Internet of Things intelligent monitoring alarm method and system for water quality detection

Abstract

The invention discloses an Internet of Things intelligent monitoring alarm method and system for water quality detection, and relates to the technical field of environmental protection monitoring, and the method comprises the steps: respectively sampling water quality electrochemical data for corresponding monitoring water area blocks based on each Internet of Things node; extracting electrochemical time sequence sensing characteristics respectively corresponding to each piece of water quality electrochemical data; extracting mode time sequence dependence features corresponding to the electrochemical time sequence sensing features based on a water quality anomaly identification model to predict water quality anomaly risks; identifying a corresponding water quality abnormity risk event type based on an abnormal event identification model; and according to the block identifier of each risk monitoring water area block and the corresponding water quality abnormity risk event type, executing water area water quality pollution alarm operation. Therefore, based on distributed water quality monitoring, intelligent risk identification and water quality pollution early warning operation of the Internet of Things nodes, water quality monitoring real-time performance, comprehensive coverage and intelligent management capability are remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of environmental monitoring technology, and in particular to an Internet of Things intelligent monitoring and alarm method and system for water quality detection. Background Art

[0002] With the rapid development of industrialization and urbanization, water pollution is becoming increasingly serious, posing a serious threat to the safety of global water resources and the ecological environment. As a key link in water pollution prevention and control, water quality monitoring is of great significance for ensuring the ecological health of water bodies and the safety of drinking water for residents.

[0003] Traditional water quality monitoring methods mainly rely on manual sampling and laboratory analysis. Usually, water quality monitoring personnel need to regularly go to the target waters to take manual samples and send the samples to the laboratory for chemical analysis and testing.

[0004] However, the process of manual sampling and laboratory analysis is time-consuming, usually requiring several hours or even days to obtain results. For dynamically changing water quality environments, especially sudden water pollution incidents (such as toxic chemical leaks, domestic sewage discharges, etc.), this lagging monitoring method is difficult to detect and respond in a timely manner, and it is easy to miss the best time to deal with pollution incidents. In addition, the manual sampling method is limited by the actual operating capabilities of the monitoring personnel and the geographical location. It can often only monitor specific points, making it difficult to achieve comprehensive coverage of the entire water area, and often leads to a large amount of human and material resources.

[0005] In response to the above problems, the industry has not yet proposed a better technical solution. Summary of the invention

[0006] The present application provides an Internet of Things intelligent monitoring and alarm method, system, storage medium, computer program product and electronic device for water quality detection, which are used to at least solve the problems of insufficient monitoring timeliness and water area coverage of water quality monitoring technology in current related technologies.

[0007] In a first aspect, an embodiment of the present application provides an IoT intelligent monitoring and alarm method for water quality detection, comprising: sampling water quality electrochemical data for corresponding monitored water areas based on each IoT node; multiple IoT nodes are distributedly arranged in different blocks in the monitored water area, and each IoT node is deployed with an electrochemical sensor; the sensing parameter types of the water quality electrochemical data include pH value, DO content and TOC content of the water body; extracting the electrochemical time series sensing features corresponding to each of the water quality electrochemical data; inputting each of the electrochemical time series sensing features into a water quality anomaly recognition model, extracting the pattern time series dependency features corresponding to the electrochemical time series sensing features, so as to predict whether there is a water quality anomaly risk in the corresponding monitored water area block; for each risk monitoring water area block predicted to have a water quality anomaly risk, inputting the pattern time series dependency features corresponding to the risk monitoring water area block into an abnormal event recognition model, so as to identify the water quality anomaly risk event type corresponding to the risk monitoring water area block; and performing a water quality pollution alarm operation according to the block identifier of each of the risk monitoring water area blocks and the corresponding water quality anomaly risk event type.

[0008] In the second aspect, the embodiment of the present application provides an IoT intelligent monitoring and alarm system for water quality detection, including: a water quality parameter sampling unit, which is used to sample water quality electrochemical data for corresponding monitored water areas based on each IoT node; multiple IoT nodes are distributedly arranged in different blocks in the monitored water area, and each of the IoT nodes is deployed with an electrochemical sensor; the sensing parameter types of the water quality electrochemical data include pH value, DO content and TOC content of the water body; a time series feature extraction unit, which is used to extract the electrochemical time series sensing features corresponding to each of the water quality electrochemical data; a water quality anomaly identification unit, which is used to identify each of the electrochemical time series sensing features; The sequence sensing features are respectively input into the water quality abnormality recognition model, and the pattern timing dependence features corresponding to the electrochemical time series sensing features are extracted to predict whether the corresponding monitored water area block has water quality abnormality risk; the abnormal event recognition unit is used to input the pattern timing dependence features corresponding to the risk monitoring water area block into the abnormal event recognition model for each risk monitoring water area block predicted to have water quality abnormality risk, so as to identify the water quality abnormality risk event type corresponding to the risk monitoring water area block; the water area water quality alarm unit is used to perform water quality pollution alarm operation according to the block identification of each risk monitoring water area block and the corresponding water quality abnormality risk event type.

[0009] According to a third aspect, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the steps of the IoT intelligent monitoring and alarm method for water quality detection according to any embodiment of the present application.

[0010] In a fourth aspect, an embodiment of the present application provides a storage medium on which a computer program is stored, characterized in that when the program is executed by a processor, the steps of the IoT intelligent monitoring and alarm method for water quality detection of any embodiment of the present application are implemented.

[0011] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the IoT intelligent monitoring and alarm method for water quality detection according to any embodiment of the present application.

[0012] The IoT intelligent monitoring and alarm method and system for water quality detection provided by this application can produce at least the following technical effects:

[0013] (1) By setting up multiple IoT nodes in the monitored water area, each node deploys an electrochemical sensor to collect electrochemical data of water quality in the monitored water area in real time, which can achieve continuous and real-time water quality data collection and transmission, so that the data processing center can obtain information on changes in water quality in the water area in a timely manner and respond quickly, effectively avoiding missing the best time to handle pollution incidents due to monitoring lags. In addition, multiple IoT nodes are deployed in different blocks of the water area, achieving fine-grained water quality monitoring coverage of each block of the water area, and can accurately locate the specific location where water quality anomalies occur, making it easier for water quality management departments to locate and track pollution sources, thereby improving the level of refined water quality management.

[0014] (2) By constructing a water quality anomaly recognition model and an abnormal event recognition model, the electrochemical time series sensing features collected by different nodes can be intelligently analyzed and pattern recognized. On the one hand, the water quality anomaly recognition model can predict the water quality anomaly risk of each monitored water area based on the time series characteristics of the electrochemical sensing data; on the other hand, the abnormal event recognition model can further identify the specific water quality anomaly event type (such as chemical pollution, excessive organic matter, heavy metal pollution, etc.) based on the pattern time series dependency characteristics of the water quality anomaly risk area, thereby achieving accurate classification of water quality anomaly events.

[0015] (3) After identifying the abnormal water quality risks and their event types, the corresponding water quality pollution alarm operation is automatically triggered according to the location and event type of each risk monitoring block, realizing a response mechanism based on event type, which can help water quality management departments quickly locate the source of pollution and formulate corresponding treatment measures, thereby controlling the spread of pollution in the shortest time and reducing the harm to the ecological environment of the water area.

[0016] Through this technical solution, through distributed water quality monitoring, real-time data collection, intelligent risk identification and water pollution warning operations based on IoT nodes, the real-time, comprehensive coverage and intelligent management capabilities of water quality monitoring have been significantly improved. It can effectively respond to the dynamically changing water quality environment, improve the detection, positioning and response capabilities of water pollution incidents, and help to ensure the ecological health of water bodies. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A flowchart of an example of an IoT intelligent monitoring and alarm method for water quality detection according to an embodiment of the present application is shown;

[0019] Figure 2 An operation flow chart of an example of extracting electrochemical time-series sensing features corresponding to water quality electrochemical data according to an embodiment of the present application is shown;

[0020] Figure 3 A schematic diagram showing a structural connection of an example of a water quality anomaly recognition model according to an embodiment of the present application is shown;

[0021] Figure 4 A schematic diagram showing the structural connection of another example of a water quality anomaly recognition model according to an embodiment of the present application is shown;

[0022] Figure 5 A schematic diagram showing a structural connection of an example of an abnormal event recognition model according to an embodiment of the present application is shown;

[0023] Figure 6 A structural block diagram of an example of an Internet of Things intelligent monitoring and alarm system for water quality detection according to an embodiment of the present application is shown;

[0024] Figure 7 It is a schematic structural diagram of an embodiment of an electronic device of the present application. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0026] In the technical solution of this application, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved shall comply with the provisions of relevant laws and regulations and shall not violate public order and good morals.

[0027] Figure 1 A flowchart of an example of an IoT intelligent monitoring and alarm method for water quality detection according to an embodiment of the present application is shown.

