Remote monitoring and early warning system for water conservancy monitoring terminal

Abstract

The invention 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 calculation module, an information transmission module, a central analysis module and an execution module. According to the method, the accuracy and continuity of water conservancy monitoring data are ensured through the technologies of timestamp detection, interpolation calculation, data cleaning, median filtering and the like; according to the method, abnormal changes in key water conservancy data such as water level, flow and pressure can be identified through real-time data analysis, and abnormal fluctuations of a trend part or a periodic part can be found in time in combination with seasonal decomposition and abnormal value analysis; according to the invention, the problems of bandwidth limitation and uneven network load in the traditional remote monitoring system are solved by dynamically adjusting the communication bandwidth, frequency and priority management. Particularly, under the conditions of multiple nodes and large data volume, data transmission delay and loss can be effectively avoided, and the real-time performance and stability are ensured.

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 related to the sustainable development of national economy, society and ecology. The construction of water conservancy projects (such as reservoirs, dams, sluice gates, etc.) and the dispatching 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, there are many problems with traditional water conservancy monitoring methods:

[0003] First, most of the existing water conservancy monitoring systems collect data periodically, with a low frequency of data collection, and cannot achieve real-time monitoring. Some high-risk waters, such as reservoirs and rivers, are prone to emergencies due to data lags. When disasters such as floods and landslides occur, it is difficult to respond in time;

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

[0005] Finally, the existing early warning mechanism of the water conservancy monitoring system is often relatively simple, relying only on traditional threshold detection methods, ignoring complex data trends, seasonal changes and multi-dimensional analysis, resulting in low accuracy of early warning information and difficulty in achieving refined management. The emergency response mechanism is also often inflexible and unable to provide specific response measures 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 comprises 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 meter, flow meter, pressure sensor, water quality sensor, etc., including:

[0011] The water level measured by the 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 recording the time of data collection and a monitoring area number recording 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. 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 by 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, and median filtering is performed to obtain filtered data.

[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. Two abnormality criteria are set according to 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.

[0021] The information transmission module transmits the water conservancy data processed by the edge computing module to the central analysis module through the long-distance transmission protocol. At the same time, the bandwidth of the communication network and the data transmission tasks of each monitoring area are evaluated in real time. The communication load is calculated and the data transmission bandwidth is adjusted dynamically. According to the abnormal judgment criteria, early warning signals and data transmission task priorities, the communication resources for data transmission in each monitoring area are adjusted. The network resource allocation is optimized by calculating the bandwidth adjustment amount and the frequency adjustment amount. The information transmission module transmits the water conservancy data processed by the edge computing module to the central analysis module through the communication network based on the 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 trend part is extracted by smoothing the data through the local weighted regression algorithm, and then the trend part is subtracted from the water conservancy data to obtain the mixed data of the periodic part and the residual noise part. The mixed data of the periodic part and the residual noise part is analyzed by fast Fourier transform to extract the periodic part.

[0024] As a preferred embodiment of the present invention, an outlier analysis is performed on the trend part and the periodic part, and an outlier analysis is performed on the trend part and the periodic part, and the calculation of the anomaly index is used to identify whether the trend part has an abnormal increase or decrease, and whether the periodic part has an abnormal fluctuation. The anomaly index is calculated to analyze the abnormal conditions of the water level, flow, 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 abnormal signals. If only a warning signal is generated, a first-level warning is output; if only an alarm signal is generated, a second-level warning is output; if both a warning signal and an alarm signal are generated, a third-level warning is output.

[0026] 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.

[0027] After detecting the second or third level warning information, the on-site alarm device is triggered and relevant personnel are notified via SMS, email or APP for quick response.

