrating self-diagnosis and wireless data transmission functions, smart sensors have greatly improved monitoring efficiency and data reliability. At the same time, combined with the Internet of Things and big data technologies, they can build a distributed monitoring network to achieve continuous monitoring and comprehensive evaluation of wide-area water environments. The application of smart sensors has significantly promoted the development of water environment monitoring from traditional manual sampling to intelligent and automated directions, providing strong technical support for ensuring water resource security and ecological environmental protection.

[0003] Existing water environment monitoring methods mainly rely on fixed-point sampling and single-point sensor monitoring, which makes it difficult to achieve hierarchical screening and multi-level positioning of anomalies in a large monitoring area. Anomaly analysis usually only stays at the simple judgment of a single data feature, lacking comprehensive consideration of the correlation between data and the global topological relationship. This makes it easy for the identification of abnormal areas to be misjudged or missed when facing complex pollution sources or dynamic water quality changes. In addition, existing technologies are difficult to effectively integrate data from different sensor nodes, lack the means to fuse and analyze the spectral characteristics of water bodies, and cannot deeply explore the multi-dimensional characteristics of the pollution state, resulting in insufficient hierarchy and depth of anomaly analysis. At the same time, the impact of environmental interference factors (such as light, temperature, etc.) on single-point sensor data further aggravates the uncertainty of anomaly judgment, thereby weakening the reliability of monitoring results. Therefore, how to achieve multi-level anomaly analysis in water environment monitoring areas and thus improve the reliability of monitoring results has become a difficult problem faced by the industry. Summary of the invention

[0004] The present application provides a water environment monitoring system and method, which can realize multi-level anomaly analysis of the water environment monitoring area, thereby improving the reliability of the monitoring results.

[0005] In a first aspect, the present application provides a water environment monitoring method based on a smart sensor, comprising the following steps: The water environment monitoring area is divided into multiple local monitoring areas, and a spectral sensor is set at the sensor node of each local monitoring area, and then the spectral sensor is used to collect the water body spectral data of each local monitoring area; Preliminarily screen out abnormal areas with abnormal water environment from all local monitoring areas according to the information difference degree of each water body spectral data, and extract the water body spectral characteristics of each sensor node in the abnormal areas; Determine the topological association relationship between each sensor node in the abnormal area, perform spectral fusion on all water body spectral characteristics according to the topological association relationship to obtain spectral fusion characteristics, and further determine the fusion state quantity of water quality pollution in the abnormal area through the spectral fusion characteristics; Obtain the interference characteristics of the external environment in the abnormal area by the spectral sensor, perform correlation analysis on the interference characteristics of the external environment and the fusion state quantity of water quality pollution, and further determine the abnormal credibility of water quality pollution in the abnormal area according to the analysis result; When the abnormal credibility is greater than the preset abnormal threshold, it is determined that there is water quality abnormality in the water environment monitoring area.

[0006] Preferably, preliminarily screening out abnormal areas with abnormal water environment from all local monitoring areas according to the information difference degree of each water body spectral data specifically includes: For each local monitoring area, determine the spectral variance of each sensor node in the local monitoring area according to the water body spectral data of the local monitoring area; Determine the information difference degree of the water body spectral data corresponding to the local monitoring area through all spectral variances; When the information difference degree is greater than the preset difference degree threshold, determine that the local monitoring area is an abnormal area, and then obtain all abnormal areas, and use all abnormal areas as the abnormal areas of water environment abnormality.

[0007] Preferably, extracting the water body spectral characteristics of each sensor node in the abnormal area specifically includes: Obtain the water body spectral data of the abnormal area; Perform projection clustering on the water body spectral data of the abnormal area to construct a spectral feature space of the water environment; Extract the water body spectral characteristics of each sensor node in the abnormal area from the water body spectral space.

[0008] Preferably, determining the topological association relationship between each sensor node in the abnormal area specifically includes: Obtain the topological structure relationship between each sensor node in the abnormal area; Obtain the historical spectral data of each sensor node in the abnormal area; Perform correlation analysis on the historical spectral data of all sensor nodes according to the topological structure relationship to obtain the topological association relationship between each sensor node.