[0028] Regarding the execution subject of the method of the embodiment of the present application, it can be any controller or processor with computing or processing capabilities. Specifically, it can be implemented by a water quality early warning platform server, by receiving sensor sampling data from each IoT node and performing corresponding risk analysis and alarm operations. By simultaneously monitoring multiple areas through distributed IoT nodes, frequent manual sampling and laboratory testing are avoided, and the consumption of material resources is reduced. At the same time, the intelligent water quality monitoring and risk identification model can automatically trigger alarms and intervention operations when water quality is abnormal, thereby improving the overall efficiency and resource utilization of water quality monitoring. In this way, through the integration of IoT technology and intelligent monitoring models, the real-time, comprehensive, intelligent and efficient water quality monitoring is achieved.

[0029] In some examples, it may be integrated and configured in an electronic device or terminal by means of software, hardware, or a combination of software and hardware, and the type of the terminal or electronic device may be diverse, such as a mobile phone, a tablet computer, or a desktop computer, etc.

[0030] like Figure 1 As shown, in step S110, water quality electrochemical data for corresponding monitored water areas are sampled based on each IoT node.

[0031] In some embodiments, IoT nodes are respectively deployed in different blocks within the monitored water area to achieve a distributed setting of IoT nodes. Electrochemical sensors are deployed in each IoT node. The electrochemical sensors can collect electrochemical data of water quality in the monitored block in real time, specifically including water quality parameters such as pH value, dissolved oxygen (DO) content and total organic carbon (TOC) content.

[0032] Specifically, the electrochemical sensors of each IoT node acquire the water quality data of the monitoring block at a certain sampling frequency (such as once a minute), and upload the collected data to the central data server in real time through the data transmission module in the IoT node. In addition, for complex water environments, the IoT node can dynamically adjust the sampling frequency and sampling time period according to the location, environmental conditions and frequency of water quality changes in the monitoring block. For example, for areas with potential pollution sources in the monitoring block (such as water bodies near factories or living areas), the sampling frequency can be appropriately increased to ensure that enough data points are obtained to reflect the trend of water quality changes.

[0033] In step S120, the electrochemical time-series sensing features corresponding to each water quality electrochemical data are extracted.

[0034] In some embodiments, first, the water quality electrochemical data collected from each IoT node is preprocessed, including noise removal, outlier correction and data smoothing, to improve the reliability of the data. Then, for each electrochemical sensor data type (such as pH value, DO content, TOC content), time series feature extraction is performed in the time dimension, including mean, variance, peak value, frequency characteristics and trend characteristics, etc., so as to reflect the fluctuation law and change trend of each sensor parameter in the time series. Therefore, through time series feature extraction, complex raw sensor data can be converted into feature data with stronger recognition ability and time series dependent characteristics, providing a favorable data basis for abnormal identification of subsequent models.

[0035] In step S130, each electrochemical time-series sensing feature is input into a water quality anomaly recognition model, and the pattern time-series dependency feature corresponding to the electrochemical time-series sensing feature is extracted to predict whether there is a risk of water quality anomaly in the corresponding monitored water area.

[0036] In some embodiments, a water quality anomaly recognition model is constructed based on historical water quality data and water quality anomaly sample data. In addition, the model type of the water quality anomaly recognition model can be diverse, for example, the model can use a machine learning algorithm (such as a decision tree, a support vector machine) or a deep learning model (such as a deep neural network, a convolutional neural network), and ensure the accuracy and generalization ability of the recognition through model training and verification.

[0037] Exemplarily, on the water quality early warning platform server, the model extracts the pattern time-series dependency features between each water quality parameter change and abnormal events by learning historical water quality data and labeled water quality abnormality sample data. Here, the pattern time-series dependency features may include time-series dependency relationships and time-series change patterns. Time-series dependency relationships refer to the dependent linkage relationship between various parameter types in the time dimension when the water quality is abnormal, which may be the learned longer-term feature information. Time-series change patterns refer to the correlation patterns between parameter fluctuation frequency, amplitude, fluctuation period, etc. and water quality abnormalities, which may be the learned shorter-term feature information. In this way, the water quality anomaly recognition model can deeply extract the pattern time-series dependency features based on the electrochemical time-series features collected by each input IoT node, analyze the water quality status of the water area block corresponding to each IoT node, and predict whether the block has a risk of water quality abnormality.

[0038] In step S140, for each risk monitoring water area block predicted to have abnormal water quality risk, the pattern time-dependent features corresponding to the risk monitoring water area block are input into the abnormal event recognition model to identify the type of abnormal water quality risk event corresponding to the risk monitoring water area block.

[0039] Here, the abnormal event recognition model can be trained by supervised learning, and a classification model is constructed based on the marked water quality abnormal event types (such as chemical pollution, excessive organic matter, heavy metal pollution, etc.), where chemical pollution events indicate the risk of toxic chemical leakage, excessive organic matter events indicate the risk of domestic sewage discharge, and heavy metal pollution indicates the risk of industrial wastewater discharge. In addition, the model type of the abnormal event recognition model can also be diversified, for example, various RNN (Recurrent Neural Network) models, multi-layer perceptrons, integrated learning models or autoregressive models can be used, and training and optimization can be performed through a large amount of marked water quality abnormal event data. Through the abnormal event recognition model, the temporal dependence characteristics of the pattern are further analyzed, and different types of water quality abnormal events can be accurately distinguished, which improves the accuracy and discrimination ability of water quality abnormal risk event classification, and provides important reference information for the source tracing and fundamental governance of water quality pollution.

[0040] In step S150, a water quality pollution alarm operation is performed in the water area according to the block identification of each risk monitoring water area block and the corresponding abnormal water quality risk event type.

[0041] In some embodiments, after the monitoring water area blocks with abnormal water quality risks and their corresponding abnormal event types are identified, the water quality warning platform server will automatically perform corresponding water quality pollution alarm operations based on the location of each risk monitoring water area block and its corresponding event type.

[0042] Exemplarily, the water quality early warning platform server maintains the corresponding pollution source event association table for the block identifiers of different monitored water blocks, and the pollution source event association table records multiple water quality abnormal risk event types and corresponding pollution sources. In this way, the water quality early warning platform server queries the pollution source event association table corresponding to the risk monitoring water block through the water quality abnormal risk event type, can effectively locate the corresponding pollution source, generate early warning messages, and provide water quality management departments with rapid positioning of pollution sources and formulation of corresponding treatment measures. For example, chemical pollution events are associated with early warnings for chemical plants near water blocks, and organic matter exceeding the standard events are associated with early warnings for domestic sewage pipes near water blocks, and so on.

[0043] As a result, automated water pollution alarm operation is achieved, which can send out alarms to water area managers or block leaders as soon as abnormal water quality events are identified, significantly improving the response efficiency of water quality monitoring and management, and effectively reducing the harm of water pollution events to the ecological environment of the water body.

[0044] Figure 2 An operational flow chart of an example of extracting electrochemical time-series sensing features corresponding to water quality electrochemical data according to an embodiment of the present application is shown.

[0045] like Figure 2 As shown, in step S210, sliding time series data corresponding to the water quality electrochemical data is extracted based on the sliding window, and the size of the sliding window is dynamically adjusted according to the fluctuation characteristics of the time series data in the sliding window.

[0046] In some embodiments, during the data acquisition phase, an initial sliding window size (such as 10 sampling points or a certain length of time) is set for each water quality electrochemical parameter (such as pH value, DO content, TOC content). Then, in each sliding window, the fluctuation characteristics of the water quality electrochemical data (such as the mean, variance, fluctuation amplitude, etc. of the data) are calculated in real time, and the water quality is judged to have obvious dynamic changes (such as sudden rise, sudden drop, etc.) according to the volatility characteristics of the data in the current sliding window. If a significant change in the fluctuation characteristics is detected (such as a sudden change in the mean or a sharp increase in the variance), the window size is appropriately reduced so that the details of the water quality fluctuation can be captured more finely in the next sampling window; conversely, when the data fluctuates smoothly, the window size is expanded to reduce the amount of data and computational complexity. In addition, the sliding window slides according to the set step size (such as sliding 1 sampling point each time), and the old data in the window is removed each time it slides, and the data of the new sampling point is added, so as to maintain the timeliness of the data in the window.

[0047] In step S220, for the sliding time series data in each sliding window, the discrete wavelet transform method is used to decompose the sliding time series data into multiple periodic amplitude features corresponding to different periodic amplitudes, and each periodic amplitude feature is weighted and fused to obtain the corresponding window comprehensive amplitude feature.

