A sensor network-based water conservancy facility operation monitoring system

Abstract

This invention relates to the field of facility monitoring technology, specifically a water conservancy facility operation monitoring system based on sensor networks. The system includes a status screening module, an anomaly review module, an aging response module, a dynamic discrimination module, and a maintenance triggering module. In this invention, by constructing a derivative change rate sequence and identifying abrupt change points, noise data caused by occasional fluctuations can be accurately eliminated, effectively improving the stability and accuracy of structural status data. Based on the comparison of the same-period responses of multiple sets of structural and flow parameters, combined with the consistency coefficient of directional changes, quantitative identification and time-series tracking of the linkage behavior between multiple components are achieved. By analyzing the cumulative intensity changes within the operating cycle, abnormal operating trends are extracted and load abrupt change segments are located, identifying the high correlation between potential structural anomalies and operating timing. By combining historical frequency and classification index comparison with the current periodic behavior, early warning of fault risks and identification of maintenance response levels for high-frequency components are achieved.

Description

Technical Field

[0001] This invention relates to the field of facility monitoring technology, and in particular to a water conservancy facility operation monitoring system based on sensor networks. Background Technology

[0002] The field of facility monitoring technology focuses on the real-time or periodic monitoring of the operational status, structural integrity, and functional performance of man-made or naturally formed facilities. This field encompasses sensor deployment, signal acquisition and processing, status identification, early warning judgment, data analysis, and visualization, and is widely applicable to various types of engineering facilities, including bridges, tunnels, reservoirs, dams, power facilities, and transportation infrastructure. Its core objective is to achieve perception and assessment of the entire facility operation process through the construction of a complete monitoring system, thereby supporting facility safety management, operation and maintenance optimization, and risk prevention and control decisions.

[0003] The water conservancy facility operation monitoring system is used to monitor the operational status of water conservancy engineering facilities. It performs real-time monitoring and data analysis of the structural and functional status, as well as environmental impact parameters, of water conservancy facilities such as dams, gates, pumping stations, and canals. By integrating various monitoring devices and information processing equipment, the system acquires and processes operational data to identify abnormal situations or potential faults in facility operation. This assists management personnel in facility maintenance, scheduling management, and risk prevention, ensuring the safety and stability of water conservancy project operations.

[0004] Traditional monitoring systems rely solely on real-time acquisition and conventional processing of sensor data to obtain structural operating status. They lack mechanisms for identifying local abrupt changes in sensor data, making them susceptible to being obscured by abnormal interference data in complex hydrological environments. This increases the probability of misjudgment. For example, during flood control or gate opening and closing, transient water pressure changes may be misjudged as structural anomalies. Furthermore, the lack of structural analysis of the behavioral linkages between multiple components makes it impossible to identify the linkage behaviors caused by load migration or synchronous operation between structures. This results in unfocused maintenance judgments or delayed risk identification. During long-term operation, the lack of trend quantification of the evolution of operation frequency leads to a slow response to the aging trends of key components, increasing facility operation risks and reducing the accuracy of maintenance responses. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a water conservancy facility operation monitoring system based on sensor networks.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a water conservancy facility operation monitoring system based on a sensor network, the system comprising:

[0007] The status screening module acquires monitoring data from the sensor network of water conservancy facilities, filters data locations where abrupt changes occur, determines the correlation coefficient between flow data and water level data based on the data locations, marks and removes noisy data points, and obtains a group of structural status segments after disturbance removal.

[0008] Based on the structural state fragment group after filtering out disturbances, the abnormal joint review module extracts the hydraulic component operation data, calculates the response value difference and the consistency coefficient of direction change. If the response value exceeds the tolerance threshold and the direction is consistent, it is marked as linkage behavior and generates a multi-component linkage time sequence record.

[0009] Based on the multi-component linkage timing record, the aging response module retrieves the number of actuations, actuation duration, and load change rate of two types of structures, namely gate hinges and pump station shaft seals, during the linkage period. Combining the material type and service life of the equipment, it calculates the response integral intensity value and obtains the list of cumulative actuation intensity.

[0010] Based on the cumulative intensity list of actions, the dynamic discrimination module extracts time periods where the load change rate is higher than the average of the cycle, extracts the recorded action trigger time points within the time periods, marks the time periods with an overlap greater than the overlap recognition threshold as response linkage segments, and generates a mutation association time group.

[0011] As a further aspect of the present invention, the structural state segment group after screening out disturbances includes a trend stable segment number, a disturbance exclusion marker set, and a state segment time index; the multi-component linkage timing record specifically includes a linkage determination time node sequence, a component pair number matching set, and a linkage event flag column; the actuation cumulative intensity list includes an actuation load integral value set, an actuation weight distribution table, and a periodic performance response ratio; and the mutation association time group specifically refers to a mutation time label, a component number index, and a response synchronization marker.

[0012] As a further aspect of the present invention, the status screening module includes:

[0013] The data derivative submodule acquires monitoring data from the water conservancy facility sensor network. The monitoring data includes structural displacement data sequences, water pressure data sequences, and flow velocity data sequences when the facility is in operation. Continuous data segments of the three types of data are extracted according to a unified time step. The first derivative difference is calculated by dividing the difference between adjacent data of each type of data at each time point by the time step, and a derivative change sequence is constructed. The derivative differences of the three types of data at each time point are used as combined data units to obtain a multi-source derivative change group.

[0014] The rate screening submodule, based on the multi-source derivative change group, compares the adjacent differences of each type of derivative change result in the time series, calculates the difference of derivative difference per unit time to form a derivative change rate sequence, selects the rate values ​​of the center point and adjacent points in each continuous time window to perform change amplitude ratio calculation, extracts the data positions where the change amplitude exceeds the rate mutation threshold, and marks the derivative mutation corresponding to the position as an anomaly point to obtain the derivative mutation position group;

[0015] The collaborative noise reduction submodule calls the derived value mutation location group, locates the corresponding time point values ​​in the flow data sequence and water level data sequence according to the mutation time point, extracts the flow value and water level value of the three time points before and after the mutation point to form a synchronization sequence, calculates the correlation coefficient of the two sets of sequences, selects the time points with correlation coefficients lower than the collaborative response benchmark value as asynchronous points, marks the data as noise points and removes them, and obtains the structural state fragment group after the disturbance is removed.