[0009] Preferably, spectral fusion is performed on all water body spectral features according to the topological association relationship to obtain spectral fusion features, which specifically includes: Determine the information difference values between each water body spectral feature; Determine the fusion weight of each water body spectral feature through all the information difference values and the topological association relationship; Fuse all the water body spectral features according to all the fusion weights to obtain spectral fusion features.

[0010] Preferably, 4 to 8 sensor nodes are arranged in each local monitoring area.

[0011] Preferably, the sampling frequency of the spectral sensor is between 10 and 100 Hz.

[0012] In a second aspect, the present application provides a water environment monitoring system. The water environment monitoring area is divided into multiple local monitoring areas, and spectral sensors are set at the sensor nodes in each local monitoring area. The water environment monitoring system includes: An acquisition module for acquiring the water body spectral data of each local monitoring area through the spectral sensor; A processing module for preliminarily screening out abnormal areas with abnormal water environment from all local monitoring areas according to the information difference degree of each water body spectral data, and extracting the water body spectral features of each sensor node in the abnormal areas; The processing module is further configured to determine the topological association relationship between each sensor node in the abnormal area, perform spectral fusion on all the water body spectral features according to the topological association relationship to obtain spectral fusion features, and then determine the fusion state quantity of water quality pollution in the abnormal area through the spectral fusion features; The processing module is further configured to obtain the interference characteristics of the external environment in the abnormal area by the spectral sensor, perform correlation analysis on the interference characteristics of the external environment and the fusion state quantity of the water quality pollution, and then determine the abnormal credibility of the water quality pollution in the abnormal area according to the analysis result; An execution module for determining that there is water quality abnormality in the water environment monitoring area when the abnormal credibility is greater than a preset abnormal threshold.

[0013] In a third aspect, the present application provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to obtain the code and execute the above-mentioned water environment monitoring method based on intelligent sensors.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned water environment monitoring method based on intelligent sensors is implemented.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: In the embodiments of this application, the water environment monitoring area is divided into multiple local monitoring areas, and spectral sensors are set at the sensor nodes in each local monitoring area. Then, the water body spectral data of each local monitoring area is collected through the spectral sensors; the abnormal areas with abnormal water environment are preliminarily screened from all local monitoring areas according to the information difference degree of each water body spectral data, and the water body spectral characteristics of each sensor node in the abnormal areas are extracted; the topological association relationship between each sensor node in the abnormal areas is determined, and all the water body spectral characteristics are spectrally fused according to the topological association relationship to obtain spectral fusion characteristics. Then, the fusion state quantity of water quality pollution in the abnormal areas is determined through the spectral fusion characteristics; the interference characteristics of the external environment at the spectral sensors in the abnormal areas are obtained, and the interference characteristics of the external environment and the fusion state quantity of water quality pollution are analyzed in association. Then, according to the analysis result, the abnormal credibility of water quality pollution in the abnormal areas is determined; when the abnormal credibility is greater than the preset abnormal threshold, it is determined that there is water quality abnormality in the water environment monitoring area.