[0048] Here, a discrete wavelet transform (DWT) is performed on the sliding time series data in each sliding window (such as a time series data segment consisting of 10 sampling points of pH value). Discrete wavelet transform decomposes the time series data into high-frequency and low-frequency components by layer-by-layer decomposition, and describes the different characteristics of the data with different period amplitudes (i.e., the level of wavelet decomposition). Among them, each decomposition level corresponds to a specific period and amplitude feature, which can capture different periodic changes and fluctuation patterns in water quality time series data. For example, Daubechies wavelet basis function or Haar wavelet basis function can be selected for decomposition, and the number of decomposition layers (such as 3 layers, 4 layers, etc.) can be determined. After each layer is decomposed, the corresponding low-frequency (Approximation) and high-frequency (Detail) sub-band coefficients are obtained.

[0049] In addition, for the low-frequency and high-frequency subband coefficients at different decomposition levels, their amplitude characteristics are calculated, including: maximum value, minimum value, mean, variance, energy distribution (square sum), etc., to obtain the corresponding periodic amplitude characteristics. The low-frequency subband coefficient mainly represents the long-term trend change of the sliding time series data, and the high-frequency subband coefficient mainly represents the short-term fluctuation characteristics of the sliding time series data. By extracting multi-level periodic amplitude characteristics, the water quality data can be comprehensively described at different time scales. Then, for the high-frequency and low-frequency characteristics obtained by decomposition at different levels, the amplitude characteristics of each cycle are extracted, and the different periodic amplitude characteristics are weighted and fused according to the target water quality monitoring requirements. In addition, the setting of weights can be based on the importance of the features (such as increasing the weight of high-frequency fluctuation features and paying attention to the sudden change of data) or determined by machine learning methods (such as automatic weight optimization). Finally, the various periodic amplitude characteristics are weighted and fused according to the weights to form the comprehensive amplitude characteristics of the sliding window, which can characterize the multi-level change pattern of water quality data captured at different periodic amplitudes within the sliding window.

[0050] In step S230, the window comprehensive amplitude features corresponding to each continuous sliding window are combined to determine the corresponding electrochemical sequential sensing feature.

[0051] In some embodiments, for the water quality electrochemical parameters collected by each IoT node, a series of window comprehensive amplitude characteristics (such as multiple comprehensive amplitude characteristic sequences in chronological order) are obtained within the sliding window. Then, these continuous window comprehensive amplitude characteristics are combined in chronological order to form an overall time series feature sequence, which can reflect the fluctuation law and multi-scale amplitude characteristics of the water quality electrochemical data in the monitoring block over a period of time. Preferably, the combined time series feature sequence can also be processed by feature dimensionality reduction methods, such as principal component analysis (PCA) or linear discriminant analysis (LDA), to remove redundant information and retain the most representative time series features.

[0052] Through the embodiment of the present application, a feature extraction strategy combining a dynamic sliding window with a discrete wavelet transform is adopted, which can accurately extract the fluctuation characteristics and periodic change patterns of water quality data at multi-level periodic amplitudes within the window, effectively improving the feature expression ability and time series pattern recognition ability of water quality data. In addition, the dynamic adjustment of the sliding window size can adaptively adjust the window scale according to the fluctuation characteristics of the water quality data, improve the sensitivity and robustness to the changes in water quality data, and provide a flexible and reliable technical means for data feature extraction in different water environments. In addition, through weighted fusion and feature combination strategies, the periodic amplitude characteristics at multiple levels and different time scales are effectively fused, and the generated electrochemical time series sensing characteristics can better characterize the overall change trend and time series dependency of water quality data, providing rich and accurate feature input information for the intelligent identification of abnormal water quality risks.

[0053] Regarding the details of the sliding window in step S210, the size of the sliding window directly affects the extraction accuracy and timeliness of the time series features. In some examples of the embodiments of the present application, a sliding window adjustment strategy based on the fluctuation amplitude and change rate of the time series data is provided.

[0054] More specifically, the size of the sliding window is dynamically adjusted in the following way:

[0055] The initial window size of the current sliding window is determined according to the fluctuation characteristics of the time series data of the previous sliding window.

[0056]

[0057] Where w(t) 初始 represents the initial window size of the t-th sliding window, f(t-1) and a(t-1) represent the fluctuation frequency and fluctuation amplitude determined by the time series analysis of the water quality parameters of the t-1-th sliding window, respectively, and λ1 and λ2 are adaptive adjustment coefficients.

[0058] More specifically, by performing frequency domain analysis (such as fast Fourier transform FFT) on the water quality parameter values ​​in the previous time window, the main frequency components are identified and their average frequency value f(t-1) is calculated. The specific calculation steps are as follows:

[0059] For the time series data X in the previous time window t-1 ={x t-1,1 ,x t-1,2 ,...,x t-1,w(t-1)}Perform FFT transformation to obtain the spectrum F t-1 ={F t-1,1 ,F t-1,2 ,...,F t-1,k}.

[0060] Calculate the energy value of each frequency component in the spectrum And select the frequency component with the largest energy value as the main frequency component f(t-1) of the current fluctuation.

[0061] On the other hand, for the calculation of the fluctuation amplitude a(t-1), the maximum value X of the water quality parameter value in the previous time window can be calculated: t-1,max With the minimum value X t-1,min The difference between the two is defined as the fluctuation amplitude a(t-1) = X t-1,max -X t-1,min .

[0062] In this way, by introducing the fluctuation frequency f(t-1) and fluctuation amplitude a(t-1) of the previous time window, the current window size can be adjusted according to the historical fluctuation characteristics, thereby improving the ability to predict short-term fluctuations and long-term trends.

[0063] Furthermore, in order to avoid unstable feature extraction caused by frequent fluctuations in window size, this paper introduces a constraint mechanism for the sliding window size change rate, that is, if the current sliding window size changes too much, it switches to a smooth adjustment strategy.

[0064] Specifically, the window size change rate corresponding to the initial window size of the current sliding window is calculated.

[0065]

[0066] Where V w (t) represents the window size change rate, and w(t-1) represents the window size of the t-1th sliding window.

[0067] If the window size change rate of the current sliding window does not exceed a preset change rate threshold, the initial window size is used as the window size of the current sliding window.

[0068] If the window size change rate of the current sliding window exceeds the change rate threshold, the window size of the current sliding window will be smoothly calculated using the following formula.

[0069] w(t)=w(t-1)+η·(w(t) 初始 -w(t-1)), Formula (3)

[0070] Where w(t) represents the window size of the tth sliding window, and η is the smoothing coefficient.

[0071] It should be noted that in the actual collection of water quality electrochemical data, water quality parameters (such as pH value, DO content and TOC content) are often affected by environmental factors and pollutant concentration fluctuations, showing complex time series characteristics. Therefore, the sliding window needs to have high sensitivity and responsiveness to these fluctuation characteristics.

[0072] In an embodiment of the present application, the initial size of the current time window is derived by calculating the fluctuation frequency and fluctuation amplitude of the previous time window, and on this basis, the window size is further smoothly adjusted by controlling the change rate, so that the current size can smoothly transition to a new state by introducing a dynamic adjustment mechanism of the window size based on the fluctuation characteristics of the historical window, avoiding sudden changes in the window size caused by fluctuations at a single moment. In addition, by introducing constraints on the window size change rate, when the change rate exceeds a preset threshold, the system can automatically smooth the window size, further ensuring the continuity and stability of the sliding window size change, thereby reducing the volatility of the feature extraction results. Thus, the current window size is adjusted according to the fluctuation frequency and fluctuation amplitude of the previous time window, so that the sliding window can adaptively capture short-term fluctuation characteristics and long-term trend changes, enhance the sliding window's ability to respond to water quality fluctuations, and thus improve the accuracy of feature extraction.

[0073] Regarding the details in step S220, the discrete wavelet transform (DWT) is used to decompose the time series data in the current sliding window, and the time series data is decomposed into feature representations at different period scales (i.e., different frequency components). It should be noted that the discrete wavelet transform has stronger localization capabilities in the time domain and frequency domain than other time-frequency analysis methods, such as the short-time Fourier transform (STFT), and can perform more accurate feature extraction of short-term fluctuations and long-term trends of water quality parameters at different time scales.

[0074] More preferably, in the extraction of time series features from water quality electrochemical data, a time series feature modeling method based on periodic analysis is used to perform multi-scale decomposition of the time series data in an adaptive sliding window by discrete wavelet transform, thereby extracting time series features at different period amplitudes.

[0075] Specifically, the sliding time series data of the current sliding window is subjected to discrete wavelet transform to obtain multiple sub-band representations.

[0076]

[0077] The time series data in the t-th sliding window is represented by x={x1,x2,...,x w(t)}, perform discrete wavelet transform on x to obtain multiple subbands representing X s ={X s1 ,X s2 ,...,X sm}, m is the number of subbands, each subband X si represents the amplitude of the ith cycle s i The characteristic component under x j represents the electrochemical parameter value of water quality at the jth time step in x, X si (t) represents the t-th sliding window in the period amplitude s i The characteristic component under ψ si (j) represents the periodic amplitude s of the tth sliding window i The wavelet basis function under different period amplitudes can capture the characteristics of time series data at different frequencies; j represents the time step offset, which is used to describe the weight of the eigenvalue of the jth time step of the tth sliding window.