[0016] As a further aspect of the present invention, the anomaly joint review module includes:

[0017] The data extraction submodule obtains the structural state fragment group after the disturbance is removed, extracts the structural displacement change amplitude of the gate and the diversion structure and the orifice water pressure fluctuation value of the corresponding time period, and simultaneously obtains the instantaneous flow rate increase value of the pump body position in the same time period. The three sets of structural response value sequences are combined according to a unified time step to generate a linkage response comparison set.

[0018] The value difference consistency submodule extracts the structural displacement change amplitude, water pressure fluctuation value and flow rate increase value for each data unit according to the linkage response comparison set. It calculates the response value difference of the three values ​​within the same time period, judges the change direction between values ​​in adjacent time periods, establishes a direction consistency identifier, calculates and obtains the composite linkage response quantity, compares the composite linkage response quantity with the response tolerance threshold, filters the time point combination that meets the direction consistency and the response quantity exceeds the threshold, and obtains the linkage trigger time series.

[0019] The behavior tagging submodule calls the linkage trigger time sequence, adds behavior tag labels to the linkage trigger time points in chronological order, counts the structural response value combinations corresponding to each tag event, summarizes them into a unified index sequence, and records the corresponding time points in the sequence to establish a multi-component linkage time sequence record.

[0020] As a further aspect of the present invention, the formula for calculating and obtaining the composite linkage response is specifically as follows:

[0021] ;

[0022] in, Indicates the composite linkage response quantity. This represents the normalized value indicating the magnitude of structural displacement variation between adjacent sampling points for the gate and the diversion structure. This represents the normalized value of the orifice water pressure fluctuation between adjacent sampling points. This represents the normalized value of the instantaneous flow rate increase of the pump body during that time period. This represents the normalized value of the sampling duration for that segment within the sampling period.

[0023] As a further aspect of the present invention, the aging response module includes:

[0024] The actuation extraction submodule obtains the linkage time period in the linkage time sequence record of the multi-component linkage, retrieves the actuation record data of the gate hinge and the pump station shaft seal, extracts three types of index data of each type of structure within the time period, namely the number of actuations, the duration of actuation and the load change rate, and integrates each data item in segments according to the time sequence to generate an actuation feature index set.

[0025] The response integration submodule, based on the set of actuation characteristic indicators, calls the actuation number, actuation duration, and load change rate corresponding to each actuation segment in each type of structure, and calculates the response integration intensity value by combining the material type and service life of the structure. The integration results under each structure number are accumulated in chronological order to obtain the sequence of actuation response integration quantities.

[0026] The numbering and aggregation submodule calls the actuation response integral sequence, sorts it in descending order according to the actuation response integral, and summarizes the integral intensity sequence data corresponding to the structure based on the number field to which the structure belongs, generating an actuation cumulative intensity list.

[0027] As a further aspect of the present invention, the formula for calculating and obtaining the integral intensity value of the response is specifically as follows:

[0028] ;

[0029] in, This represents the integral strength value of the response. The normalized value representing the number of actions. The normalized value representing the rate of change of load. The normalized value representing the duration of the action. The normalized value representing the number of years of use. Indicates the material type grade factor. This indicates the intensity of the influence factor in high humidity and high temperature environments.

[0030] As a further aspect of the present invention, the dynamic discrimination module includes:

[0031] The high-load extraction submodule obtains the list of cumulative intensity of actuation, extracts the normalized value of the load change rate corresponding to each actuation record, calculates the average value of the load change rate, identifies the time period above the average value, classifies and summarizes the high-load time period by number, and generates a high-load actuation time group.

[0032] The slope recognition submodule extracts the structural displacement trend curve and water pressure trend curve within the corresponding time period based on the high-load operation time group. Taking the continuous data interval in each curve segment as input, it performs linear segment fitting on the curve data, extracts the curve slope value of the fitted segment, filters the time interval where the difference between adjacent slopes is greater than the jump discrimination threshold, and establishes a slope jump segment set.

[0033] The time overlap submodule calls the slope jump segment set and extracts all time values ​​within each segment. It calls the recorded actuation trigger time point sequence, compares the two sets of time sequences, counts the number of time overlaps between each slope jump segment and the actuation trigger point, calculates the time overlap ratio of each segment, compares it with the overlap degree identification threshold, filters the segment numbers and time points that meet the conditions, and generates a mutation-related time group.

[0034] As a further aspect of the present invention, the system further includes:

[0035] Based on the mutation-related time group and the cumulative intensity list of operations, the maintenance triggering module numbers the components with the highest frequency of operations, calculates the frequency growth ratio, and if the growth ratio is higher than the facility maintenance critical frequency ratio, it extracts the component's category and the corresponding water conservancy facility failure mode index number, retrieves the corresponding item in the maintenance response table, and generates a high-frequency operation component maintenance index list.

[0036] The maintenance index list for high-frequency actuating components includes maintenance level numbers, fault mode label sets, and component number lists.

[0037] As a further aspect of the present invention, the maintenance triggering module includes:

[0038] The frequency extraction submodule, based on the mutation association time group and the cumulative intensity list of actions, identifies the component number with the most frequent actions according to the action frequency of each component number, extracts the number of actions of the number in the current period and the number of actions in the corresponding historical quarterly period, calculates the ratio of the two, and establishes a set of action frequency growth ratio values.

[0039] The grade matching submodule identifies the number whose ratio is higher than the facility maintenance critical frequency ratio based on the set of actuation frequency growth ratios, extracts the structural classification code and failure mode index number corresponding to the number, locates the maintenance grade field that matches the number in the maintenance response table, extracts the maintenance response grade and pairs it with the component number to obtain a maintenance grade matching pair.

