A remote monitoring and early warning system for water conservancy monitoring terminal

Abstract

The application relates to the technical field of water conservancy monitoring, in particular to a remote monitoring and early warning system for a water conservancy monitoring terminal, which comprises a data acquisition module, an edge computing module, an information transmission module, a central analysis module and an execution module. The accuracy and continuity of water conservancy monitoring data are ensured through technologies such as timestamp detection, interpolation calculation, data cleaning and median filtering; the application can identify abnormal changes in key water conservancy data such as water level, flow and pressure through real-time data analysis, and timely discover abnormal fluctuations in trend parts or periodic parts through seasonal decomposition and abnormal value analysis; the application solves the problems of bandwidth limitation and uneven network load in traditional remote monitoring systems through dynamic adjustment of communication bandwidth, frequency and priority management. Especially in the case of multiple nodes and large data volume, the application can effectively avoid data transmission delay and loss, and ensure real-time performance and stability.

Description

Technical Field

[0001] The present invention relates to the technical field of water conservancy monitoring, and in particular to a remote monitoring and early warning system for a water conservancy monitoring terminal. Background Art

[0002] With the rapid development of society and economy, water resources have become one of the key factors affecting the sustainable development of a country's economy, society, and ecology. The construction of water conservancy projects (such as reservoirs, dams, and sluice gates) and the scheduling and management of water resources play a vital role in ensuring regional water security, reducing the impact of natural disasters, and promoting the rational use of water resources. However, traditional water conservancy monitoring methods have many problems:

[0003] First, most existing water conservancy monitoring systems collect data periodically, with a low frequency, and cannot achieve real-time monitoring. In some high-risk water areas, such as reservoirs and rivers, data lags, making it difficult to respond to emergencies such as floods and landslides in a timely manner.

[0004] Secondly, traditional water conservancy monitoring systems are unable to effectively deal with problems such as missing data, noise, and outliers during data collection. Especially during extreme weather or emergencies, data may be missing or inaccurate on a large scale, affecting monitoring effectiveness.

[0005] Finally, existing water monitoring systems often have simplistic early warning mechanisms, relying solely on traditional threshold detection methods and ignoring complex data trends, seasonal variations, and multi-dimensional analysis. This results in low early warning accuracy and makes it difficult to implement refined management. Emergency response mechanisms also often lack flexibility, failing to provide specific responses based on different abnormal situations.

[0006] In view of the above problems, it is necessary to propose a remote monitoring and early warning system for water conservancy monitoring terminals. Summary of the Invention

[0007] The purpose of the present invention is to solve the problems existing in the background technology and to propose a remote monitoring and early warning system for a water conservancy monitoring terminal.

[0008] The purpose of the present invention can be achieved through the following technical solutions:

[0009] A remote monitoring and early warning system for a water conservancy monitoring terminal includes a data acquisition module, an edge computing module, an information transmission module, a central analysis module, and an execution module.

[0010] The data acquisition module obtains water conservancy data of each monitoring area i through different water conservancy monitoring terminals such as water level meters, flow meters, pressure sensors, and water quality sensors, including:

[0011] Water levels measured by water level gauges in each monitoring area;

[0012] Water flow measured by flow meters in each monitoring area;

[0013] The hydrostatic pressure measured by the pressure sensors in each monitoring area;

[0014] Dissolved oxygen concentration and chemical oxygen demand measured by water quality sensors in each monitoring area.

[0015] Each water conservancy data is accompanied by a timestamp that records the time of data collection and a monitoring area number that records the spatial location of data collection.

[0016] Subsequently, the water conservancy data collected by the sensor is converted into digital signals through an analog-to-digital converter and sent to the edge computing module.

[0017] The edge computing module is responsible for the preliminary processing and analysis of water conservancy data, including data missing processing, noise suppression and outlier detection.

[0018] As a preferred method of the present invention, missing data is processed for water conservancy data. Every preset period T, the data timestamps of each monitoring point are checked. If the interval between adjacent timestamps is too large, it is determined that data is missing and the missing data is supplemented through interpolation. The interpolation calculation uses a formula to estimate the location and specific value of the missing value based on the existing data.

[0019] As a preferred method of the present invention, water conservancy data is cleaned and noise suppressed, a sliding window width k is preset, k data before and after each data point are obtained, a sequence of length 2k+1 is generated, median filtering is performed, and filtered data is obtained.