[0078] Then, the energy value of each periodic amplitude feature is calculated. Here, in order to measure the significance of each periodic scale feature, the energy value is introduced as an importance measurement indicator of the periodic feature.

[0079]

[0080] In the formula, E si (t) represents the period amplitude s in the t-th sliding window i The energy value of the lower subband feature can reflect the importance of the characteristic component under the current cycle amplitude; Indicates subband X si The sum of squares of the eigenvalues ​​over all time steps within the t-th sliding window.

[0081] Furthermore, based on the significance measurement index, the comprehensive period significance of each period amplitude feature is calculated. The significance measurement index includes periodic intensity and relative energy proportion.

[0082] In the embodiment of the present application, on the basis of the basic energy value measurement, a multi-dimensional measurement index based on the significance of periodic features is introduced, including periodic strength and relative energy proportion. Here, periodic strength represents the significance of the periodic feature in the current time window, which is defined as the product of the characteristic fluctuation amplitude and the energy value of the frequency component. In addition, the relative energy proportion represents the relative importance of the periodic feature in all period scales.

[0083] θ si (t) = l1·P si (t)+l2·REP si (t), Formula (6)

[0084] P si (t) = A si (t)·E si (t), Formula (7)

[0085] A si (t) = max{X si (j)}-min{X si (j)}, Formula (8)

[0086]

[0087] Where P si (t) represents the period amplitude s in the t-th sliding window i The periodic intensity under A si (t) represents the period amplitude s in the t-th sliding window i The fluctuation amplitude of the lower subband feature is defined as the maximum value of the subband feature value max{X si (j)} and the minimum value min{X si (j)}; REP si (t) represents the period amplitude s in the t-th sliding window i The relative energy proportion of the characteristic components under ; is a normalized term, which represents the sum of the energy values ​​of the sub-band features of all periodic amplitudes in the t-th sliding window; θ si (t) represents the period amplitude s in the t-th sliding window i The comprehensive period significance of , l1 and l2 represent the weight coefficients of the corresponding significance measurement indicators.

[0088] It should be noted that when the energy value of a certain period amplitude accounts for a large proportion of the overall energy value, it means that the period feature has a higher importance in the current time window.

[0089] By performing multi-scale decomposition of time series data through discrete wavelet transform, characteristic components can be extracted at different periodic scales, and the periodic intensity P si(t) and relative energy proportion REP si (t) Measures the significance of periodic features, thereby accurately capturing the complex fluctuation patterns of water quality parameters within the current sliding window.

[0090] Furthermore, based on the comprehensive period significance, each period amplitude feature is weighted and fused to obtain the corresponding window comprehensive amplitude feature.

[0091]

[0092] Where F(t) represents the window comprehensive amplitude feature corresponding to the t-th sliding window.

[0093] Here, after integrating the characteristic components at all periodic scales, the generated comprehensive feature representation F(t) can reflect the overall fluctuation and periodic change characteristics of the water quality parameters in the current sliding window, improving the feature representation capability. Therefore, it can not only reflect the main fluctuation patterns in the water quality data, but also reveal the periodic change laws hidden in the data, so that potential water quality abnormalities can be more accurately identified during anomaly detection.

[0094] It should be noted that the temporal changes of water quality parameters are often affected by environmental conditions (such as water temperature, water flow velocity, external pollutant concentration, etc.) and internal system states (such as monitoring equipment drift, data sampling frequency changes, etc.). Traditional fixed window or single-scale feature extraction methods are difficult to adapt to the dynamic changes in these complex environments.

[0095] Through the embodiment of the present application, an adaptive sliding window adjustment strategy is introduced to dynamically determine the size of the current sliding window according to the fluctuation characteristics of the previous time window, so that the sliding window can be adaptively adjusted to follow the fluctuation of water quality parameters. Furthermore, DWT is used to perform multi-scale decomposition of the time series data in the sliding window, and through energy value calculation and significance weight allocation strategy, the time series characteristics can be modeled and optimized at different period scales. The comprehensive scale feature representation of each sliding window can simultaneously reflect short-term characteristics (such as instantaneous fluctuations, sudden anomalies) and long-term characteristics (such as overall trend changes, long-term pollutant accumulation effects), which improves the system's adaptability to multiple characteristic patterns in complex environments.

[0096] Figure 3 A schematic diagram of the structural connection of an example of a water quality anomaly identification model according to an embodiment of the present application is shown.

[0097] like Figure 3 As shown, the water quality anomaly identification model 300 includes a cascaded time series feature modeling module 310 and a feature anomaly identification module 320 .

[0098] The time series feature modeling module 310 is used to model the input electrochemical time series sensing features, extract the dynamic change pattern and time dependency of the features for the monitored water area block, and obtain the corresponding pattern time series dependency features.

[0099] In some embodiments, the time series feature modeling module 310 first normalizes the input data (such as standardizing each feature value to an interval of 0 to 1) to eliminate the dimensional differences between the various electrochemical features. Then, various neural network models suitable for time series data modeling, such as LSTM (Long Short-Term Memory, Long Short-Term Memory Network), GRU (Gated Recurrent Unit, Gated Recurrent Unit) or ARIMA (Autoregressive Integrated Moving Average, Autoregressive Integrated Moving Average Model), can be used to model the input electrochemical time series sensing features. Exemplarily, LSTM and GRU have the ability to process long time series dependencies, and can effectively capture the dynamic patterns and internal dependencies of water quality electrochemical characteristics over time. For example, through the LSTM network, it is possible to identify the hysteresis effect of changes in pH value on changes in DO content, or whether an increase in TOC content will cause a decrease in DO content. Furthermore, the hidden layer features extracted in the process of temporal feature modeling (such as the hidden state vector of the neural network) can be regarded as pattern temporal dependency features to reflect the dynamic change trend and periodic characteristics of the water quality characteristics of the current monitoring block.

[0100] The feature anomaly identification module 320 is used to process the pattern time-dependent features to predict whether there is a risk of water quality anomaly in the corresponding monitored water area.

[0101] Here, an anomaly recognition model suitable for processing pattern time-dependent features is constructed. For example, a traditional machine learning algorithm (such as SVM (Support Vector Machine), RF (Random Forest)) or a deep learning model (such as MLP (Multilayer Perceptron), CNN (Convolutional Neural Network)) can be used. Exemplarily, the label data used in the data sample set of the anomaly recognition model comes from the water quality anomaly records of the historical monitoring block, and supervised learning or semi-supervised learning (if some data are not labeled) is performed based on the actual data of various types of water quality anomaly records. When training the model, special attention is paid to the relationship between the time-dependent features and the water quality anomaly records, such as when the water quality characteristics suddenly fluctuate violently (such as a sudden drop in pH), whether there is a similar pattern to a certain type of water quality anomaly in the historical data. In this way, in the reasoning application stage, the pattern time-dependent features are input into the trained anomaly recognition model, which can better predict whether there is a risk of water quality anomaly in the current monitoring block.

[0102] Preferably, in some examples of the embodiments of the present application, the time series feature modeling module adopts a combination of deep neural network and time series analysis, so as to effectively deal with abnormal water quality in complex environments.

[0103] Figure 4 A schematic diagram of the structural connection of another example of a water quality anomaly identification model according to an embodiment of the present application is shown.

[0104] like Figure 4 As shown, the water quality anomaly recognition model 400 includes a time series feature modeling module 410 and a feature anomaly recognition module 420. The time series feature modeling module 410 includes an LSTM layer 411, a TCN (Temporal Convolutional Network) layer 412 and a feature fusion layer 413, and the LSTM layer 411 and the TCN layer 412 are connected in parallel to the feature fusion layer 413. The feature anomaly recognition module 420 includes an MLP layer 421, a Softmax layer 422 and an anomaly classification layer 423.

[0105] The LSTM layer 411 is used to extract long-term trend features in the electrochemical time series sensing features through a gating structure to capture the trend changes and long-term dependencies of water quality parameters.

[0106] h g,t =σ1(W h ·[h g,t-1 ,V g,t ]+b h ), Formula (11)

[0107] H g =[h g,1 ,h g,2 ,...,h g,T ], Formula (12)

[0108] In the formula, h g,t Indicates the LSTM layer monitors the water area The hidden state output at the t-th sliding window, W h represents the weight matrix of the LSTM layer, h g,t-1 Indicates the LSTM layer monitors the water area The hidden state output at the t-1th sliding window; V g,t Indicates monitored water area The electrochemical input feature vector in the tth sliding window, b h represents the bias term, σ1(·) represents the activation function of the LSTM layer; T represents the total number of time steps corresponding to the electrochemical temporal sensing feature, H g Indicates that the LSTM layer is used to monitor the water area The extracted long-term trend features.