[0040] The list generation submodule calls the maintenance level matching pair group and the cumulative intensity of operation list, arranges the component numbers with maintenance level records, establishes a correspondence table between component numbers and response levels, and summarizes the maintenance index list of high-frequency operation components.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0042] In this invention, by constructing a derivative change rate sequence and identifying the location of abrupt change points, noise data caused by occasional fluctuations can be accurately eliminated, effectively improving the stability and accuracy of structural state data. Based on the comparison of the same-period responses of multiple sets of structural and flow parameters, combined with the consistency coefficient of directional change, quantitative identification and time-series tracking of the linkage behavior between multiple components are realized, forming a traceable linkage record data chain. By analyzing the cumulative intensity changes within the operation cycle, abnormal operation trends are extracted and load abrupt change segments are located. Combined with the matching analysis of operation triggering time points, the high correlation between potential structural anomalies and operation timing is identified. By comparing the current cycle behavior with historical frequency and classification index, early warning of fault risks of high-frequency components and identification of maintenance response levels are realized, improving the dynamic identification capability and the pertinence of maintenance response of the monitoring system under multi-component linkage conditions. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a system flowchart of the present invention;

[0045] Figure 2 This is a schematic diagram of the system framework of the present invention;

[0046] Figure 3 This is a flowchart of the status screening module of the present invention;

[0047] Figure 4 This is a flowchart of the abnormal joint review module of the present invention;

[0048] Figure 5 This is a flowchart of the aging response module of the present invention;

[0049] Figure 6 This is a flowchart of the dynamic discrimination module of the present invention;

[0050] Figure 7 The flowchart for maintaining the trigger module of this invention is shown. Detailed Implementation

[0051] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0052] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0053] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0054] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0055] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0056] Please see Figure 1 A water conservancy facility operation monitoring system based on sensor network, the system includes a status screening module, an anomaly joint review module, an aging response module, a dynamic discrimination module and a maintenance triggering module;

[0057] The status screening module acquires monitoring data from the sensor network of the water conservancy facility. The monitoring data includes structural displacement data sequences, water pressure data sequences, and flow velocity data sequences when the facility is in operation. Based on the time sequence, continuous data segments are extracted, the first derivative difference of each data item in the current time period is calculated, adjacent derivative difference results are called, and a derivative change rate sequence is constructed. In the derivative change rate sequence, the rate corresponding to each time point is compared within the time window, and the data locations where abrupt changes occur at a single data point are screened. The correlation coefficient between the flow data and water level data based on the data location is determined. If there is no synchronous response trend, the time point data is marked as noise data point and the noise data points are removed to obtain a group of structural status segments after disturbance removal.

[0058] The anomaly joint review module extracts the structural displacement change amplitude and orifice water pressure fluctuation value between the gate and the diversion structure in the hydraulic components based on the structural state fragment group after filtering out disturbances. Combined with the instantaneous flow increase value of the pump body position in the same time period, a comparison set with the same period is constructed for the three sets of values. The response value difference and the direction change consistency coefficient are calculated in the comparison set respectively. If the response value exceeds the tolerance threshold and the direction is consistent, the behavior of this pair of components is marked as linkage behavior. The linkage behavior marked events are arranged in chronological order, and the corresponding time points are summarized to generate a multi-component linkage time sequence record.

[0059] The aging response module is based on the multi-component linkage timing record. It retrieves the number of actuations, actuation duration and load change rate of two types of structures, namely gate hinge and pump station shaft seal, during the linkage period. Combined with the material type and service life of the equipment, it calculates the response integral intensity value within the actuation cycle and arranges it according to the component number to obtain the cumulative actuation intensity list.

[0060] The fatigue weighting factor for metal structures is selected by referring to the slope of the SN curve for materials such as stainless steel, cast iron, and low alloy steel in the metal fatigue design code, and combining the corresponding material of the component.

[0061] The dynamic discrimination module extracts time periods where the load change rate is higher than the average period based on the cumulative intensity list of actuations. It then performs intra-segment fitting between the structural displacement trend curve and the water pressure trend curve within the time period, extracts the slope change segments of the curves, identifies time periods with slope jumps within the segments, and extracts the recorded actuation trigger time points within the time period. It compares the time overlap between the two and marks the time periods with an overlap greater than the overlap recognition threshold as response linkage segments. It records the time and component numbers to generate abrupt change association time groups.

[0062] The maintenance triggering module, based on the mutation association time group and the cumulative intensity list of operations, extracts the operation frequency in the historical quarterly cycle and the operation frequency in the current cycle for the component number with the highest operation frequency, calculates the frequency growth ratio, and if the growth ratio is higher than the facility maintenance critical frequency ratio, it extracts the component's category and the corresponding water conservancy facility failure mode index number, retrieves the corresponding item in the maintenance response table, marks the current maintenance response level, lists the over-frequency component number, and generates a high-frequency operation component maintenance index list;

[0063] The structural state segment group after disturbance removal includes trend stable segment number, disturbance removal mark set and state segment time index. The multi-component linkage time sequence record specifically includes linkage judgment time node sequence, component pair number matching set and linkage event flag column. The cumulative actuation intensity list includes actuation load integral value set, actuation weight distribution table and periodic performance response ratio. The mutation association time group specifically refers to mutation time label, component number index and response synchronization mark. The high-frequency actuation component maintenance index list includes maintenance level number, fault mode label set and component number list.

[0064] Please see Figure 2 and Figure 3 The status screening module includes a data output submodule, a rate screening submodule, and a collaborative noise reduction submodule.

[0065] The data derivative submodule acquires monitoring data from the water conservancy facility sensor network. The monitoring data includes structural displacement data sequences, water pressure data sequences, and flow velocity data sequences when the facility is in operation. Continuous data segments of the three types of data are extracted according to a unified time step. The first derivative difference is calculated by dividing the difference between adjacent data of each type of data at each time point by the time step, and a derivative change sequence is constructed. The derivative differences of the three types of data at each time point are used as combined data units to obtain a multi-source derivative change group.