[0020] As a preferred method of the present invention, outlier detection is performed on water conservancy data, and outlier detection is performed on each filtered data set. Two outlier criteria are set based on the data's maximum value, minimum value, mean, and standard deviation. If the data exceeds the threshold, it is marked as abnormal data and a warning signal is output.

[0021] The information transmission module transmits the water conservancy data processed by the edge computing module to the central analysis module via a long-distance transmission protocol. It also assesses the communication network bandwidth and the data transmission tasks in each monitoring area in real time. It calculates the communication load and dynamically adjusts the data transmission bandwidth. It adjusts the communication resources for data transmission in each monitoring area based on anomaly criteria, warning signals, and data transmission task priorities. It optimizes network resource allocation by calculating bandwidth and frequency adjustments. The information transmission module transmits the water conservancy data processed by the edge computing module to the central analysis module via a communication network based on a long-distance transmission protocol.

[0022] The central analysis module receives the water conservancy data sent by the information transmission module, and decomposes the water conservancy data into three parts on the time scale through seasonal decomposition: trend part, periodic part and residual noise part.

[0023] The data is smoothed using a local weighted regression algorithm to extract the trend component. The trend component is then subtracted from the water conservancy data to obtain a mixture of periodicity and residual noise. This mixture of periodicity and residual noise is then analyzed using a fast Fourier transform to extract the periodic component.

[0024] As a preferred embodiment of the present invention, outlier analysis is performed on both the trend and periodic components. By calculating an anomaly index, it is determined whether the trend component exhibits abnormal increases or decreases, and whether the periodic component exhibits abnormal fluctuations. The anomaly index is calculated to analyze abnormalities in water level, flow rate, pressure, dissolved oxygen concentration, and chemical oxygen demand data. If the anomaly index exceeds a preset threshold, an alarm signal is generated.

[0025] As a preferred embodiment of the present invention, different levels of warning information are output according to the abnormal signal. If only a warning signal is generated, a level one warning is output; if only an alarm signal is generated, a level two warning is output; if both a warning signal and an alarm signal are generated, a level three warning is output.

[0026] The execution module is responsible for performing corresponding response operations after receiving the first, second and third level warning information sent by the central analysis module.

[0027] After detecting a level 2 or level 3 warning message, the on-site alarm device will be triggered and relevant personnel will be notified via SMS, email or APP for quick response.

[0028] Choose different emergency response strategies based on the specific warning level:

[0029] Level 1 warning: Immediately notify staff and initiate manual intervention.

[0030] Level 2 warning: Start sluice gates, water pumps and other automated regulating equipment.

[0031] Level 3 warning: triggers public safety alarm and conducts large-scale evacuation or emergency response.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. This invention uses an edge computing module to perform preliminary processing on water level, flow, pressure, water quality and other data collected by sensors, solving the problems of data loss, noise interference and outliers in traditional monitoring. Through technologies such as timestamp detection, interpolation calculation, data cleaning and median filtering, the accuracy and continuity of water conservancy monitoring data are ensured; this enables water conservancy data to more accurately reflect the real-time water conservancy status, providing a more reliable basis for decision-making;

[0034] 2. Through real-time data analysis, this invention can identify abnormal changes in key water resources data, such as water level, flow, and pressure, which can be early warning signs of natural disasters such as floods and droughts. Combined with seasonal decomposition and outlier analysis, it can promptly detect abnormal fluctuations in trend or cyclical components, triggering early warning signals. This enables managers to take effective countermeasures before potential risks occur, thereby improving the safety and efficiency of water resource management.

[0035] 3. This invention solves the problems of bandwidth limitations and uneven network load in traditional remote monitoring systems by dynamically adjusting communication bandwidth, frequency, and priority management. Especially in multi-node, large-scale data scenarios, it can effectively avoid data transmission delays and loss, ensuring real-time performance and stability. Furthermore, dynamic optimization of network load and communication resources can improve overall system performance, ensuring priority transmission of critical monitoring data in emergency situations and enhancing the system's emergency response capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings:

[0037] Figure 1 is a system block diagram of the present invention;

[0038] Figure 2 Schematic diagram of seasonal decomposition of the present invention. DETAILED DESCRIPTION

[0039] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0040] See also Figure 1 As shown, a remote monitoring and early warning system for water conservancy monitoring terminals includes a data acquisition module, an edge computing module, an information transmission module, a central analysis module and an execution module.