[0109] The LSTM layer can effectively capture the long-term dependencies in time series data through its internal gating structure, and is suitable for extracting trend changes in time series features. Therefore, for each monitored water area block, the hidden state output of LSTM at all time steps constitutes its time series feature representation sequence, which contains the long-term dependencies in the electrochemical time series data and can reflect the trend changes and long-term fluctuation patterns of water quality parameters.

[0110] The TCN layer 412 extracts local fluctuation pattern features in the electrochemical time series sensing features by stacking multiple layers of convolution operations to capture short-term fluctuations and local change patterns of water quality parameters.

[0111] y g,t =W p *V g,t +b p , Formula (13)

[0112] Y g =[y g,1 ,y g,2 ,...,y g,T ], Formula (14)

[0113] In the formula, y g,t Indicates the TCN layer for monitoring water area blocks The convolution output feature of the t-th sliding window, W p and b pThey represent the convolution kernel weight and bias term of the TCN layer respectively, * represents a one-dimensional convolution operation; Y g Indicates that the LSTM layer is used to monitor the water area Extracted local fluctuation pattern features.

[0114] Here, TCN extracts features at multiple scales through multi-layer convolution operations. Furthermore, by stacking multiple layers of convolution operations, TCN can gradually extract multi-scale features in time series data and concatenate the outputs of all convolution layers into local feature representations.

[0115] The feature fusion layer 413 is used to perform weighted fusion on the output results of the LSTM layer and the output results of the TCN layer to obtain corresponding pattern temporal dependency features.

[0116] F g =α·H g +(1-α)·Y g , Formula (15)

[0117] In the formula, F g Indicates monitoring water area The pattern time-series dependency feature, α is the feature fusion coefficient.

[0118] In this way, based on the feature fusion strategy of weighted fusion, the long-term feature representation of LSTM output and the local feature representation of TCN output are weightedly fused to achieve long-term and short-term modeling of time series features, which can effectively extract the dynamic change pattern and time dependency of each monitoring block.

[0119] On the other hand, the feature anomaly identification module 420 is used to process the pattern time-dependent features through the first MLP layer 421, and predict the probability value of the risk of water quality anomaly in the monitored water area through the Softmax layer 422.

[0120]

[0121] In the formula, Indicates monitoring water area The output feature representation of the fth hidden layer is W (f) and b (f) represents the weight matrix and bias term of the fth hidden layer; Indicates the monitoring water area block β g The output feature representation of the f-1th hidden layer is that in the calculation of the first hidden layer P g Indicates monitored water area The abnormal probability of water quality abnormality risk; L1 is the total number of hidden layers in the first MLP layer, σ2(·) represents the activation function of the hidden layer, and Represent the weight matrix and bias term of the Softmax layer respectively, The last hidden layer of the first MLP layer is used to monitor the water area. The output feature representation, Softmax(·) represents the Softmax activation function.

[0122] The abnormal classification layer 423 is used to compare the predicted probability value of the water quality abnormal risk with the preset abnormal probability threshold, and determine whether the corresponding monitored water area has a water quality abnormal risk based on the comparison result.

[0123] Here, the extracted features are further integrated through the feature anomaly recognition module, and the abnormal risk probability of each monitoring block is output, and the threshold comparison is used to determine whether the monitoring block has abnormal water quality risk. The combined model based on deep time series feature modeling and feature anomaly recognition module provided by the embodiment of the present application focuses on modeling and identifying abnormal states of the time series features of each monitoring block, which belongs to the detection stage. The time series data is dynamically extracted through the deep time series feature modeling module (LSTM and TCN) to identify long-term dependencies and local fluctuation patterns in the data; then, feature fusion and anomaly detection are performed through the feature anomaly recognition module to determine whether each monitored water area block has abnormal water quality. In this way, the abnormal change pattern in the time series data can be accurately captured, and the abnormal state of each monitoring block (that is, the presence or absence of anomalies) can be output.

[0124] It should be noted that in actual water quality monitoring scenarios, the chemical composition of water bodies (such as pH, DO content and TOC content) will be affected by a variety of complex factors, such as water flow changes, pollution source diffusion, temperature fluctuations, etc. These factors often cause water quality parameters to show different scales of change patterns over time, including long-term trends (such as seasonal changes) and short-term fluctuations (such as sudden pollutant injection).

[0125] Through the embodiment of the present application, the LSTM layer and the TCN layer are used in parallel modeling in the water quality anomaly identification model. The LSTM layer can effectively capture the long-term trend characteristics in the electrochemical time-series sensing characteristics, and extract the long-term dependence of water quality parameters over time, such as the gradual change trend of the pH value of the water body over many days. In addition, the TCN layer can extract the local fluctuation pattern in the electrochemical time-series sensing characteristics through multi-layer convolution operations, and capture the short-term fluctuations (such as the sudden increase in TOC content within a certain period of time) and local change patterns in water quality parameters. Therefore, through the parallel time-series feature extraction method, not only can the model better understand the complex changes in water quality data, but also can avoid the problem of insufficient feature representation caused by using only a single model, and significantly improve the sensitivity and detection accuracy of abnormal water quality conditions.

[0126] In addition, in actual water quality monitoring, different types of water quality anomalies (such as sudden changes in pH, decreased DO content, excessive TOC content, etc.) will show different temporal characteristic patterns. By using the MLP layer in the water quality anomaly recognition model to perform nonlinear mapping and transformation of the fused features, and using the Softmax layer and classification layer to make refined judgments on the probability of water quality anomalies, the recognition ability and sensitivity of various types of water quality anomalies can be significantly improved.

[0127] Figure 5 A schematic diagram of the structural connection of an example of an abnormal event recognition model according to an embodiment of the present application is shown.

[0128] like Figure 5 As shown, the abnormal event recognition model 500 includes a cascaded multi-scale convolution module 510 and an event type classification module 520, and the event type classification module 520 includes an attention layer 521, a second MLP layer 522, a confidence calculation layer 523 and an abnormal event classification layer 524.

[0129] The multi-scale convolution module 510 is used to extract the multi-scale time series features corresponding to the pattern time series dependency features of the risk monitoring water area block.

[0130] Here, the multi-scale convolution module is used to process the temporal dependence characteristics of the input pattern, and by designing convolution kernels of multiple time scales, the multi-scale convolution operation is used to model and fuse the features of different time lengths or scales, so as to obtain a richer multi-scale temporal feature representation and capture the fluctuation and trend information in the features at different time lengths, thereby improving the model's ability to recognize complex water quality abnormal event types.

[0131]

[0132] In the formula, Indicates monitoring block The convolution output feature representation at the e-th convolution kernel scale includes the local fluctuation and trend characteristics at this convolution kernel scale; and They represent the convolution kernel weight matrix and bias term of the e-th convolution kernel scale respectively; F D Indicates risk monitoring waters The temporal dependency characteristics of the pattern; σ3(·) represents the activation function of the multi-scale convolution module; M D Indicates risk monitoring waters The comprehensive multi-scale feature representation, μ e represents the scale fusion weight of the e-th convolution kernel scale, and U represents the total number of convolution kernel scales in the multi-scale convolution module.

[0133] It should be noted that in actual water quality monitoring scenarios, abnormal water quality events (such as chemical pollution, excessive organic matter, heavy metal pollution, etc.) are often accompanied by interference from a variety of complex factors, such as changes in water flow, diffusion of external pollution sources, etc. These factors make the abnormal features in water quality data present a variety of temporal change patterns and spatial distribution characteristics. By adopting a multi-scale convolutional feature extraction and fusion strategy, the model can dynamically model the input pattern temporal dependency features at different time scales, generate a more accurate multi-scale feature representation, and significantly improve the model's recognition accuracy for multiple types of water quality abnormal events.

[0134] The attention layer 521 is used to calculate the weight allocation matrix for the multi-scale temporal features to determine the corresponding attention features.

[0135]

[0136] In the formula, Indicates risk monitoring waters The attention weight of the e-th convolution kernel scale; and Represents the attention weight matrix and bias term of the e-th convolution kernel scale; Indicates that the attention layer monitors the water area for risks Output attention features.

[0137] It should be noted that different types of abnormal water quality events often present different distribution patterns in the feature space. In view of this, through the embodiment of the present application, the feature weight allocation strategy based on the attention mechanism can weight the feature representations of each time scale to highlight the feature representations with high discrimination for specific event types. The weight allocation matrix calculated by the attention mechanism can dynamically adjust the importance of each feature in the feature representations of different time scales, thereby enhancing the model's ability to recognize different types of abnormal water quality events.

[0138] In addition, by adopting multi-scale convolution feature extraction and feature fusion strategies, the model can independently model features at different time scales in the feature extraction stage, and then generate a comprehensive feature representation through a dynamic weighted combination of feature fusion and attention mechanism. The model structure design of hierarchical modeling and feature combination can effectively reduce the complexity of the model's feature representation, thereby improving the model's feature representation capability and training efficiency.