[0066] The data acquisition submodule obtains monitoring data from the sensor network of the hydraulic facilities. The monitoring data consists of raw sequence data of displacement, water pressure, and flow velocity, which are recorded in real time by strain gauges, pressure sensors, and flow meters deployed on the hydraulic structure, and are processed according to a uniform time step. Extract data from 100 consecutive time points to form an observation period. For the raw data sequence corresponding to each sensor type, first obtain the first... Time and the The original value at time is taken and the difference is calculated. Then the difference is divided by . The difference in the first derivative at the corresponding time points is denoted as follows: (Difference in displacement derivative) (Water pressure conductance difference) and (Difference in velocity derivative), for example, within a certain observation period, the original displacement data sequence is... ,but The difference of the first derivative at time 1 The three types of derivative differences combine into vector data units at the same time point. The values ​​are arranged sequentially to form a multi-source derivative change group with a length of 99. An initial time should be set during the acquisition process. To ensure data alignment with the corresponding synchronous triggering time, and to perform linear interpolation to complete missing data points when generating multi-source derivative change groups, a derivative sequence group with the same time base and equal length is obtained for subsequent analysis.

[0067] The rate screening submodule is based on a multi-source derivative change group. It compares the differences between adjacent derivative change results for each type in the time series, calculates the difference of derivative difference per unit time to form a derivative change rate sequence, selects the rate values ​​of the center point and adjacent points in each continuous time window to perform change amplitude ratio calculation, extracts the data positions where the change amplitude exceeds the rate mutation threshold, and marks the derivative mutation corresponding to the position as an anomaly point to obtain the derivative mutation position group.

[0068] The rate screening submodule performs data mutation detection based on multi-source derivative change groups, targeting each type of derivative change sequence. At every moment The derivative difference between two adjacent time points is extracted and a second-order difference calculation is performed. The calculation formula is as follows: This forms the corresponding derivative rate of change sequence. , and Set the time window width to At the [number] time point, at the [number]th time point The time window at time point is With the center point rate value Calculate the ratio of the rate of change of all points within the window. When there is Greater than the set rate mutation threshold Time, i.e., mark For the point of abrupt change in the derivative, such as setting and , ,but If the criteria are met, the mutation point location is determined independently for each type of derivative, and the time points that meet the mutation criteria are stored as derivative mutation location groups, with a rate mutation threshold. The setting is based on the upper limit of sensor acquisition error. and the upper limit of the monitoring range The ratio, when using an electromagnetic current meter and its error range is... The maximum range is The reference ratio is In actual settings, considering sensitivity to environmental disturbances, the following should be considered: Set as Within the range, considering the disturbance sensitivity period of this type of structure during peak flood periods, a value of [value to be filled in] is recommended. This ultimately forms a set of anomalous time points for changes in three types of derivatives: displacement, water pressure, and flow velocity.

[0069] The collaborative noise reduction submodule calls the derivative mutation location group, locates the corresponding time point values ​​in the flow data sequence and water level data sequence according to the mutation time point, extracts the flow and water level values ​​of the three time points before and after the mutation point to form a synchronization sequence, calculates the correlation coefficient of the two sets of sequences, and filters the time points with correlation coefficients lower than the collaborative response benchmark value as asynchronous points. The data is marked as noise points and removed to obtain the structural state fragment group after the disturbance is removed.

[0070] The collaborative noise reduction submodule calls the derivative mutation location group and performs a check on each mutation time point. In the traffic data sequence and water level data sequence Searching for the corresponding time point numerical value and ,by Extract three time points forward and three backward from the center to form a time window sequence of length 7. and Calculate the Pearson correlation coefficient between the two sets of sequences. The formula for calculating the correlation coefficient is: ,in For covariance, , These are the standard deviations of flow rate and water level, respectively. If a certain abrupt change point corresponds to... If the point is not a synchronous mutation point, it is considered a noise point and is removed. The cooperative response benchmark value is... The setting is based on the correlation statistics of known synchronous abrupt change points in typical hydrological events. The setting process selects 100 sets of data on historical abrupt change events of downstream flow and downstream water level to calculate the average correlation. Standard deviation The baseline value is set at 95% confidence interval. When the flow rate sequence corresponding to a certain mutation point is {2.5, 2.6, 2.8, 3.0, 2.9, 2.7, 2.6} and the water level sequence is {1.2, 1.3, 1.4, 1.5, 1.5, 1.4, 1.3}, the calculated values ​​are... If, then it will be retained; if If the data is noisy, it will be marked as noisy data and removed. After the process is completed, the structure state fragment group with the noise interference removed will be output for subsequent processing.

[0071] Please see Figure 2 and Figure 4 The anomaly joint review module includes a data extraction submodule, a value difference consistency submodule, and a behavior marking submodule;

[0072] The data extraction submodule obtains the structural state fragment group after filtering out disturbances, extracts the structural displacement change amplitude of the gate and the diversion structure and the orifice water pressure fluctuation value of the corresponding time period, and simultaneously obtains the instantaneous flow rate increase value of the pump body position in the same time period. The three sets of structural response value sequences are combined according to a unified time step to generate a linkage response comparison set.

[0073] The data extraction submodule acquires the structural state fragment group after disturbance removal. First, it extracts the monitoring data segment located at the gate and guide structure positions from the structural displacement data sequence. This data segment consists of measuring points numbered 01 and 02 of the displacement sensor deployment points. The time step is set to... Data was extracted from the period from 18:00 to 18:05 on July 28, 2024. The displacement data are as follows: The variation amplitude of each pair of sampling points is calculated using the adjacent difference, and the variation amplitude is... And normalized, using the historical maximum response amplitude as... After normalization Get value sequence The pressure value sequence recorded at orifice pressure monitoring point 03 within the same time period is as follows: The difference in water pressure fluctuation values ​​is obtained If the difference between the upper and lower limits of the working pressure at the orifice is , Then the normalized water pressure fluctuation value have to Then acquire the instantaneous flow rate data of the pump body, and the flow rate sequence. Rated flow rate is The increase would be Normalization process get The three data items were combined into a response value comparison set, and the results are shown in the table below.