[0016] Thus, this application analyzes the interference characteristics of the external environment and the fusion state quantity of water quality pollution in association, and determines the abnormal credibility of water quality pollution in the abnormal areas. Then, according to the abnormal credibility, it is judged whether there is water quality abnormality in the water environment monitoring area; First, the abnormal areas with abnormal water environment are preliminarily screened from all local monitoring areas according to the information difference degree of each water body spectral data. By analyzing the information difference degree of the water body spectral data, the abnormal areas in the water environment can be preliminarily screened, avoiding the limitation of only judging abnormality through a single data feature in the prior art, thus enhancing the hierarchy of abnormal analysis; Then, all the water body spectral characteristics are spectrally fused through the topological association relationship between each sensor node, and then the fusion state quantity of water quality pollution in the abnormal areas is determined. By analyzing the topological relationship between the sensor nodes in the abnormal areas, the fusion of spectral characteristics is implemented, making full use of the spectral information of different nodes, eliminating the information loss and analysis deviation caused by insufficient data integration in the traditional method. It can not only improve the depth of analysis, but also reveal the multi-dimensional characteristics of water quality pollution, and then more accurately reflect the abnormal state of the water environment monitoring area; Finally, by comprehensively analyzing the fusion state quantity of water quality pollution and the environmental interference characteristics, the scheme can quantify the abnormal credibility of water quality pollution in the abnormal areas, thus realizing more accurate abnormal determination. When the abnormal credibility exceeds the preset threshold, the system can timely issue a water quality abnormality warning, reducing the probability of missed detection or misjudgment in the traditional method, thus improving the reliability of the water environment monitoring results; In summary, the solution of this application can realize multi-level abnormal analysis of the water environment monitoring area, thus improving the reliability of the monitoring results. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 FIG. 1 is an exemplary flowchart of a water environment monitoring method based on intelligent sensors according to some embodiments of the present application; Figure 2 FIG. 2 is a schematic structural diagram of the distribution of sensor nodes in a local monitoring area according to some embodiments of the present application; Figure 3 FIG. 3 is a schematic flowchart of determining a topological association relationship according to some embodiments of the present application; Figure 4 FIG. 4 is a schematic structural diagram of a water environment monitoring system according to some embodiments of the present application; Figure 5 FIG. 5 is a schematic structural diagram of a computer device for implementing a water environment monitoring method based on intelligent sensors according to some embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] In order to better understand the technical solutions of the present application, the technical solutions of the present application will be described in detail below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0019] Referring to Figure 1 FIG. 1, which is an exemplary flowchart of a water environment monitoring method based on intelligent sensors according to some embodiments of the present application, the water environment monitoring method 100 based on intelligent sensors mainly includes the following steps: In step 101, the water environment monitoring area is divided into a plurality of local monitoring areas, and a spectral sensor is set at the sensor node of each local monitoring area, and then the water body spectral data of each local monitoring area is collected through the spectral sensor.

[0020] Specifically, first, the water environment monitoring area can be divided into a plurality of local areas based on the geographical distribution information of the target water area, and each local area is used as a local monitoring area; then, referring to Figure 2 FIG. 2, which is a schematic structural diagram of the distribution of sensor nodes in a local monitoring area in some embodiments of the present application, 4 to 8 sensor nodes are arranged in each local monitoring area, and a spectral sensor is set at each sensor node. The sampling frequency of the spectral sensor is between 10 and 100 Hz. Further, the spectral data of the water body at each sensor node in each local monitoring area can be collected through the spectral sensor, and the set of all spectral data of the water body collected through the spectral sensor in each local monitoring area is used as the water body spectral data of the corresponding local monitoring area. It should be noted that the water body spectral data refers to the information of the reflected light or absorbed light of the water body under the irradiation of light with different wavelengths collected by the spectral sensor.

[0021] It should be noted that in this application, a large - scale water environment monitoring area is divided into multiple local monitoring areas, which can monitor and identify the pollution status of specific local monitoring in real time, thereby reducing the error of overall monitoring.

[0022] In step 102, based on the information difference degree of each water body spectral data, abnormal areas with abnormal water environment are preliminarily screened out from all local monitoring areas, and the water body spectral characteristics of each sensor node in the abnormal areas are extracted.

[0023] In some embodiments, the preliminary screening of abnormal areas with abnormal water environment from all local monitoring areas based on the information difference degree of each water body spectral data can be achieved by the following steps: For each local monitoring area, determine the spectral variance of each sensor node in the local monitoring area according to the water body spectral data of the local monitoring area; Determine the information difference degree of the water body spectral data corresponding to the local monitoring area through all the spectral variances; When the information difference degree is greater than the preset difference degree threshold, determine that the local monitoring area is an abnormal area, and then obtain all the abnormal areas, and take all the abnormal areas as the abnormal areas with abnormal water environment.

[0024] Specifically, when implementing, for each local monitoring area, first separate the spectral data of the water body at each sensor node (each sensor node is a sensor node in the local monitoring area) from the water body spectral data of the local monitoring area, then substitute the difference between each spectral data and the preset reference spectrum into the variance calculation formula, and take the calculation result as the spectral variance of the corresponding sensor node. Among them, the reference spectrum is the reference value for evaluating whether the water body spectral data is abnormal, and it is usually the average value of the water body spectral data in the normal state of the water environment. The spectral variance represents the variance value of the difference between the spectral data and the reference spectrum; then, the average value of the spectral variances of all sensor nodes can be used as the information difference degree of the water body spectral data corresponding to the local monitoring area; finally, compare the information difference degree with the preset difference degree threshold. Among them, the preset difference degree threshold can be determined according to historical spectral data. In the embodiments of this application, the preset difference degree threshold is the average value of the variances of historical spectral data. When the information difference degree is greater than the preset difference degree threshold, determine the local monitoring area as an abnormal area. Through the above method, all the abnormal areas can be obtained, and all the abnormal areas are taken as the abnormal areas with abnormal water environment.