[0041] The data acquisition module obtains water conservancy data of each monitoring area i through different water conservancy monitoring terminals such as water level meters, flow meters, pressure sensors, and water quality sensors, including:

[0042] Water levels measured by water level gauges in each monitoring area ;

[0043] Water flow measured by flow meters in each monitoring area ;

[0044] Hydrostatic pressure measured by pressure sensors in each monitoring area ;

[0045] Dissolved oxygen concentration measured by water quality sensors in each monitoring area and chemical oxygen demand .

[0046] Where t is the timestamp, i.e. the time when the water conservancy data is collected; i is the number of the monitoring area, i=1,2,3,...,n; and n is the total number of monitoring areas.

[0047] Subsequently, the water conservancy data collected by the sensor is converted into digital signals through an analog-to-digital converter and sent to the edge computing module.

[0048] The edge computing module is responsible for the preliminary processing and analysis of water conservancy data.

[0049] First, the missing data is identified through timestamp detection, the missing values ​​are supplemented, and the location and specific value of the missing values ​​are determined through interpolation calculation.

[0050] The timestamp detection is specifically as follows:

[0051] Every preset statistical period T, for specific water conservancy data , extract all water conservancy data recorded during the statistical period The timestamps are arranged by time size to generate a timestamp sequence: {t1, t2, t3, ...}. Including water depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand .

[0052] In the time stamp sequence, the time interval Δt of all two adjacent time stamps is calculated. When it is detected that the time interval Δt of two adjacent time stamps t1 and t2 is greater than the preset threshold ΔtMax, the water conservancy data corresponding to time stamps t1 and t2 are determined. and The data recording time interval is too large, that is, and There are missing data, and Fill in missing values .make .

[0053] The interpolation calculation is specifically as follows: Calculate the timestamp t' and missing value of missing values The specific value of .

[0054] Every preset statistical period T, the timestamp detection and interpolation calculation are performed on all monitoring area IDs i, all timestamps t and all water conservancy data, including water level depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand . All missing values ​​of water conservancy data collected within the statistical period T are supplemented.

[0055] Subsequently, the noise of the collected water conservancy data is suppressed through data cleaning and median filtering. The specific process is as follows:

[0056] The preset sliding window width k is used for each water conservancy data. , get the first k water conservancy data , ,..., and the last k water conservancy data , ,..., , create a water conservancy data sequence of length 2k+1 { , ,..., ,..., , }.

[0057] Through the median filter formula =Median[{ , ,..., ,..., , For each water conservancy data Filtering is performed, where is the filtered data; Median is the median operator, which represents the median of the water conservancy data series.

[0058] It should be noted that median filtering is a nonlinear filtering method that removes noise by taking the median of a set of data. In water conservancy monitoring, high-frequency noise may appear in water conservancy data due to sensor drift or transient changes. Median filtering can effectively eliminate this noise.

[0059] Finally, abnormal value detection is used to analyze abnormal data changes. , through the formula Calculate the water conservancy data of monitoring area i The first abnormal criterion and the second anomaly criterion , including water conservancy data Including water depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand .in For water conservancy data The maximum threshold value, For water conservancy data The minimum threshold of ) is the filtered data The mean of ) is the filtered data The standard deviation of , where β is a preset excess constant, which is an integer greater than 0.

[0060] When detected Greater than the preset threshold or Greater than the preset threshold When determining the collection of water conservancy data The monitoring area i detects an abnormal value, marks the water conservancy monitoring terminal of the monitoring area i as an abnormal terminal, and outputs the water conservancy data through the water conservancy monitoring terminal of the monitoring area i. early warning signal.

[0061] The information transmission module transmits the water conservancy data processed by the edge computing module to the central analysis module through a communication network based on a long-distance transmission protocol.

[0062] Subsequently, the status of the communication network is evaluated in real time to obtain the current network available bandwidth AvailableBW of the entire communication network. At the same time, the communication bandwidth BW(i) and data transmission volume V(i) of the data transmission task of the water conservancy monitoring terminal in each monitoring area i are recorded in real time.

[0063] By formula Calculate the communication network load Load(i) of the water conservancy monitoring terminal in each monitoring area i.