[0074] Table 1. Comparison Set of Linked Responses and Calculation Table of Linked Response Quantities:

[0075]

[0076] As shown in Table 1, the response value data is generated through a unified sampling period, normalization processing, and synchronization sequence, and the linkage response quantity is obtained through subsequent calculations. The value is used to determine the validity of the linked behavior.

[0077] The value difference consistency submodule extracts the structural displacement change amplitude, water pressure fluctuation value, and flow rate increase value for each data unit based on the linkage response comparison set. It calculates the response value difference of the three values ​​within the same time period, judges the direction of change between values ​​in adjacent time periods, establishes a direction consistency identifier, calculates and obtains the composite linkage response quantity, compares the composite linkage response quantity with the response tolerance threshold, and selects the time point combination that meets the direction consistency and whose response quantity exceeds the threshold to obtain the linkage trigger time series.

[0078] The specific formula for calculating and obtaining the composite linkage response is as follows:

[0079] ;

[0080] in, This represents the composite linkage response quantity, serving as a quantitative estimate of the linkage strength among the three types of structural response terms within the same time period. This represents the normalized value of the structural displacement variation between adjacent sampling points of the gate and the diversion structure. The original data comes from the time-series monitoring values ​​of the displacement sensor, and the normalization process is based on the historical maximum response amplitude of the structure. This represents the normalized value of the orifice water pressure fluctuation between adjacent sampling points. The original data is pressure sensor sampling data, and the normalization process is based on the difference between the upper and lower limits of water pressure fluctuation within the working cycle of this component. This represents the normalized value of the instantaneous flow rate increase of the pump body during this time period. The original data is the flow meter sampling value, and the normalization process is performed with the pump set's rated flow rate as the denominator. This represents the normalized value of the sampling duration within the sampling period, used for the time difference weighting term. The original data source is the integrated value of the interval between monitoring sampling frequencies. The normalization process is based on the maximum sampling duration. The response tolerance threshold is the acceptable limit index of the structural response behavior, usually set to 0.75 within the normalization intensity range. It is constructed based on the 75th percentile value of the response quantity in the system's historical data, or it can be calculated based on the upper quartile value of long-term operating samples.

[0081] The value difference consistency submodule applies the formula based on the values ​​of the linkage response comparison set in Table 1:

[0082] ;

[0083] The linkage strength is calculated for each data unit, and the meaning of the parameters and the calculation logic are as follows: This represents the normalized value of the structural displacement response amplitude, used to measure the degree of movement of the physical component during the sampling period. This represents the normalization of water pressure fluctuations, used to reflect the severity of hydraulic pressure changes. This represents the normalized result of the flow rate increase, used to reflect the degree to which changes in pump flow correspond to structural behavior. The logarithmic function is used to stretch the percentage of the effect of the flow rate increase on the total response, preventing... The unreasonableness of the time value The result of the square root of the period duration serves as a dynamic adjustment factor for the linkage amplitude, controlling the impact of the time span on the degree of linkage. The formula as a whole expresses the multivariate physical linkage intensity through the product of the total structural response amplitude and the pump body response elasticity term. The denominator is adjusted by taking the square root of time to achieve synergistic quantification of different physical indicators, ultimately resulting in the calculated... The value is a numerical estimate of the linkage strength. The larger the value, the stronger the synergistic trend of the three types of responses, and vice versa.

[0084] Taking T2 as an example, , , , ,but:

[0085] ;

[0086] Other time points are substituted into the calculations sequentially, as detailed in Table 1.

[0087] Response tolerance threshold The setup method is as follows: Select a total of 300 samples with complete linkage behavior labels from the historical operation data of this system, and calculate the normalized linkage response amount for each sample. The median was obtained as The 75th percentile is Therefore, the threshold is set to Used to identify linked behaviors with a strength reaching a identifiable level, all of which meet the following criteria. The time points were selected as linked trigger events. Based on the results in the table, T1 and T2 meet this condition, while other time points are excluded.

[0088] The behavior tagging submodule calls the linkage trigger time sequence, adds behavior tag labels to the linkage trigger time points in chronological order, counts the structural response value combinations corresponding to each tag event, summarizes them into a unified index sequence, and records the corresponding time points in the sequence to establish a multi-component linkage time sequence record;

[0089] The behavior tagging submodule calls the linkage trigger time sequence T1 and T2, assigning behavior identifiers B001 and B002 respectively. It records the three types of raw response data corresponding to each behavior event in chronological order: displacement value is the difference between monitoring points 01 and 02; water pressure is the difference in the same period value at point 03; and flow rate is the time difference between the pump body measuring points. The calculation process is as follows: the raw displacement difference in time period T1 is... Water pressure is Traffic is The displacement during time period T2 is Water pressure is Traffic is Finally, the above information is constructed into an event index sequence and uniformly entered into the system behavior index pool for subsequent analysis. The final behavior sequence is used to construct a time feature sequence of cross-component physical linkage and provides subsequent module calls for fatigue strength accumulation and abnormal load trend analysis. Its structure is a mapping table of time-behavior number-original response value.

[0090] Please see Figure 2 and Figure 5 The aging response module includes an action extraction submodule, a response integration submodule, and a number collection submodule.

[0091] The actuation extraction submodule obtains the linkage time period in the linkage time sequence record of multiple components, retrieves the actuation record data of the gate hinge and the pump station shaft seal, extracts three types of index data of each type of structure within the time period, namely the number of actuations, the duration of actuation and the load change rate, and integrates each data item in segments according to the time sequence to generate an actuation feature index set.

[0092] The actuation extraction submodule retrieves the linkage time period information filtered from the multi-component linkage timing records. This time period includes three linkage events occurring between 18:00 and 18:10 on July 28, 2024. These events involve gate number G01, diversion structure number F02, and pump station number P03, which are respectively located at the main control node of the regulating area. The module retrieves the component status log table and extracts actuation records within the aforementioned time period. It then calculates the number of actuations, actuation duration (seconds), and the average slope of the load change curve (unit load change rate) for each structure number. Taking gate G01 as an example, three actuations are recorded within this segment, with durations of 32s, 35s, and 29s respectively. The average actuation duration is then calculated. The load change rate is calculated from the derivative of the force value time series recorded by the sensor, and the average change rate of G01 is... To achieve unified and normalized processing, the maximum cumulative number of operations per year for each facility type is used as the benchmark. For example, if the maximum number of operations for a gate per year is 900, then the normalized number of operations is... The action time within a single segment Comparison of the maximum single actuation time of the same structure Normalized time value Similarly, the load change rate Comparison of maximum load change rate in structural design The normalized load change rate was obtained. The above three normalization results form the action feature index set for each segment. After processing, the segments are merged in order of timestamp. The structure number is matched with the timestamp to form a complete feature index, which is used to calculate the response integral intensity in the next step.