[0028] Depending on the specific warning level, different emergency response strategies are selected:

[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. The present invention uses the edge computing module to perform preliminary processing on the water level, flow, pressure, water quality and other data collected by the sensor, solving the problems of data missing, 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. The present invention can identify abnormal changes in key water conservancy data such as water level, flow, pressure, etc. through real-time data analysis, such as early signs of natural disasters such as floods and droughts. Combined with seasonal decomposition and outlier analysis, it can timely detect abnormal fluctuations in trend parts or cyclical parts and trigger early warning signals. This enables managers to take effective countermeasures before potential risks occur, thereby improving the safety and efficiency of water resources management.

[0035] 3. The present invention solves the problems of bandwidth limitation and uneven network load in traditional remote monitoring systems by dynamically adjusting communication bandwidth, frequency and priority management. Especially in the case of multiple nodes and large amounts of data, it can effectively avoid data transmission delays and losses, and ensure real-time performance and stability. In addition, dynamic optimization of network load and communication resources can improve the performance of the overall system, ensure that key monitoring data can be transmitted first in an emergency, and enhance the emergency response capability of the system. 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 It is a seasonal decomposition schematic diagram of the present invention. DETAILED DESCRIPTION

[0039] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] See also Figure 1 As shown, 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.

[0041] 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, water quality sensor, etc., 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 timestamp sequence, the time interval Δt of all two adjacent timestamps is calculated. When it is detected that the time interval Δt between two adjacent timestamps t1 and t2 is greater than the preset threshold ΔtMax, the water conservancy data corresponding to the timestamps t1 and t2 are determined. and The data recording time interval is too large, that is, and There are missing data between and Fill in missing values make

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

[0054] Every preset statistical period T, the timestamp detection and interpolation calculation traverse 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 The missing values ​​of all water conservancy data collected within the statistical period T are supplemented.

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

[0056] The preset sliding window width k is Get the first k water conservancy data And the last k water conservancy data Create a water conservancy data sequence with a length of 2k+1

[0057] Through the median filter formula 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 instantaneous changes. Median filtering can effectively eliminate this noise.

[0059] Finally, outlier detection is used to analyze data abnormal changes. By formula Calculate the water conservancy data x of monitoring area i i The first abnormal criterion α1(x i ) and the second abnormality criterion α2(x i ), where water conservancy data xi Including water level depth h i , water flow Q i , hydrostatic pressure P i , dissolved oxygen concentration DO i and chemical oxygen demand COD i . Where x i Max is water conservancy data The maximum threshold value, x i Min is water conservancy data The minimum threshold of The filtered data The mean of The filtered data The standard deviation of , where β is a preset excess constant, which is an integer greater than 0.

[0060] When α1 is detected i Greater than the preset threshold α1Max or α2 i When it is greater than the preset threshold α2Max, it is determined that water conservancy data is collected The monitoring area i detects an abnormal value, marks the water conservancy monitoring terminal in the monitoring area i as an abnormal terminal, and outputs the water conservancy data x through the water conservancy monitoring terminal in the monitoring area i 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 state 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 abnormal criterion α1 of the water conservancy monitoring terminal in each monitoring area i i , The second abnormality criterion α2 i , 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 abnormality 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 ΔBW(i) of the water conservancy monitoring terminal in each monitoring area i. When ΔBW(i) is greater than 0, it means the bandwidth is increased; when ΔBW(i) is less than 0, it means the bandwidth is reduced. The specific value of the reduction or increase is equal to |ΔBW(i)|.

[0067] By formula Calculate the communication network frequency adjustment Δf(i) of the water conservancy monitoring terminal in each monitoring area i. BaseFrequency is the preset base 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 |Δf(i)|.

[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, the water conservancy data is decomposed into three parts on a time scale through seasonal decomposition: trend part, periodic part and residual noise. Subsequently, outlier analysis is performed on the trend part and the periodic part to identify whether the trend part contains abnormally rising or falling data, and whether the periodic part contains abnormal fluctuation information, and whether to generate alarm information based on the identification results.