[0025] It should be noted that the information difference degree in this application is an index for measuring the difference degree of water body spectral data. By quantifying the information difference degree of water body spectral data, the difference degree of the spectral data collected between each sensor node in the local monitoring area can be reflected.

[0026] In some embodiments, in some embodiments, the water body spectral features of each sensor node in the abnormal area can be extracted by the following steps: Obtain the water body spectral data of the abnormal area; Perform projection clustering on the water body spectral data of the abnormal area to construct a spectral feature space of the water body environment; Extract the water body spectral features of each sensor node in the abnormal area from the water body spectral space.

[0027] It should be noted that the water body spectral features in this application refer to the spectral performance characteristics presented during the processes of light absorption, reflection, and scattering by different water body components. It is the optical response characteristic of the water body components, including absorption peaks, reflection peaks at specific wavelengths, and the overall shape of the spectral curve, and is used to identify the substance composition and state changes in the water body.

[0028] In specific implementation, first, collect the spectral data of the water body at all sensor nodes in the abnormal area, and use the spectral data of the water body at all sensor nodes as the water body spectral data of the abnormal area; then, use principal component analysis to project the high-dimensional water body spectral data into a low-dimensional space, thereby reducing the redundant information in the water body spectral data. Further, in the projected low-dimensional space, apply a clustering algorithm (such as K-means) to group the water body spectral data according to the Euclidean distance between spectral features (wavelengths), obtaining spectral data groups in different wavelength ranges. Then, screen out all spectral data groups with representative water quality characteristics (i.e., spectral bands with wavelengths between 700 and 2500 nm) from the spectral data groups in different wavelength ranges, and use the data segment composed of all the screened spectral data groups as the spectral data segment of the water body. For example, if the value range of the spectral characteristic band of the water quality characteristics is between 700 and 2500 nm, the characteristic bands with spectra between 700 and 2500 nm can be screened out from all spectral data groups, and the space composed of all the screened spectral characteristic bands is used as the spectral feature space of the water body. Finally, select the spectral characteristic bands corresponding to each sensor node from the water body spectral space, and extract the water body spectral features of the corresponding sensor nodes from the spectral characteristic bands. The water body spectral features refer to the spectral performance characteristics presented during the processes of light absorption, reflection, and scattering by different water body components.

[0029] In step 103, determine the topological association relationship between each sensor node in the abnormal area, perform spectral fusion on all the water body spectral features according to the topological association relationship to obtain spectral fusion features, and further determine the fusion state quantity of water quality pollution in the abnormal area through the spectral fusion features.

[0030] In some embodiments, refer to Figure 3As shown in the figure, this is a schematic flowchart for determining the topological association relationship in some embodiments of the present application. In this embodiment, the topological association relationship between each sensor node in the abnormal area can be implemented by the following steps: In step 1031, obtain the topological structure relationship between each sensor node in the abnormal area; In step 1032, obtain the historical spectral data of each sensor node in the abnormal area; In step 1033, perform correlation analysis on the historical spectral data of all sensor nodes according to the topological structure relationship to obtain the topological association relationship between each sensor node.

[0031] Specifically, first, obtain the initial topological structure relationship between each sensor node through the deployment information of the sensor network (such as physical location or communication network structure), which can be represented by a spatial adjacency matrix or a network connection table; then, collect the historical spectral data of each sensor node and perform time series analysis on these data to extract the correlation between the spectral feature changes of different sensor nodes; finally, correlate the spectral data between sensor nodes according to the topological structure relationship, use a correlation analysis method (such as Pearson correlation coefficient) to evaluate the similarity of spectral changes between each two sensor nodes, and further use the similarity of spectral changes between each two sensor nodes as the adjustment coefficient of the topological structure between these two sensor nodes, so that the similarity can be multiplied by the topological structure to obtain the topological association value between these two sensor nodes. Through the above method, the topological association value between each two sensor nodes can be obtained, and then the topological relationship structure composed of all topological association values is used as the topological association relationship between each sensor node.