[0093] The response integral submodule, based on the set of actuation characteristic indicators, calls the actuation number, actuation duration, and load change rate corresponding to each actuation segment in each type of structure. Combined with the material type and service life of the structure, it calculates and obtains the response integral intensity value. The integral results under each structure number are accumulated in chronological order to obtain the sequence of actuation response integral quantities.

[0094] The specific formula for calculating the integral intensity value of the response is as follows:

[0095] ;

[0096] in, This represents the integral strength value of the response, which is the result of the aging risk assessment of the structure within a single operating cycle. This represents the normalized value of the number of operations. The original data is the number of operations recorded by the facility, which is normalized based on the maximum cumulative number of operations per year for similar components. This represents the normalized value of the load change rate, with data sourced from the average slope of the load curve within the operating segment, and normalized based on the maximum design load fluctuation rate of the component. The normalized value representing the duration of the actuation is derived from the actuation control timestamp record, with the maximum single actuation duration of the same structure as the normalization benchmark. The normalized value representing the service life is derived from the difference between the construction date and the current date, with the design service life as the normalization benchmark. This represents the material type grade factor, a constant set for the standard material fatigue grade (such as the steel fatigue classification in the "Code for Design of Steel Structures"), and is a positive integer ranging from 1 to 10. The intensity influence factor representing a high-humidity and high-temperature environment is derived from environmental temperature and humidity records during operation. Daily average values ​​of temperature and humidity are extracted separately, their product is calculated, and then normalized based on the historical maximum temperature and humidity product of the facility's operating environment. The unit is a dimensionless normalized value. The calculation is expressed as follows:

[0097] ;

[0098] in This represents the average temperature over 24 hours. This represents the average humidity over 24 hours, with the maximum value taken from the facility's historical operating records.

[0099] The response integral submodule calculates the response intensity based on the set of actuation feature indicators extracted from paragraph 1, and extracts the normalization times for each actuation segment. Normalized load change rate Normalized action time Service life normalization value Material grade factor With temperature and humidity environmental factors Substituting into the formula for the integral strength of the response:

[0100] ;

[0101] Taking structure number G01 as an example, the service life is 7 years, the design service life is 30 years, and the normalized service life value is... The material is Q345 steel, corresponding to a fatigue grade coefficient. Temperature and humidity during operation were recorded as 24-hour average temperature. The humidity is The product of the maximum historical temperature and humidity is Therefore, it is attributed to environmental factors. Substitute all the values ​​you have already obtained:

[0102] ;

[0103] First, calculate the step-by-step values:

[0104] molecular ;

[0105] denominator ;

[0106] The first part is ;

[0107] ;

[0108] Part Two is ;

[0109] The final integral strength of the response is:

[0110] ;

[0111] The formula structure is explained as follows: Part One This represents the frequency of structural actuation per unit time multiplied by the load variation capacity, standardized, and then square-rooted over the duration, emphasizing the weighting effect of short-duration, high-load actuation; Part Two This factor is used to reflect the effect of aging on response risk; the longer the service life and the more severe the environment, the greater the overall factor improvement; the final response integral strength. This represents the estimated risk intensity of the current structural segment under the dual factors of long-term aging and short-term high load. The larger the value, the higher the risk accumulation.

[0112] Table 2 Example of Response Integral Strength Calculation:

[0113]

[0114] Table 2 lists the integral intensity results of the single-segment actuation response under the G01 number.

[0115] The numbering and collection submodule calls the actuation response integral sequence, sorts it in descending order according to the actuation response integral, and summarizes the integral intensity sequence data corresponding to the structure based on the number field of the structure to generate an actuation cumulative intensity list.

[0116] The numbering and aggregation submodule extracts the intensity integral values ​​corresponding to all structure numbers based on the actuation response integral sequence, sorts them in descending order, and accumulates the integral values ​​of multiple time periods under the same number. The results are then summarized using the number field as the grouping primary key. The results under each number are sorted by the total accumulated intensity value, serving as the input data source for subsequent modules used for fatigue trend identification. For example, in the actuation sequence, G01 experienced 3 actuations, with integral values ​​of... , , The total is accumulated as After numbering and aggregation, the following structure is generated: {G01:0.004387, P03:0.003212, F02:0.002889}, and a list of cumulative actuation intensity is established based on this. This list is sorted from high to low according to the integral value, reflecting the risk level input sequence corresponding to the structural aging trend, for use by the subsequent abrupt change segment identification and maintenance module. The final list structure is a ternary mapping table of number—total integral value—sorting position, which is stored in the system master control index table after its establishment.

[0117] Please see Figure 2 and Figure 6 The dynamic discrimination module includes a high-load extraction submodule, a slope recognition submodule, and a time overlap submodule;

[0118] The high-load extraction submodule obtains the list of cumulative actuation intensity, extracts the normalized value of the load change rate corresponding to each actuation record, calculates the average value of the load change rate, identifies the time period above the average value, classifies and summarizes the high-load time periods by number, and generates high-load actuation time groups.