[0139] The second MLP layer 522 is used to perform nonlinear mapping and feature conversion processing on the attention features, and determine the confidence matrix of the risk monitoring water area block for a preset set of risk event types through the confidence calculation layer 523.

[0140]

[0141] In the formula, Indicates risk monitoring waters The output feature representation of the rth hidden layer; W (r) and b (r) represents the weight matrix and bias term of the rth hidden layer; Indicates risk monitoring waters The output feature representation of the r-1th hidden layer is that in the calculation of the first hidden layer L2 is the total number of hidden layers in the second MLP layer, σ4(·) represents the activation function of the hidden layer, The last hidden layer of the second MLP layer is used to monitor the risk water area. The output comprehensive risk characterization; and Represent the weight matrix and bias term of the confidence calculation layer, P D Indicates risk monitoring water area The confidence matrix contains the confidence scores for each risk event type in the risk event type set.

[0142] Here, the feature representation based on the attention mechanism is input into the second MLP layer. Through the layer-by-layer calculation of the second MLP layer, the feature representation can be mapped into a more compact feature representation to achieve the classification and judgment of the event type.

[0143] The abnormal event classification layer 524 is used to process the confidence matrix to identify the water quality abnormal risk event type corresponding to the risk monitoring water area block.

[0144] Here, the processing methods for the confidence matrix can be diverse and can be selected or adjusted according to business needs, for example, a highest confidence selection strategy for outputting a single event and a confidence threshold comparison strategy that allows outputting multiple types of events.

[0145] Through the embodiments of the present application, based on the principle of differences in the distribution patterns of different types of abnormal water quality events in the feature space, by introducing the attention mechanism, the model can effectively strengthen the selection and weighting of specific time scale features, thereby improving the ability to distinguish different event types and classification accuracy. In addition, in the face of the situation where water quality data in different monitoring blocks may present highly unbalanced and complex distribution patterns, by introducing multi-scale feature extraction and attention mechanisms, the model can dynamically adapt to different data distribution patterns, and refine the prediction results in combination with the confidence classification module, effectively reducing the probability of false positives and false negatives.

[0146] It should be noted that in the embodiment of the present application, two independent intelligent models are used to identify water quality anomalies. The water quality anomaly identification model focuses on the abnormal risk detection task, while the abnormal event identification model focuses on the prediction task of specific abnormal risk events. They can be refined and optimized for their specific tasks respectively, and the model can be designed, optimized and maintained more efficiently. In addition, different algorithms and strategies are used in the detection stage and the prediction stage to improve the adaptability and flexibility of the model. For example, different time series modeling networks are used in the detection stage, and different classification and time series prediction algorithms are used in the prediction stage to achieve algorithm decoupling of detection and prediction and improve the overall system performance.

[0147] It should be understood that in the embodiments of the present application, the types of abnormal water quality risk events included in the risk event type set can be diverse. In addition to the chemical pollution, excessive organic matter, and heavy metal pollution described above, there can also be other risk event types, such as trend-based water quality deterioration events, sudden pollution events, and long-term stability deviation events.

[0148] Specifically, a trending water quality deterioration event refers to a situation in which water quality characteristics show a continuous deterioration trend over a long period of time (such as weeks or months), rather than a sudden abnormality. Such events usually reflect chronic pollution of the water environment or a trend of gradually deteriorating water quality. A sudden pollution event refers to a sudden and drastic change in a water quality characteristic (or multiple characteristics) within a short period of time (such as hours or a day), and the change is obviously beyond the normal fluctuation range. Such events are usually caused by sudden pollution sources (such as factory emissions, sudden leakage accidents, etc.). A long-term stability deviation event refers to a water quality characteristic (such as pH value, DO content) that deviates from the normal range continuously and steadily over a long period of time, but no obvious trend changes occur. Such events are usually related to stable pollution sources (such as groundwater pollution) or insufficient self-purification capacity of water bodies.

[0149] The abnormal event recognition model adopts modules such as multi-layer feature modeling, multi-scale time series feature fusion and attention mechanism, which can accurately identify various types of abnormal water quality risk events in complex water quality monitoring scenarios, effectively solving the limitations of traditional models that cannot handle multi-dimensional complex features and cannot identify complex and trend water quality events. The model can comprehensively analyze the complex relationships and dynamic change patterns between multiple water quality parameters, thereby identifying more complex and diverse risk event types and achieving more comprehensive and accurate risk event identification capabilities.

[0150] In some examples of the embodiments of the present application, the loss function of the abnormal event recognition model is used to measure the performance of the model when processing data samples of multiple risk monitoring water areas in complex water quality monitoring scenarios, thereby improving its classification accuracy, time series feature modeling capabilities and feature allocation capabilities in complex environments by optimizing the model training.

[0151] Here, the data sample set contains multiple data samples for risk monitoring water blocks, and each data sample contains characteristic data and corresponding labels. Each data sample contains multiple water quality electrochemical parameter characteristics (such as pH value, DO content, TOC content, etc.), and can be provided with spatiotemporal information or identification related to the monitoring block. The label of each data sample indicates the type of abnormal water quality risk event to which it belongs.

[0152] Specifically, the abnormal event recognition model uses a comprehensive loss function that includes classification cross entropy loss, adaptive balance loss term, and temporal consistency loss term. Specifically, the loss function is:

[0153]

[0154]

[0155] In the formula, represents the loss function of the abnormal event recognition model, represents the classification cross entropy loss term, represents the adaptive balancing loss term, and represents the time series consistency loss term, β1, β2, β3 represent the weight coefficients of each corresponding loss term respectively; N represents the total number of samples in the risk data sample set for the risk monitoring waters, C represents the total number of water quality abnormal risk event types in the risk event type set, y d,c represents the actual label of the d-th risk data sample on the c-th risk event type. If the block belongs to the c-th event type, then y d,c =1, otherwise y d,c =0; p d,c represents the confidence that the d-th risk data sample is predicted by the abnormal event recognition model as the c-th risk event type; γ is the adaptive balance coefficient, which is used to adjust the balance between feature weight and classification confidence; w d,c represents the feature weight distribution value of the d-th risk data sample on the c-th risk event type; Q represents the total time steps of the multi-scale time series feature, Represents the feature representation of the qth time step of the dth risk data sample at the eth convolution kernel scale; Represents the feature representation of the d-th risk data sample at the q-1th time step at the e-th convolution kernel scale.

[0156] In the comprehensive loss function provided in the embodiment of the present application, the classification cross entropy loss can measure the classification accuracy of the model on each event type by comparing with the label data of the data sample set, and improve the model's fitting ability to the data sample set by optimization, thereby improving the model's classification performance for different risk event types in practical applications. Based on the adaptive balance loss of feature weights and confidence, it can measure the balance between feature weight allocation and classification confidence. By combining the feature data and label data in the data sample set, it can improve the model's dynamic weight allocation ability on each data sample, thereby improving the model's dynamic allocation ability for sample features. Based on the time series consistency loss of multi-scale time series feature modeling, it measures the feature consistency of the model at different time scales. By combining with the time series feature data in the data sample set, it ensures that the model's judgment on the same risk event at multiple time steps is consistent, which can help the model better understand the feature relationship at different time steps.

[0157] It should be noted that, unlike the traditional single loss function design, the comprehensive loss function provided in the embodiment of the present application introduces an adaptive balance mechanism and time consistency constraints. The model can maintain a high generalization ability in a complex and changeable water quality monitoring environment. When dealing with complex risk events (such as complex pollution events and long-term trend events), the model can target the distribution characteristics of multi-dimensional water quality characteristics, adaptively adjust feature weights and time consistency, and help the model maintain high classification accuracy and stability in different monitoring environments (such as environments with large fluctuations in water quality parameters and areas with complex and changeable pollution sources), thereby effectively improving the generalization ability and robustness of the model in complex environments.

[0158] Through the embodiments of the present application, the comprehensive loss function can optimize the classification accuracy, feature weight distribution and time series consistency in the same training process while ensuring the classification accuracy, avoiding the problem of imbalanced model performance caused by a single loss term, improving the comprehensive performance and overall performance of the model, and significantly improving the comprehensive capabilities of the model in multi-type event recognition, composite feature representation and multi-time scale modeling, providing more comprehensive and powerful technical support for risk warning of intelligent water quality monitoring systems.

[0159] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of actions combined, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application. In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0160] Figure 6 A structural block diagram of an example of an Internet of Things intelligent monitoring and alarm system for water quality detection according to an embodiment of the present application is shown.

[0161] like Figure 6 As shown, the IoT intelligent monitoring and alarm system 600 for water quality detection includes a water quality parameter sampling unit 610, a time series feature extraction unit 620, a water quality anomaly identification unit 630, an abnormal event identification unit 640 and a water area water quality alarm unit 650.