[0070] It should be noted that the trend part represents the long-term change trend of water conservancy data. It reflects the overall directional change of the data over time, that is, whether the data is rising, falling or remaining stable. The trend part is usually used to identify long-term change patterns, such as water levels rising or falling year by year. If the trend part shows an abnormal increase or abnormal decrease in a certain period of time, it usually means that there are significant changes or problems in the water conservancy system. For example, an abnormally rapid rise or fall in water levels may mean floods, droughts, equipment failures and other problems. The periodic part represents the regular fluctuations in the time series due to seasonal or periodic factors. These fluctuations are usually caused by seasonal changes in nature, such as changes in temperature and seasonal changes in precipitation. The periodic part can help identify seasonal patterns in water conservancy data, such as increased water flow in summer or decreased water levels in winter. Abnormal fluctuations in the periodic part may indicate abnormal changes in natural conditions, or abnormal responses of equipment and facilities to periodic fluctuations. For example, if the precipitation in a certain year is significantly different from previous periodic fluctuations, it may indicate abnormal climate or failure of some equipment. The residual part is the part of the time series data that remains after removing the trend and cyclical components, which usually contains random noise, occasional events or unpredictable fluctuations. Residual noise reflects the random factors in the data, which may be caused by accidental factors or data 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, each water conservancy data is generated 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

[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] 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] Subsequently, the trend part was 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. Where ω (t, tk) 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 The slope of the fitted line is obtained 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 Wherein τ1, τ2 and τ3 are respectively the preset slope criterion, amplitude criterion and frequency criterion; wherein r1, r2 and r3 are all 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 The 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 early warning information are output according to the generation of early warning signals and alarm signals:

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

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

[0086] If the water conservancy monitoring terminal in a certain monitoring area i generates both a warning signal and an alarm signal within the preset period T, a third-level warning information 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 executing corresponding response operations after receiving the first, second and third level warning information sent by the central analysis module.

[0089] After detecting the second or third level warning information, the on-site alarm device is triggered and relevant personnel are notified via SMS, email or APP for quick response.

[0090] Depending on the specific warning level, different emergency response strategies are selected:

[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 exclude 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 only for the purpose of describing specific embodiments 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 combination of one or more of the associated listed items and all possible combinations, and includes these combinations;

[0096] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. 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 a water conservancy monitoring terminal, 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; Process missing water conservancy data and check the data timestamps of each monitoring point. If the interval between adjacent timestamps is too large, it is determined that data is missing and the missing data is supplemented by 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 on 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 through a long-distance transmission protocol; at the same time, it evaluates the bandwidth of the communication network and the data transmission tasks of each monitoring area in real time; calculates the communication load and dynamically adjusts the data transmission bandwidth; adjusts the communication resources for data transmission in each monitoring area according to abnormal judgment criteria, early warning signals and data transmission task priorities; and optimizes network resource allocation by calculating bandwidth adjustment and frequency adjustment. 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 part and the periodic part, calculate the anomaly index, and analyze the abnormal conditions 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, water quality sensor, etc., 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; where i is the number of the monitoring area, i = 1, 2, 3, ..., n; where 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 the second or third level warning information, the alarm device on site is triggered and relevant personnel are notified via SMS, email or APP for quick response; Choose different emergency response strategies according to the specific warning level: Level 1 warning: Immediately notify staff and initiate manual intervention; Second level warning: start automatic regulating equipment such as sluice gates and water pumps; 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 P t i , dissolved oxygen concentration and chemical oxygen demand Where t is the timestamp, i is the monitoring area number; In the timestamp sequence, the time interval Δt of all two adjacent timestamps is calculated; when it is detected that the time interval Δt of two adjacent timestamps t1 and t2 is greater than the preset threshold ΔtMax, the water conservancy data corresponding to the timestamps t1 and t2 are determined. and The data recording time interval is too large, that is, and There are missing data between and Fill in missing values make By formula Calculate the timestamp t' and missing value of the missing value Specific value of Every preset statistical period T, the timestamp detection and interpolation calculation traverse all monitoring area IDs i, all timestamps t and all water conservancy data, including water level depth Water flow Hydrostatic pressure P t i , dissolved oxygen concentration and chemical oxygen demand The missing values ​​of all water conservancy data collected within the statistical period T are supplemented.

4. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 1, characterized in that: The specific process of noise suppression of the collected water conservancy data through data cleaning and median filtering is as follows: The preset sliding window width k is Get the first k water conservancy data And the last k water conservancy data Create a water conservancy data sequence with a length of 2k+1 Through the median filter formula 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. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 1, characterized in that: The specific process of outlier detection for water conservancy data is as follows: For each filtered data By formula Calculate the water conservancy data x of monitoring area i i The first abnormal criterion α1(x i ) and the second abnormality criterion α2(x i ), where water conservancy data x i Including water level depth h i , water flow Q i , hydrostatic pressure P i , dissolved oxygen concentration DO i and chemical oxygen demand COD i ; where x i Max is water conservancy data The maximum threshold value, x i Min is water conservancy data The minimum threshold of The filtered data The mean of The filtered data The standard deviation of , where β is a preset excess constant, which is an integer greater than 0; When α1 is detected i Greater than the preset threshold α1Max or α2 i When it is greater than the preset threshold α2Max, it is determined that water conservancy data is collected The monitoring area i detects an abnormal value, marks the water conservancy monitoring terminal in the monitoring area i as an abnormal terminal, and outputs the water conservancy data x through the water conservancy monitoring terminal in the monitoring area i i early warning signal.

6. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 1, 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 network available bandwidth AvailableBW of the entire communication network; at the same time, record in real 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; By formula Calculate the communication network load Load(i) of the water conservancy monitoring terminal in each monitoring area i; Comprehensively evaluate the first abnormal criterion α1 of the water conservancy monitoring terminal in each monitoring area i i , The second abnormality criterion α2 i , 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 abnormality criteria: Where priority(i) is the data transmission task priority of the water conservancy monitoring terminal in monitoring area i; Furthermore, through the formula Calculate the communication bandwidth adjustment ΔBW(i) of the water conservancy monitoring terminal in each monitoring area i; when ΔBW(i) is greater than 0, it means that the bandwidth is increased; When ΔBW(i) is less than 0, it means the bandwidth is reduced; the specific value of reduction or increase is equal to |ΔBW(i)|; By formula Calculate the communication network frequency adjustment Δf(i) of the water conservancy monitoring terminal in each monitoring area i; where BaseFrequency is the preset base frequency, and ω is the preset load adjustment factor; when Δf(i) is greater than 0, it means increasing the frequency; when Δf(i) is less than 0, it means reducing the frequency; the specific value of the reduction or increase is equal to |Δf(i)|.

7. A remote monitoring and early warning system for a water conservancy monitoring terminal according to claim 1, characterized in that: The specific process of decomposing water conservancy data into trend part, periodic part and residual noise on the time scale is: 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 includes: water level depth Water flow Hydrostatic pressure P t i , 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 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 Represents the unexplained part of the water resources data time series, including data noise and abnormal fluctuations; Subsequently, the trend part was 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; k is the timestamp number in the water conservancy data time series, and k = 1, 2, 3, ..., m; ω (t, tk) is a 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: Wherein 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 The slope of the fitted line is obtained Get the periodic part T t i The maximum amplitude FFT (T t i )max and its corresponding frequency f(T t i )max; through the formula Calculate the anomaly index Wherein τ1, τ2 and τ3 are preset slope criteria, amplitude criteria and frequency criteria respectively; wherein r1, r2 and r3 are 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 P t i , dissolved oxygen concentration and chemical oxygen demand The abnormal index β(P t i ), and When the abnormal index β(P t i ), or When it is greater than the preset threshold, a corresponding alarm signal is generated.

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

Images