[0119] The high-load extraction submodule obtains all actuation records from the cumulative actuation intensity list and extracts the corresponding normalized load change rate value for each actuation segment in the list. Suppose the sample size is 10 data sets, numbered G01 to G10, and the recorded normalized loading rate sequence is as follows: First, the sequence is averaged to obtain the average rate of change of load for all segments. Then, each record was compared one by one. Is the value greater than the mean? Those meeting the criteria are considered high-load segments, and the obtained numbers are: G02 (0.28 ≈ mean, retained), G04 (0.33), G06 (0.41), G07 (0.36), G08 (0.29), and G10 (0.31), a total of 6 records are classified as high-load segments. The time information of these segments is aggregated to establish "high-load operation time groups", which are then classified according to their respective structure numbers. For example, G04 corresponds to the time period from 18:05:00 to 18:05:30 on 2024-07-28, and G06 corresponds to the time period from 18:07:10 to 18:08:00 on 2024-07-28. In this way, multiple sets of time series are generated with the structure number as the primary key and the time period as the value, and then uniformly input into the next identification module.

[0120] The slope recognition submodule extracts the structural displacement trend curve and water pressure trend curve within the corresponding time period based on the high-load operation time group. Taking the continuous data interval in each curve segment as input, it performs linear segment fitting on the curve data, extracts the curve slope value of the fitted segment, filters the time interval where the difference between adjacent slopes is greater than the jump discrimination threshold, and establishes a slope jump segment set.

[0121] The slope recognition submodule acquires the aforementioned high-load operation time groups, locates the structural displacement trend curves and water pressure trend curves within the corresponding time periods, performs linear processing on the raw data of each interval, and applies a sliding window to the sample interval, setting the sliding window size to 10 time points and the overlap ratio to 50%. For each sliding window segment, the least squares method is applied for linear fitting to obtain the corresponding fitted slope value. Within the high-load segment numbered G04, the slope value of the displacement curve obtained within the four sliding windows is... The slope of the water pressure curve fitting is By performing a difference calculation on adjacent slope values, the slope difference of the displacement curves is obtained as follows: The slope difference of the water pressure curve is Set the threshold for transition detection Based on the set rules, determine whether there is a sudden change: for example, if the difference between the second group and the third group is 0.09 > 0.07, it meets the jump condition. Mark the time interval where the window is located as the "slope jump segment". Repeat the processing of all high-load segment data to generate several slope jump segments. Record the start time, end time, structure number and source curve type (displacement / water pressure) of each segment and store them in the "slope jump segment set" to prepare for the next step of coincidence analysis with the operation time point.

[0122] The time overlap submodule calls the slope jump segment set and extracts all time values ​​within each segment. It calls the recorded actuation trigger time point sequence, compares the two sets of time sequences, counts the number of time overlaps between each slope jump segment and the actuation trigger point, calculates the time overlap ratio of each segment, compares it with the overlap degree identification threshold, filters the segment numbers and time points that meet the conditions, and generates a mutation-related time group.

[0123] The time overlap submodule calls the slope jump segment set to extract the set of all time points for each jump segment. Let a jump segment be numbered G04, with a time period from 18:05:10 to 18:05:30 on 2024-07-28. This segment records 21 sampling points (sampling interval 1 second), then its corresponding time set is... Then, the actuation trigger time sequence is called. For example, if the actuation trigger time points recorded for G04 on this day are 18:05:05, 18:05:17, and 18:06:02, then the number of times the time overlaps between this transition segment and the trigger point sequence is calculated to be 1 (18:05:17 is within the segment), and the overlap ratio is calculated to be... If the time overlap recognition threshold is set to Then the segment satisfies The selected time segments are identified as valid mutation segments. This process is repeated to compare all mutation segments, ultimately filtering out all time segments that meet the overlap ratio requirement. The structure number and start and end times of each segment are output, generating mutation-related time groups for subsequent structural behavior anomaly assessment modules to analyze. This process emphasizes rigorous registration of time period granularity and sampling intervals. Through high-precision interval identification, it avoids misidentifying asynchronous events as abnormal responses, thereby improving the accuracy and robustness of association identification.

[0124] Please see Figure 2 and Figure 7 The maintenance trigger module includes a frequency extraction submodule, a level matching submodule, and a list generation submodule;

[0125] The frequency extraction submodule is based on the mutation association time group and the cumulative intensity list of operations. According to the operation frequency of each component number, it identifies the component number with the most operation frequency, extracts the number of operations of the number in the current period and the number of operations in the corresponding historical quarterly period, calculates the ratio of the two, and establishes a set of operation frequency growth ratio values.

[0126] The frequency extraction submodule, based on the mutation-related time group and the cumulative intensity list of actions, first establishes a mapping index between component numbers and their action trigger records. Frequency statistics are then performed on all component numbers. Within the monitoring period of the second quarter of 2025, components numbered G01, P02, F03, and G04 recorded 34, 51, 18, and 51 actions respectively. Since P02 and G04 have the highest frequency, they are selected as the highest frequency set. Next, historical data records are retrieved to locate the action counts of the same component number in the previous quarter, i.e., the first quarter of 2025. The read values ​​are: G01: 26 times, P02: 30 times, F03: 22 times, G04: 21 times. Based on this, the frequency ratio of the current period to the historical period is calculated. Specifically, G01 is... P02 is F03 is G04 is The ratio of the number of each component to its frequency of operation is summarized to generate a set of frequency of operation growth ratios. This set of ratios serves as a quantitative representation of the trend of component behavior changes and is used to identify abnormal frequency increases in the structural operating state. All calculations are completed under the premise of a uniform cycle duration to eliminate the influence of deviations caused by inconsistent sampling days.

[0127] The grade matching submodule identifies numbers whose ratios are higher than the facility maintenance critical frequency ratios based on the set of actuation frequency growth ratios, extracts the structural classification code and failure mode index number corresponding to the number, locates the maintenance grade field that matches the number in the maintenance response table, extracts the maintenance response grade and pairs it with the component number to obtain a maintenance grade matching pair.