[0064] Comprehensively evaluate the first abnormality criterion of the water conservancy monitoring terminal in each monitoring area i , Second abnormality criterion , warning signal output, network load, and the data transmission rate and communication bandwidth of each data transmission task, and dynamically adjust the communication resources for data transmission. The specific process is as follows:

[0065] Assign data transmission task priorities based on the first and second exception criteria: Where priority(i) is the data transmission task priority of the water conservancy monitoring terminal in monitoring area i;

[0066] Furthermore, through the formula Calculate the communication bandwidth adjustment of the water conservancy monitoring terminal in each monitoring area i .when When it is greater than 0, it means increasing the bandwidth; when When it is less than 0, it means the bandwidth is reduced. The specific value of reduction or increase is equal to .

[0067] By formula Calculate the communication network frequency adjustment Δf(i) of the water conservancy monitoring terminal in each monitoring area i. is the preset reference frequency, and ω is the preset load adjustment factor. When Δf(i) is greater than 0, it means the frequency is increased; when Δf(i) is less than 0, it means the frequency is reduced. The specific value of the reduction or increase is equal to .

[0068] The central analysis module receives the water conservancy data sent by the information transmission module.

[0069] See also Figure 2 As shown in the figure, seasonal decomposition is used to decompose water conservancy data into three components on a temporal scale: a trend component, a cyclical component, and residual noise. Subsequently, outlier analysis is performed on the trend and cyclical components to identify whether the trend component contains abnormally high or low data, and whether the cyclical component contains abnormal fluctuations. Based on the identification results, an alert is generated.

[0070] It's important to note that the trend component represents the long-term trend in water conservancy data. It reflects the overall directional change in the data over time, that is, whether the data is rising, falling, or remaining stable. The trend component is often used to identify long-term patterns of change, such as annual increases or decreases in water levels. If the trend component shows an unusual increase or decrease over a period of time, it often indicates significant changes or problems in the water conservancy system. For example, an unusually rapid increase or decrease in water levels may indicate floods, droughts, or equipment failures. The periodic component represents regular fluctuations in the time series due to seasonal or cyclical factors. These fluctuations are often caused by natural seasonal variations, such as temperature and precipitation. The periodic component can help identify seasonal patterns in water conservancy data, such as increased water flow in summer or decreased water levels in winter. Unusual fluctuations in the periodic component may indicate unusual changes in natural conditions or abnormal responses of equipment or facilities to cyclical fluctuations. For example, precipitation in a particular year that differs significantly from previous cyclical fluctuations may indicate climate anomalies or equipment failure. The residual is what remains after removing trend and cyclical components from time series data. It typically contains random noise, occasional events, or unpredictable fluctuations. Residual noise reflects the randomness in the data, which may be caused by chance or measurement errors.

[0071] For water conservancy monitoring data, seasonal fluctuations have clear periodic characteristics on different time scales such as years, quarters, and months.

[0072] The specific process of seasonal decomposition is:

[0073] First, every preset statistical period T, generate various water conservancy data Time series of water conservancy data: , where m is the total number of water conservancy data contained in the water conservancy data time series. The water conservancy data include: water level depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand .

[0074] Assume that each water conservancy data in the water conservancy data time series consists of three parts: trend part , periodic part and the residual noise part .Right now .

[0075] Among them, the trend part Represents the long-term trend of water conservancy data time series; cyclical part Represents the fluctuations in the water conservancy data time series caused by seasonal changes or periodic factors; the residual noise part It represents the unexplained part of the water conservancy data time series, including data noise and abnormal fluctuations.

[0076] Then, the trend part is extracted by smoothing the data using the locally weighted regression LOESS algorithm. , the calculation formula is: Where w is the half width of the preset time window, which controls the flexibility of the smoothing operation; For water conservancy data time series Various water conservancy data The trend points are obtained after seasonal decomposition. Where k is the timestamp number in the water conservancy data time series, and k=1, 2, 3, ..., m. is a smoothing operator.

[0077] Then, through the formula Subtract the trend part from the water conservancy data time series , get the superimposed data sequence of the periodic part and the residual noise part .

[0078] Furthermore, the superimposed data sequence of the periodic part and the residual noise part is transformed by fast Fourier transform. Perform frequency domain decomposition to obtain the amplitude and phase of each frequency component.