[0162] The water quality parameter sampling unit 610 is used to sample the water quality electrochemical data for the corresponding monitored water area blocks based on each IoT node; multiple IoT nodes are distributedly arranged in different blocks in the monitored water area, and each of the IoT nodes is deployed with an electrochemical sensor; the sensing parameter types of the water quality electrochemical data include pH value, DO content and TOC content of the water body.

[0163] The time series feature extraction unit 620 is used to extract the electrochemical time series sensing features corresponding to each of the water quality electrochemical data.

[0164] The water quality anomaly identification unit 630 is used to input each of the electrochemical time series sensing features into the water quality anomaly identification model, extract the pattern time series dependency features corresponding to the electrochemical time series sensing features, so as to predict whether there is a water quality anomaly risk in the corresponding monitored water area.

[0165] The abnormal event identification unit 640 is used to input the pattern time-dependent characteristics corresponding to each risk monitoring water area block predicted to have abnormal water quality risk into the abnormal event identification model to identify the type of water quality abnormal risk event corresponding to the risk monitoring water area block.

[0166] The water area water quality alarm unit 650 is used to perform a water area water quality pollution alarm operation according to the block position of each risk monitoring water area block and the corresponding water quality abnormality risk event type.

[0167] In some embodiments, an embodiment of the present application provides a non-volatile computer-readable storage medium, which stores one or more programs including execution instructions, and the execution instructions can be read and executed by an electronic device (including but not limited to a computer, a server, or a network device, etc.) to execute the steps of any of the above-mentioned IoT intelligent monitoring and alarm methods for water quality detection in the present application.

[0168] In some embodiments, the embodiments of the present application also provide a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes any step of the above-mentioned Internet of Things intelligent monitoring and alarm method for water quality detection.

[0169] In some embodiments, the embodiments of the present application also provide an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the steps of the Internet of Things intelligent monitoring and alarm method for water quality detection.

[0170] Figure 7 is a schematic diagram of the hardware structure of an electronic device for executing an IoT intelligent monitoring and alarm method for water quality detection provided by another embodiment of the present application, such as Figure 7 As shown, the device includes:

[0171] One or more processors 710 and memory 720, Figure 7 A processor 710 is taken as an example.

[0172] The device for executing the IoT intelligent monitoring and alarm method for water quality detection may also include: an input device 730 and an output device 740 .

[0173] The processor 710, the memory 720, the input device 730 and the output device 740 may be connected via a bus or other means. Figure 7 The example of connecting through bus is taken in the following.

[0174] The memory 720, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules corresponding to the IoT intelligent monitoring and alarm method for water quality detection in the embodiment of the present application. The processor 710 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 720, that is, the IoT intelligent monitoring and alarm method for water quality detection in the above method embodiment is realized.

[0175] The memory 720 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 720 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 720 may optionally include a memory remotely arranged relative to the processor 710, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0176] The input device 730 may receive input digital or character information and generate signals related to user settings and function control of the electronic device. The output device 740 may include a display device such as a display screen.

[0177] The one or more modules are stored in the memory 720, and when executed by the one or more processors 710, the IoT intelligent monitoring and alarm method for water quality detection in any of the above method embodiments is executed.

[0178] The above-mentioned product can execute the method provided in the embodiment of the present application, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in this embodiment, please refer to the method provided in the embodiment of the present application.

[0179] The electronic devices of the embodiments of the present application exist in various forms, including but not limited to:

[0180] (1) Mobile communication equipment: This type of equipment is characterized by having mobile communication functions and its main purpose is to provide voice and data communications. This type of terminal includes: smart phones, multimedia phones, functional phones, and low-end phones.

[0181] (2) Ultra-mobile personal computer devices: These devices belong to the category of personal computers, have computing and processing functions, and generally also have mobile Internet access features. These terminals include: PDA, MID and UMPC devices, etc.

[0182] (3) Portable entertainment devices: These devices can display and play multimedia content. They include audio and video players, handheld game consoles, e-books, smart toys, and portable car navigation devices.

[0183] (4) Other onboard electronic devices with data interaction functions, such as on-board devices installed in vehicles.

[0184] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0185] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0186] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An IoT intelligent monitoring and alarm method for water quality detection, comprising: Based on each IoT node, water quality electrochemical data for the corresponding monitored water area is sampled; A plurality of IoT nodes are distributedly arranged in different blocks in the monitored water area, and each IoT node is deployed with an electrochemical sensor; the sensing parameter types of the water quality electrochemical data include pH value, DO content and TOC content of the water body; Extracting the electrochemical time-series sensing features corresponding to each of the water quality electrochemical data; Input each of the electrochemical time-series sensing features into a water quality anomaly recognition model, extract the pattern time-series dependency features corresponding to the electrochemical time-series sensing features, and predict whether there is a risk of water quality anomaly in the corresponding monitored water area; For each risk monitoring water area block predicted to have abnormal water quality risk, the pattern time-series dependency feature corresponding to the risk monitoring water area block is input into the abnormal event recognition model to identify the type of abnormal water quality risk event corresponding to the risk monitoring water area block; According to the block identification of each risk monitoring water area block and the corresponding water quality abnormal risk event type, the water area water quality pollution alarm operation is executed.

2. The method according to claim 1, wherein: The step of extracting the electrochemical time-series sensing features corresponding to each of the water quality electrochemical data comprises: Extracting sliding time series data corresponding to the water quality electrochemical data based on a sliding window; wherein the size of the sliding window is dynamically adjusted according to the fluctuation characteristics of the time series data within the sliding window; For the sliding time series data in each sliding window, the sliding time series data is decomposed into a plurality of period amplitude features corresponding to different period amplitudes by using a discrete wavelet transform method, and each of the period amplitude features is weightedly fused to obtain a corresponding window comprehensive amplitude feature; The window comprehensive amplitude characteristics corresponding to each continuous sliding window are combined to determine the corresponding electrochemical timing sensing characteristics.

3. The method according to claim 1, wherein: The size of the sliding window is dynamically adjusted in the following manner: According to the fluctuation characteristics of the time series data of the previous sliding window, determine the initial window size of the current sliding window: Where w(t) 初始 represents the initial window size of the t-th sliding window, f(t-1) and a(t-1) represent the fluctuation frequency and fluctuation amplitude determined by the time series analysis of the water quality parameters of the t-1-th sliding window, respectively, and λ1 and λ2 are adaptive adjustment coefficients; Calculate the window size change rate corresponding to the initial window size of the current sliding window: Where V w (t) represents the window size change rate, w(t-1) represents the window size of the t-1th sliding window; If the window size change rate of the current sliding window does not exceed the preset change rate threshold, the initial window size is used as the window size of the current sliding window; as well as If the window size change rate of the current sliding window exceeds the change rate threshold, the window size of the current sliding window will be smoothly calculated by the following formula: w(t)=w(t-1)+η·(w(t) 初始 -w(t-1)), Where w(t) represents the window size of the tth sliding window, and η is the smoothing coefficient.

4. The method according to claim 3, wherein: The sliding time series data in each sliding window is decomposed into a plurality of period amplitude features corresponding to different period amplitudes by using a discrete wavelet transform method, and each of the period amplitude features is weightedly fused to obtain a corresponding window comprehensive amplitude feature, including: Perform discrete wavelet transform on the sliding time series data of the current sliding window to obtain multiple sub-band representations: The time series data in the t-th sliding window is represented by x={x1,x2,...,x w(t) }, perform discrete wavelet transform on x to obtain multiple subbands representing X s ={X s1 ,X s2 ,...,X sm }, m is the number of subbands, each subband X si represents the amplitude of the ith cycle s i The characteristic component under x j represents the electrochemical parameter value of water quality at the jth time step in x, X si (t) represents the t-th sliding window in the period amplitude s i The characteristic component under ψ si (j) represents the periodic amplitude s of the tth sliding window i The wavelet basis function under t; j represents the time step offset, which is used to describe the weight of the eigenvalue of the jth time step of the tth sliding window; Calculate the energy value of each period amplitude feature: In the formula, E si (t) represents the period amplitude s in the t-th sliding window i Energy value of the lower subband feature; Indicates subband X si The sum of squares of eigenvalues ​​at all time steps in the t-th sliding window; Based on the significance metric, the comprehensive period significance of each period amplitude feature is calculated. The significance metric includes period strength and relative energy proportion: θ si (t)=l1·P si (t)+l2·REP si (t), P si (t)=A si (t)·E si (t), HAS si (t)=max{X si (j)}-min{X si (j)}, Where P si (t) represents the period amplitude s in the t-th sliding window i The periodic intensity under A si (t) represents the period amplitude s in the t-th sliding window i The fluctuation amplitude of the lower subband feature is defined as the maximum value of the subband feature value max{X si (j)} and the minimum value min{X si (j)}; REP si (t) represents the period amplitude s in the t-th sliding window i The relative energy proportion of the characteristic components under ; is a normalized term, which represents the sum of the energy values ​​of the sub-band features of all periodic amplitudes in the t-th sliding window; θ si (t) represents the period amplitude s in the t-th sliding window i The comprehensive period significance of , l1 and l2 represent the weight coefficients of the corresponding significance measurement indicators; Based on the comprehensive period significance, each of the period amplitude features is weighted and fused to obtain the corresponding window comprehensive amplitude feature: Where F(t) represents the window comprehensive amplitude feature corresponding to the t-th sliding window.