[0032] It should be noted that the topological association relationship in the present application is an index reflecting the spatial adjacency relationship between sensor nodes.

[0033] In some embodiments, the spectral fusion of all water body spectral features according to the topological association relationship to obtain the spectral fusion feature can be implemented by the following steps: Determine the information difference value between each water body spectral feature; Determine the fusion weight of each water body spectral feature through all the information difference values and the topological association relationship; Fuse all the water body spectral features according to all the fusion weights to obtain the spectral fusion feature.

[0034] In specific implementation, first, select a sensor node as the selected sensor node, and calculate the information difference value of the water body spectral characteristics between the selected sensor node and each other sensor node. The information difference value can be represented by the Euclidean distance between the water body spectral characteristics, which will not be elaborated here. Further, obtain the topological association value between the selected sensor node and each other sensor node from the topological association relationship. The topological association value is specifically explained in the step of determining the topological association relationship between each sensor node, which will not be elaborated here. Then, take the product of the average value of all information difference values and the average value of all topological association values as the fusion weight corresponding to the water body spectral characteristics of the selected sensor node. Repeat the above steps to obtain the fusion weights corresponding to the water body spectral characteristics of the remaining sensor nodes, and then the weighted fusion of all water body spectral characteristics can be performed according to each fusion weight, and the fusion result is used as the spectral fusion feature.

[0035] It should be noted that the spectral fusion feature in this application is a fusion feature that can characterize the optical properties and spatial distribution of different components in the water body.

[0036] In addition, it should be noted that determining the fusion state quantity of water quality pollution in the abnormal area through the spectral fusion feature means taking the water body spectral fusion feature as the input, and inversely evaluating the water environment state of the abnormal area through a pre-trained water quality inversion model to obtain the fusion state quantity of water quality pollution in the abnormal area. In specific implementation, a large number of experimental data sets can be collected first, and the data in the experimental data sets are labeled. When labeling, the water body spectral data can be corresponded to the corresponding water quality indicators (such as COD, ammonia nitrogen concentration, and chlorophyll-a concentration). Then, use machine learning or deep learning algorithms (such as random forest, support vector machine, or neural network) to learn the relationship between the water body spectral data and water quality indicators in the experimental data set and establish the mapping relationship between the spectral features and water quality parameters. Through multiple iterations to optimize the model parameters, it can accurately capture the non-linear relationship between the spectral features and water quality changes and achieve a high prediction accuracy on the validation set. Through the above method, the pre-training of the water quality inversion model can be completed. Further, input the spectral fusion feature into the pre-trained water quality inversion model, and the water environment state of the abnormal area can be inversely evaluated through the water quality inversion model, and the output result is used as the fusion state quantity of water quality pollution in the abnormal area.

[0037] It should be noted that the fusion state quantity of water quality pollution in this application is a fusion index for measuring the overall water quality pollution degree of the abnormal area. Through the fusion state quantity of water quality pollution, the uncertainty of the data of a single sensor node can be effectively reduced, providing a reliable basic support for the accurate diagnosis and pollution source tracing of water quality pollution.

[0038] In step 104, obtain the interference characteristics of the external environment in the abnormal area by the spectral sensor, perform correlation analysis on the interference characteristics of the external environment and the fusion state quantity of the water quality pollution, and then determine the abnormal credibility of the water quality pollution in the abnormal area according to the analysis result.

[0039] In some embodiments, obtaining the interference characteristics of the external environment in the abnormal area by the spectral sensor can be implemented by the following steps: Obtain all environmental interference factors outside the spectral sensor in water environment monitoring; Determine the interference amount of each environmental interference factor through pre-collected test data; Determine the interference characteristics of the external environment in the abnormal area by the spectral sensor according to all the interference amounts.