[0079] The fast Fourier transform formula is: Where q is the imaginary unit and q×q = -1. Where f is the frequency.

[0080] Finally, the extracted trend part and the periodic part Perform outlier analysis to identify abnormal fluctuations. Calculate the trend part using the least squares method. Get the slope of the fitted line . Get the periodic part The maximum amplitude in the preset characteristic frequency range f1 to f2 and its corresponding frequency By formula Calculate the anomaly index , where τ1, τ2 and τ3 are the preset slope criterion, amplitude criterion and frequency criterion respectively; where r1, r2 and r3 are the preset weight factors.

[0081] Every preset statistical period T, the seasonal decomposition and abnormal index calculation traverses all the number symbols i and all the water conservancy data: water level, flow, pressure, dissolved oxygen concentration and chemical oxygen demand, and obtains the water level depth of the water conservancy monitoring terminal of each monitoring area i. , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand Abnormal index 、 、 、 and .

[0082] When the abnormal index 、 、 、 or When it is greater than the preset threshold, a corresponding alarm signal is generated.

[0083] Furthermore, different levels of warning information are output according to the generation of warning signals and alarm signals:

[0084] If the water conservancy monitoring terminal in a monitoring area i only generates an early warning signal within the preset period T, a first-level early warning message is output;

[0085] If the water conservancy monitoring terminal in a monitoring area i only generates an alarm signal within the preset period T, a second-level warning message is output;

[0086] If the water conservancy monitoring terminal in a monitoring area i generates both a warning signal and an alarm signal within the preset period T, a level 3 warning message is output;

[0087] The generated first-level, second-level and third-level warning information is sent to the execution module.

[0088] The execution module is responsible for performing corresponding response operations after receiving the first, second and third level warning information sent by the central analysis module.

[0089] After detecting a level 2 or level 3 warning message, the on-site alarm device will be triggered and relevant personnel will be notified via SMS, email or APP for quick response.

[0090] Choose different emergency response strategies based on the specific warning level:

[0091] Level 1 warning: Immediately notify staff and initiate manual intervention.

[0092] Level 2 warning: Start sluice gates, water pumps and other automated regulating equipment.

[0093] Level 3 warning: triggers public safety alarm and conducts large-scale evacuation or emergency response.

[0094] It should be understood that the terms “include” and “comprising” used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0095] It should also be understood that the terms used in this disclosure are for the purpose of describing particular embodiments only and are not intended to limit the disclosure. As used in this disclosure and the claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and the claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations;

[0096] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A remote monitoring and early warning system for water conservancy monitoring terminals, comprising an edge computing module, an information transmission module, and a central analysis module, characterized in that: The edge computing module is responsible for the preliminary processing and analysis of water conservancy data, including data missing processing, noise suppression, and outlier detection; Missing data is processed for water conservancy data. The data timestamps of each monitoring point are checked. If the interval between adjacent timestamps is too large, it is determined that data is missing and the missing data is supplemented through interpolation. The interpolation calculation uses a formula to estimate the location and specific value of the missing value based on the existing data. Perform data cleaning and noise suppression on water conservancy data, preset the sliding window width k, obtain k data before and after each data point, generate a sequence of length 2k+1, perform median filtering, and obtain filtered data; Perform outlier detection on water conservancy data and each filtered data; set two anomaly criteria based on the maximum value, minimum value, mean value and standard deviation of the data; If the data exceeds the threshold, it is marked as abnormal data and an early warning signal is output; The information transmission module transmits the water conservancy data processed by the edge computing module to the central analysis module via a long-distance transmission protocol. It also evaluates the bandwidth of the communication network and the data transmission tasks of each monitoring area in real time. It calculates the communication load and dynamically adjusts the data transmission bandwidth. It adjusts the communication resources for data transmission in each monitoring area based on abnormality criteria, warning signals, and data transmission task priorities. It optimizes network resource allocation by calculating bandwidth and frequency adjustments. The central analysis module receives the water conservancy data sent by the information transmission module and decomposes the water conservancy data into trend part, periodic part and residual noise part on the time scale through seasonal decomposition; Perform outlier analysis on the trend and periodic parts, calculate the anomaly index, and analyze the anomalies of water level, flow, pressure, dissolved oxygen concentration, and chemical oxygen demand data; if the anomaly index exceeds the preset threshold, generate an alarm signal; output different levels of warning information based on the anomaly signal.

2. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 1, characterized in that: It also includes a data acquisition module and an execution module; The data acquisition module obtains water conservancy data of each monitoring area i through different water conservancy monitoring terminals such as water level meter, flow meter, pressure sensor, and water quality sensor, including: Water levels measured by water level gauges in each monitoring area ; Water flow measured by flow meters in each monitoring area ; Hydrostatic pressure measured by pressure sensors in each monitoring area ; Dissolved oxygen concentration measured by water quality sensors in each monitoring area and chemical oxygen demand ; Where t is the timestamp, i.e. the time when the water conservancy data is collected; i is the number of the monitoring area, i=1,2,3,...,n; n is the total number of monitoring areas; Subsequently, the water conservancy data collected by the sensor is converted into digital signals through an analog-to-digital converter and sent to the edge computing module; The execution module is responsible for executing corresponding response operations after receiving the first, second and third level warning information sent by the central analysis module; After detecting a level 2 or 3 warning signal, the on-site alarm device is triggered and relevant personnel are notified via SMS, email or APP for quick response; Choose different emergency response strategies based on the specific warning level: Level 1 warning: Immediately notify staff and initiate manual intervention; Level 2 warning: Activate sluice gates and water pump automatic regulating equipment; Level 3 warning: triggers public safety alarm and conducts large-scale evacuation or emergency response.

3. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 1, characterized in that: The specific process of missing data processing for water conservancy is as follows: Every preset statistical period T, for specific water conservancy data , extract all water conservancy data recorded during the statistical period The timestamps are arranged by time size to generate a timestamp sequence: {t1, t2, t3, ...}; the water conservancy data Including water depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand ; Where t is the timestamp and i is the monitoring area number; In the time stamp sequence, calculate the time interval Δt of all two adjacent time stamps; when it is detected that the time interval Δt between two adjacent time stamps t1 and t2 is greater than the preset threshold ΔtMax, determine the water conservancy data corresponding to time stamps t1 and t2 and The data recording time interval is too large, that is, and There are missing data, and Fill in missing values ;make ; By formula Calculate the timestamp t' and missing value of missing values Specific value of Every preset statistical period T, the timestamp detection and interpolation calculation are performed on all monitoring area IDs i, all timestamps t and all water conservancy data, including water level depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand ; Supplement missing values ​​for all water conservancy data collected within the statistical period T.

4. The remote monitoring and early warning system for water conservancy monitoring terminals according to claim 1 is characterized in that: The specific process of noise suppression of collected water conservancy data through data cleaning and median filtering is as follows: The preset sliding window width k is used for each water conservancy data. , get the first k water conservancy data , ,..., and the last k water conservancy data , ,..., , create a water conservancy data sequence of length 2k+1 { , ,..., ,..., , }; Through the median filter formula =Median[{ , ,..., ,..., , For each water conservancy data Filtering is performed, where is the filtered data; Median is the median operator, which represents the median of the water conservancy data series.

5. The remote monitoring and early warning system for water conservancy monitoring terminals according to claim 1 is characterized in that: The specific process of outlier detection for water conservancy data is as follows: For each filtered data , through the formula ; Calculate the water conservancy data of monitoring area i The first abnormal criterion and the second anomaly criterion , including water conservancy data Including water depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand ;in For water conservancy data The maximum threshold value, For water conservancy data The minimum threshold of ) is the filtered data The mean of ) is the filtered data The standard deviation of , where β is a preset over-limit constant, which is an integer greater than 0; When detected Greater than the preset threshold or Greater than the preset threshold When determining the collection of water conservancy data The monitoring area i detects an abnormal value, marks the water conservancy monitoring terminal of the monitoring area i as an abnormal terminal, and outputs the water conservancy data through the water conservancy monitoring terminal of the monitoring area i. early warning signal.

6. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 5, characterized in that: The specific process of real-time evaluation of the communication network status and dynamic adjustment of data transmission bandwidth is as follows: Obtain the current available network bandwidth AvailableBW of the entire communication network; at the same time, record the communication bandwidth BW(i) and data transmission volume V(i) of the data transmission task of the water conservancy monitoring terminal in each monitoring area i in real time; By formula Calculate the communication network load Load(i) of the water conservancy monitoring terminal in each monitoring area i; Comprehensively evaluate the first abnormality criterion of the water conservancy monitoring terminal in each monitoring area i , Second abnormality criterion , warning signal output, network load, and the data transmission rate and communication bandwidth of each data transmission task, and dynamically adjust the communication resources for data transmission. The specific process is as follows: Assign data transmission task priorities based on the first and second exception criteria: ; priority(i) is the data transmission task priority of the water conservancy monitoring terminal in monitoring area i; By formula Calculate the communication bandwidth adjustment of the water conservancy monitoring terminal in each monitoring area i ;when When it is greater than 0, it means increasing the bandwidth; when When it is less than 0, it means reducing the bandwidth; the specific value of reduction or increase is equal to ; By formula Calculate the communication network frequency adjustment Δf(i) of the water conservancy monitoring terminal in each monitoring area i; is the preset reference frequency, ω is the preset load adjustment factor; when Δf(i) is greater than 0, it means the frequency is increased; when Δf(i) is less than 0, it means the frequency is reduced; the specific value of the reduction or increase is equal to .

7. The remote monitoring and early warning system for water conservancy monitoring terminals according to claim 1 is characterized in that: The specific process of decomposing water conservancy data into trend part, periodic part and residual noise on the time scale is as follows: Generate various water conservancy data every preset statistical period T Time series of water conservancy data: , where m is the total number of water conservancy data contained in the water conservancy data time series; the water conservancy data include: water level depth , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand ; Assume that each water conservancy data in the water conservancy data time series consists of three parts: trend part , periodic part and the residual noise part ;Right now ; Among them, the trend part Represents the long-term trend of water conservancy data time series; cyclical part Represents the fluctuations in the water conservancy data time series caused by seasonal changes or periodic factors; the residual noise part It represents the unexplained part of the water conservancy data time series, including data noise and abnormal fluctuations; Then, the trend part is extracted by smoothing the data using the locally weighted regression LOESS algorithm. , the calculation formula is: ; Where w is the half width of the preset time window, which controls the flexibility of the smoothing operation; For water conservancy data time series Various water conservancy data The trend points obtained after seasonal decomposition; where k is the timestamp number in the water conservancy data time series, and k=1, 2, 3, ..., m; is the smoothing operator; Then, through the formula Subtract the trend part from the water conservancy data time series , get the superimposed data sequence of the periodic part and the residual noise part ; The superimposed data sequence of the periodic part and the residual noise part is transformed by fast Fourier transform Perform frequency domain decomposition to obtain the amplitude and phase of each frequency component, and obtain a frequency spectrum containing characteristic information of the periodic part; The fast Fourier transform formula is: ; where q is the imaginary unit and q×q=-1; where f is the frequency.

8. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 7, characterized in that: The specific process of performing outlier analysis on the extracted trend part and periodic part is as follows: Calculate the trend component by the least squares method Get the slope of the fitted line ; Get the periodic part The maximum amplitude in the preset characteristic frequency range f1 to f2 and its corresponding frequency ; Through the formula Calculate the anomaly index , where τ1, τ2 and τ3 are the preset slope criterion, amplitude criterion and frequency criterion respectively; where r1, r2 and r3 are the preset weight factors; Every preset statistical period T, the seasonal decomposition and abnormal index calculation traverses all the number symbols i and all the water conservancy data: water level, flow, pressure, dissolved oxygen concentration and chemical oxygen demand, and obtains the water level depth of the water conservancy monitoring terminal of each monitoring area i. , water flow , hydrostatic pressure , dissolved oxygen concentration and chemical oxygen demand Abnormal index 、 、 、 and ; When the abnormal index 、 、 、 or When it is greater than the preset threshold, a corresponding alarm signal is generated.

9. The remote monitoring and early warning system for water conservancy monitoring terminals according to claim 1 is characterized in that: The specific process of outputting different levels of warning information according to the generation of early warning signals and alarm signals is as follows: If the water conservancy monitoring terminal in a monitoring area i only generates an early warning signal within the preset period T, a first-level early warning message is output; If the water conservancy monitoring terminal in a monitoring area i only generates an alarm signal within the preset period T, a second-level warning message is output; If the water conservancy monitoring terminal in a monitoring area i generates both a warning signal and an alarm signal at the same time within the preset synchronous period T, a third-level warning information is output.

Images