5. The method according to claim 4, wherein: The water quality anomaly identification model comprises a cascaded temporal feature modeling module and a feature anomaly identification module; The time series feature modeling module is used to model the input electrochemical time series sensing features, extract the dynamic change pattern and time dependency of the features of the monitored water area block, and obtain the corresponding pattern time series dependency features; The feature anomaly identification module is used to process the pattern time-dependent characteristics to predict whether there is a risk of water quality anomaly in the corresponding monitored water area.

6. The method according to claim 5, wherein: The temporal feature modeling module comprises an LSTM layer, a TCN layer and a feature fusion layer, wherein the LSTM layer and the TCN layer are connected in parallel to the feature fusion layer; The feature anomaly recognition module includes an MLP layer, a Softmax layer and an anomaly classification layer; The LSTM layer is used to extract long-term trend features in electrochemical time-series sensing features through a gating structure to capture the trend changes and long-term dependencies of water quality parameters: h g,t =σ1(W h ·[h g,t-1 ,V g,t ]+b h ), H g =[h g,1 ,h g,2 ,...,h g,T ], In the formula, h g,t Indicates the LSTM layer monitors the water area The hidden state output at the t-th sliding window, W h represents the weight matrix of the LSTM layer, h g,t-1 Indicates the LSTM layer monitors the water area The hidden state output at the t-1th sliding window; V g,t Indicates monitored water area The electrochemical input feature vector in the tth sliding window, b h represents the bias term, σ1(·) represents the activation function of the LSTM layer; T represents the total number of time steps corresponding to the electrochemical temporal sensing feature, H g Indicates that the LSTM layer is used to monitor the water area The extracted long-term trend features; The TCN layer extracts local fluctuation pattern features in electrochemical time-series sensing features by stacking multiple layers of convolution operations to capture short-term fluctuations and local change patterns of water quality parameters: y g,t =W p *V g,t +b p , AND g =[and g,1 ,and g,2 ,...,and g,T ], In the formula, y g,t Indicates the TCN layer for monitoring water area blocks The convolution output feature of the t-th sliding window, W p and b p They represent the convolution kernel weight and bias term of the TCN layer respectively, * represents a one-dimensional convolution operation; Y g Indicates that the LSTM layer is used to monitor the water area The extracted local fluctuation pattern features; The feature fusion layer is used to perform weighted fusion on the output result of the LSTM layer and the output result of the TCN layer to obtain the corresponding pattern time-dependent features: F g =α·H g +(1-a)·Y g , In the formula, F g Indicates monitoring water area The pattern time-dependent features, α is the feature fusion coefficient; The feature anomaly recognition module is used to process the pattern time-dependent features through the first MLP layer, and predict the probability value of the risk of water quality anomaly in the monitored water area through the Softmax layer: In the formula, Indicates monitoring water area The output feature representation of the fth hidden layer is W (f) and b (f) represents the weight matrix and bias term of the fth hidden layer; Indicates monitoring water area The output feature representation of the f-1th hidden layer is that in the calculation of the first hidden layer P g Indicates monitored water area The abnormal probability of water quality abnormality risk; L1 is the total number of hidden layers in the first MLP layer, σ2(·) represents the activation function of the hidden layer, and Represent the weight matrix and bias term of the Softmax layer respectively, The last hidden layer of the first MLP layer is used to monitor the water area. Output feature representation, Softmax(·) represents the Softmax activation function; The anomaly classification layer is used to compare the predicted probability value of the water quality anomaly risk with a preset anomaly probability threshold, and determine whether the corresponding monitored water area has a water quality anomaly risk based on the comparison result.

7. The method according to claim 1, wherein: The abnormal event recognition model includes a cascaded multi-scale convolution module and an event type classification module; The multi-scale convolution module is used to extract the multi-scale time series features corresponding to the pattern time series dependency features of the risk monitoring water area block: In the formula, Indicates monitoring block The convolution output feature representation at the e-th convolution kernel scale includes the local fluctuation and trend characteristics at this convolution kernel scale; and They represent the convolution kernel weight matrix and bias term of the e-th convolution kernel scale respectively; F D Indicates risk monitoring waters The temporal dependency characteristics of the pattern; σ3(·) represents the activation function of the multi-scale convolution module; M D Indicates risk monitoring waters The comprehensive multi-scale feature representation, μ e represents the scale fusion weight of the e-th convolution kernel scale, and U represents the total number of convolution kernel scales in the multi-scale convolution module; The event type classification module includes an attention layer, a second MLP layer, a confidence calculation layer and an abnormal event classification layer; the attention layer is used to calculate the weight distribution matrix for the multi-scale time series features to determine the corresponding attention features: In the formula, Indicates risk monitoring waters The attention weight of the e-th convolution kernel scale; and Represents the attention weight matrix and bias term of the e-th convolution kernel scale; Indicates that the attention layer monitors the water area for risks Output attention features; The second MLP layer is used to perform nonlinear mapping and feature conversion processing on the attention features, and determine the confidence matrix of the risk monitoring water area block for a preset risk event type set through the confidence calculation layer: In the formula, Indicates risk monitoring waters The output feature representation of the rth hidden layer; W (r) and b (r) represents the weight matrix and bias term of the rth hidden layer; Indicates risk monitoring waters The output feature representation of the r-1th hidden layer is that in the calculation of the first hidden layer L2 is the total number of hidden layers in the second MLP layer, σ4(·) represents the activation function of the hidden layer, The last hidden layer of the second MLP layer is used to monitor the risk water area. The output comprehensive risk characterization; and Represent the weight matrix and bias term of the confidence calculation layer, P D Indicates risk monitoring water area The confidence matrix includes the confidence scores for each risk event type in the risk event type set; The abnormal event classification layer is used to process the confidence matrix to identify the water quality abnormal risk event type corresponding to the risk monitoring water area block.

8. The method according to claim 7, wherein: The loss function of the abnormal event recognition model is: In the formula, represents the loss function of the abnormal event recognition model, represents the classification cross entropy loss term, represents the adaptive balancing loss term, and represents the time series consistency loss term, β1, β2, β3 represent the weight coefficients of each corresponding loss term respectively; N represents the total number of samples in the risk data sample set for the risk monitoring waters, C represents the total number of water quality abnormal risk event types in the risk event type set, y d,c represents the actual label of the d-th risk data sample on the c-th risk event type. If the block belongs to the c-th event type, then y d,c =1, otherwise y d,c =0; p d,c represents the confidence that the d-th risk data sample is predicted by the abnormal event recognition model as the c-th risk event type; γ is the adaptive balance coefficient, which is used to adjust the balance between feature weight and classification confidence; w d,c represents the feature weight distribution value of the d-th risk data sample on the c-th risk event type; Q represents the total time steps of the multi-scale time series feature, Represents the feature representation of the qth time step of the dth risk data sample at the eth convolution kernel scale; Represents the feature representation of the d-th risk data sample at the q-1th time step at the e-th convolution kernel scale.

9. An IoT intelligent monitoring and alarm system for water quality detection, comprising: A water quality parameter sampling unit is used to sample the water quality electrochemical data of the corresponding monitored water area based on each IoT node; A plurality of IoT nodes are distributedly arranged in different blocks in the monitored water area, and each IoT node is deployed with an electrochemical sensor; the sensing parameter types of the water quality electrochemical data include pH value, DO content and TOC content of the water body; A time series feature extraction unit, used to extract the electrochemical time series sensing features corresponding to each of the water quality electrochemical data; A water quality anomaly identification unit, used to input each of the electrochemical time-series sensing features into a water quality anomaly identification model, extract the mode time-series dependency features corresponding to the electrochemical time-series sensing features, so as to predict whether there is a water quality anomaly risk in the corresponding monitored water area; An abnormal event identification unit is used to input the pattern time-series dependency characteristics corresponding to each risk monitoring water area block predicted to have abnormal water quality risk into an abnormal event identification model to identify the type of abnormal water quality risk event corresponding to the risk monitoring water area block; The water quality alarm unit of the water area is used to perform the water quality pollution alarm operation of the water area according to the block identification of each risk monitoring water area block and the corresponding water quality abnormal risk event type.

Images