[0040] In specific implementation, first, obtain the environmental interference factors affecting the normal operation of the spectral sensor by referring to literature materials, such as light change, temperature fluctuation, water turbidity, and particulate suspension; then, use the spectral sensor to conduct tests under various environmental conditions (different light conditions, different temperature conditions, different water turbidity conditions, and different particulate suspension conditions), record the test data, and then analyze the interference amount received by the spectral sensor under different environmental interference factors through test data analysis. For example, statistically analyze the difference between the spectral data and the reference spectral data under different light conditions, and use the difference as the interference amount under the corresponding light condition. It should be noted that the reference spectral data is the spectral data collected under normal light conditions; finally, perform principal component analysis on the interference amounts of each environmental interference factor, screen out the representative environmental interference factors, and use the interference amounts of the screened environmental interference factors as the interference characteristics of the external environment in the abnormal area. The interference characteristics can describe the degree of environmental interference received by the spectral sensor in the abnormal area and provide a basis for subsequent abnormal analysis of spectral data.

[0041] In some embodiments, performing correlation analysis on the interference characteristics of the external environment and the fusion state quantity of the water quality pollution, and then determining the abnormal credibility of the water quality pollution in the abnormal area according to the analysis result can be implemented by the following steps: Obtain all environmental interference factors outside the spectral sensor in water environment monitoring; For each environmental interference factor, extract the interference amount of the environmental interference factor on the spectral sensor from the interference characteristics of the external environment; Determine the mapping relationship between the water quality pollution and the environmental interference factors in the abnormal area according to the interference amount and the water body spectral data in the abnormal area, and then obtain the mapping relationship between the water quality pollution and each environmental interference factor in the abnormal area; Associate and evaluate the water quality pollution degree in the abnormal area through all the mapping relationships and the fusion state quantity of the water quality pollution, and obtain the abnormal credibility of the water quality pollution in the abnormal area.

[0042] In specific implementation, first, obtain the environmental interference factors affecting the normal operation of the spectral sensor by consulting literature and materials, such as light changes, temperature fluctuations, water turbidity, and particulate suspension; second, obtain the interference amount of the environmental interference factors on the spectral data from the interference characteristics of the external environment; then, combine the interference amount with the fusion state quantity of the water quality pollution, and learn the mapping relationship between the interference amount and the water quality pollution by constructing a multi-dimensional regression model. The mapping relationship can reflect the influence degree of the environmental interference factors on the water quality pollution assessment. It should be noted that the multi-dimensional regression model can be specifically trained through the following steps, that is: the first step, data preparation, collect a multi-dimensional data set including water body spectral data, the fusion state quantity of the water quality pollution (such as pollutant concentration), and environmental interference factors (such as light intensity, temperature), and clean and normalize the multi-dimensional data set; the second step, feature extraction and annotation, use the interference amount of the environmental interference factors on the spectral sensor and the water body spectral data in the abnormal area as feature variables, and at the same time use the water quality index of the water quality pollution as the target variable to provide the input and output relationship for the model, and label the relationship between the input and output as the mapping relationship between the water quality pollution and each environmental interference factor; the third step, model selection and initialization, can select the support vector regression algorithm as the model framework, set hyperparameters and initialize the model according to the complexity of the spectral data (the complexity can be described by variance) and the non-linear degree (the non-linear degree can be described by the second derivative); the fourth step, training and optimization, use the training data to iteratively train the model, evaluate the model performance through an error function (such as mean square error or absolute error), and adjust the model parameters through optimization methods such as gradient descent and regularization (such as L1 or L2) to reduce overfitting and improve the generalization ability; the fifth step, verification and testing, evaluate the trained model on an independent validation set and a test set, evaluate the model accuracy by comparing the error between the predicted value and the true value, and confirm the stability of the model through cross-validation; then, output the water quality index of the water quality pollution through the multi-dimensional regression model, and use the absolute difference between the water quality index and the fusion state quantity of the water quality pollution as the degree factor of the environmental interference. Finally, the natural exponential function value of the reciprocal of the degree factor can be used as the abnormal credibility of the water quality pollution in the abnormal area, where the value range of the abnormal credibility is between 0 and 1. The greater the abnormal credibility, the greater the probability of water quality pollution in the abnormal area, rather than a misjudgment caused by the sensor being affected by external environmental changes.

[0043] It should be noted that the abnormal credibility of water quality pollution in this application is a reliability index for measuring water quality anomalies in the water environment monitoring area. A higher abnormal credibility indicates a greater probability of water quality pollution in the abnormal area, rather than an abnormal judgment caused by the sensor being interfered by the external environment. A lower abnormal credibility indicates that the spectral data in the abnormal area may be a spectral anomaly caused by environmental interference.

[0044] In step 105, when the abnormal credibility is greater than a preset abnormal threshold, it is determined that there is water quality abnormality in the water environment monitoring area.

[0045] Specifically, first, a preset abnormal threshold is set, which represents the sensitivity limit of water quality abnormality in the monitoring data. Secondly, during the water environment monitoring process, if the calculated result of the abnormal credibility of water quality pollution is greater than this abnormal threshold, it indicates that the water quality in the abnormal area has been significantly affected by pollution, and the abnormal area is judged as a water quality abnormal area. Then, by combining the spatial information of the abnormal area and the water body spectral data, using spatial positioning algorithms (such as hotspot analysis, regional clustering algorithm or geographic information system GIS), the precise location of the abnormal area is obtained, and the specific monitoring areas with more serious pollution or interference are identified. Finally, the location information and pollution type of the abnormal area are fed back to the management personnel or the system to support further water quality restoration, pollution source tracing and regional governance work.

[0046] On the other hand, in some embodiments, this application provides a water environment monitoring system. The water environment monitoring area is divided into multiple local monitoring areas, and spectral sensors are set at the sensor nodes of each local monitoring area. Refer to Figure 4 , this figure is a schematic structural diagram of the water environment monitoring system shown in some embodiments of this application. The water environment monitoring system 400 includes: an acquisition module 401, a processing module 402, and an execution module 403, which are described as follows: Acquisition module 401. In this application, the acquisition module 401 mainly acquires the water body spectral data of each local monitoring area through spectral sensors. Processing module 402. In this application, the processing module 402 is used to preliminarily screen out the abnormal areas with water environment abnormalities from all local monitoring areas according to the information difference degree of each water body spectral data, and extract the water body spectral characteristics of each sensor node in the abnormal areas. In this application, the processing module 402 is further used to determine the topological association relationship between each sensor node in the abnormal area, perform spectral fusion on all water body spectral characteristics according to the topological association relationship to obtain spectral fusion characteristics, and then determine the fusion state quantity of water quality pollution in the abnormal area through the spectral fusion characteristics. In this application, the processing module 402 is further configured to obtain the interference characteristics of the external environment in the abnormal area by the spectral sensor, perform correlation analysis on the interference characteristics of the external environment and the fusion state quantity of the water quality pollution, and then determine the abnormal credibility of the water quality pollution in the abnormal area according to the analysis result. Execution module 403. In this application, the execution module 403 is mainly configured to determine that there is water quality abnormality in the water environment monitoring area when the abnormal credibility is greater than a preset abnormal threshold.

[0047] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to obtain the code and execute the above-mentioned water environment monitoring method based on intelligent sensors.

[0048] In some embodiments, refer to Figure 5 , this figure is a schematic structural diagram of a computer device for implementing the water environment monitoring method based on intelligent sensors according to some embodiments of this application. The water environment monitoring method based on intelligent sensors in the above embodiments can be implemented by Figure 5 the computer device shown. The computer device 500 includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0049] The processor 501 can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0050] The communication bus 502 can be used to transmit information between the above components.

[0051] The memory 503 can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM), or other types of dynamic storage devices that can store information and instructions. It can also be an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 503 can exist independently and be connected to the processor 501 through the communication bus 502. The memory 503 can also be integrated with the processor 501.

[0052] Among them, the memory 503 is used to store the program code for executing the solution of this application, and is controlled by the processor 501 for execution. The processor 501 is used to execute the program code stored in the memory 503. The program code can include one or more software modules. The above-mentioned water environment monitoring method based on intelligent sensors in the embodiment can be implemented by one or more software modules in the program code in the processor 501 and the memory 503.

[0053] The communication interface 504 uses any device such as a transceiver to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0054] In a specific implementation, as an embodiment, the computer device can include multiple processors, and each of these processors can be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, the processor can refer to one or more devices, circuits, and / or processing cores for processing data (such as computer program instructions).

[0055] The above computer device may be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device may be a desktop computer, a laptop computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. The embodiments of the present application do not limit the type of the computer device.

[0056] In addition, the present application also provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the above water environment monitoring method based on intelligent sensors is implemented.

[0057] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present application.

[0058] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. A water environment monitoring method based on intelligent sensors, characterized in that: The steps include: The water environment monitoring area is divided into multiple local monitoring areas, and a spectral sensor is set at the sensor node of each local monitoring area, and then the spectral sensor is used to collect the water body spectral data of each local monitoring area; According to the information difference of each water body spectral data, the abnormal areas with abnormal water environment are preliminarily screened out from all local monitoring areas, and the water body spectral characteristics of each sensor node in the abnormal area are extracted; Determine the topological association relationship between each sensor node in the abnormal area, perform spectral fusion on all water spectral features according to the topological association relationship to obtain spectral fusion features, and then determine the fusion state quantity of water quality pollution in the abnormal area through the spectral fusion features; Obtaining interference features of the external environment of the spectral sensor in the abnormal area, performing correlation analysis on the interference features of the external environment and the fusion state quantity of the water pollution, and then determining the abnormal credibility of the water pollution in the abnormal area according to the analysis results; When the abnormality credibility is greater than a preset abnormality threshold, it is determined that there is water quality abnormality in the water environment monitoring area.

2. The method according to claim 1, characterized in that Based on the information difference of each water body spectral data, the abnormal areas with abnormal water environment are initially screened out from all local monitoring areas, including: For each local monitoring area, the spectral variance of each sensor node in the local monitoring area is determined according to the water spectral data of the local monitoring area; The information difference of the water body spectral data corresponding to the local monitoring area is determined through all spectral variances; When the information difference is greater than a preset difference threshold, the local monitoring area is determined to be an abnormal area, and then all abnormal areas are obtained and regarded as abnormal areas of water environment abnormality.

3. The method according to claim 1, characterized in that Extracting the water spectral characteristics of each sensor node in the abnormal area specifically includes: Obtain water spectral data in abnormal areas; Project and cluster the water spectral data in the abnormal area to construct the spectral feature space of the water environment; The water body spectral features of each sensor node in the abnormal area are extracted from the water body spectral space.

4. The method according to claim 1, characterized in that Determining the topological association relationship between the sensor nodes in the abnormal area specifically includes: Obtain the topological structure relationship between each sensor node in the abnormal area; Obtain historical spectral data of each sensor node in the abnormal area; According to the topological structure relationship, the historical spectral data of all sensor nodes are subjected to association analysis to obtain the topological association relationship between the sensor nodes.

5. The method according to claim 1, characterized in that According to the topological association relationship, all water spectral features are spectrally fused to obtain spectral fusion features including: Determine the information difference value between the spectral characteristics of each water body; Determine the fusion weight of each water body spectral feature through all information difference values ​​and the topological association relationship; All water spectral features are fused according to all fusion weights to obtain spectral fusion features.

6. The method according to claim 1, characterized in that 4 to 8 sensor nodes are deployed in each local monitoring area.

7. The method according to claim 1, characterized in that The sampling frequency of the spectrum sensor is between 10 and 100 Hz.

8. A water environment monitoring system, characterized in that: The water environment monitoring area is divided into multiple local monitoring areas, and a spectral sensor is set at the sensor node of each local monitoring area. The water environment monitoring system includes: A collection module, used to collect water spectral data of each local monitoring area through a spectral sensor; A processing module is used to preliminarily screen out abnormal areas with abnormal water environment from all local monitoring areas according to the information difference of each water body spectral data, and extract the water body spectral characteristics of each sensor node in the abnormal area; The processing module is further used to determine the topological association relationship between each sensor node in the abnormal area, perform spectral fusion on all water spectral features according to the topological association relationship to obtain spectral fusion features, and then determine the fusion state quantity of water pollution in the abnormal area through the spectral fusion features; The processing module is further used to obtain interference characteristics of the external environment of the spectral sensor in the abnormal area, perform correlation analysis on the interference characteristics of the external environment and the fusion state quantity of the water pollution, and then determine the abnormal credibility of the water pollution in the abnormal area according to the analysis result; The execution module is used to determine that there is water quality abnormality in the water environment monitoring area when the abnormality credibility is greater than a preset abnormality threshold.

9. A computer device, comprising a memory and a processor, wherein the memory stores a code, characterized in that: The processor is configured to obtain the code and execute the water environment monitoring method based on the smart sensor according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the water environment monitoring method based on the smart sensor as described in any one of claims 1 to 7 is implemented.

Images