[0128] After obtaining the set of actuation frequency growth ratios, the level matching submodule sets the facility maintenance critical frequency ratio as follows: Based on this threshold, the ratio of each component number is judged. If the frequency increase ratio corresponding to the component number is greater than 1.5, the high-frequency behavior response judgment process is initiated. From the calculation results of the previous module, the ratio of P02 is 1.70 and that of G04 is 2.43, both higher than the threshold. Therefore, they are identified as target numbers requiring maintenance level matching judgment. Subsequently, the structural classification codes and failure mode index numbers of P02 and G04 are extracted. The structural classification of P02 is set as "pump body subsystem" and the failure mode index number is set as "FM-006". G04 is set as "control gate system" and the index number is set as "FM-006". 02”, in the maintenance response table, the corresponding row item is retrieved according to the index number field. It is found that the maintenance level corresponding to FM-006 is “M2 (medium level)”, and the maintenance level corresponding to FM-002 is “M3 (high level)”. The maintenance level field is paired with the original part number respectively to finally obtain the maintenance level matching pair, in the form of {P02:M2, G04:M3}. This pair is used for the subsequent maintenance list sorting and scheduling system maintenance strategy attachment process. All fields used are derived from the structure classification mapping field and failure level classification field of the standard maintenance response table to avoid the use of fuzzy rule judgment.

[0129] The list generation submodule calls the maintenance level matching pair group and the cumulative intensity of operation list, arranges the part numbers with maintenance level records, establishes a correspondence table between part numbers and response levels, and summarizes the maintenance index list of high-frequency operation parts;

[0130] The inventory generation submodule calls the maintenance level matching pair group and the cumulative intensity list of operations. It filters the component numbers with clearly defined maintenance levels in the list, retaining only those appearing in the matching pair group. It then reads the corresponding maintenance level tags and performs a structured arrangement according to sorting rules such as ascending order of component number or descending order of operation integral value, generating the final set of correspondences between component numbers and maintenance response levels. For example, if the currently identifiable matching numbers are P02 and G04, the output maintenance inventory index is {G04:M3, P02:M2}. No uncertain fields are introduced during the generation process, ensuring that each result comes from the judgment criteria determined by the preceding module. This inventory is output as a maintenance index inventory for subsequent maintenance scheduling, job arrangement, and risk level layer matching processes, ultimately completing the output of the high-frequency operation component maintenance index inventory. This inventory is ultimately formed as a static set of two fields: number and level, and can be written to the structure database or maintenance control platform as needed.

[0131] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0132] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0133] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0134] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0135] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0136] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0137] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0138] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0139] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0140] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A water conservancy facility operation monitoring system based on sensor networks, characterized in that, The system includes: The status screening module acquires monitoring data from the sensor network of water conservancy facilities, filters data locations where abrupt changes occur, determines the correlation coefficient between flow data and water level data based on the data locations, marks and removes noisy data points, and obtains a group of structural status segments after disturbance removal. Based on the structural state fragment group after filtering out disturbances, the abnormal joint review module extracts the hydraulic component operation data, calculates the response value difference and the consistency coefficient of direction change. If the response value exceeds the tolerance threshold and the direction is consistent, it is marked as linkage behavior and generates a multi-component linkage time sequence record. Based on the multi-component linkage timing record, the aging response module retrieves the number of actuations, actuation duration, and load change rate of two types of structures, namely gate hinges and pump station shaft seals, during the linkage period. Combining the material type and service life of the equipment, it calculates the response integral intensity value and obtains the list of cumulative actuation intensity. Based on the cumulative intensity list of actions, the dynamic discrimination module extracts time periods where the load change rate is higher than the average of the cycle, extracts the recorded action trigger time points within the time periods, marks the time periods with an overlap greater than the overlap recognition threshold as response linkage segments, and generates a mutation association time group.

2. The water conservancy facility operation monitoring system based on sensor networks according to claim 1, characterized in that, The structural state segment group after perturbation removal includes trend stable segment number, perturbation exclusion mark set and state segment time index. The multi-component linkage time sequence record specifically includes linkage judgment time node sequence, component pair number matching set and linkage event flag column. The actuation cumulative intensity list includes actuation load integral value set, actuation weight distribution table and periodic performance response ratio. The mutation association time group specifically refers to mutation time label, component number index and response synchronization mark.

3. The water conservancy facility operation monitoring system based on sensor networks according to claim 2, characterized in that, The status screening module includes: The data derivative submodule acquires monitoring data from the water conservancy facility sensor network. The monitoring data includes structural displacement data sequences, water pressure data sequences, and flow velocity data sequences when the facility is in operation. Continuous data segments of the three types of data are extracted according to a unified time step. The first derivative difference is calculated by dividing the difference between adjacent data of each type of data at each time point by the time step, and a derivative change sequence is constructed. The derivative differences of the three types of data at each time point are used as combined data units to obtain a multi-source derivative change group. The rate screening submodule, based on the multi-source derivative change group, compares the adjacent differences of each type of derivative change result in the time series, calculates the difference of derivative difference per unit time to form a derivative change rate sequence, selects the rate values ​​of the center point and adjacent points in each continuous time window to perform change amplitude ratio calculation, extracts the data positions where the change amplitude exceeds the rate mutation threshold, and marks the derivative mutation corresponding to the position as an anomaly point to obtain the derivative mutation position group; The collaborative noise reduction submodule calls the derived value mutation location group, locates the corresponding time point values ​​in the flow data sequence and water level data sequence according to the mutation time point, extracts the flow value and water level value of the three time points before and after the mutation point to form a synchronization sequence, calculates the correlation coefficient of the two sets of sequences, selects the time points with correlation coefficients lower than the collaborative response benchmark value as asynchronous points, marks the data as noise points and removes them, and obtains the structural state fragment group after the disturbance is removed.

4. The water conservancy facility operation monitoring system based on sensor networks according to claim 3, characterized in that, The anomaly joint review module includes: The data extraction submodule obtains the structural state fragment group after the disturbance is removed, extracts the structural displacement change amplitude of the gate and the diversion structure and the orifice water pressure fluctuation value of the corresponding time period, and simultaneously obtains the instantaneous flow rate increase value of the pump body position in the same time period. The three sets of structural response value sequences are combined according to a unified time step to generate a linkage response comparison set. The value difference consistency submodule extracts the structural displacement change amplitude, water pressure fluctuation value and flow rate increase value for each data unit according to the linkage response comparison set. It calculates the response value difference of the three values ​​within the same time period, judges the change direction between values ​​in adjacent time periods, establishes a direction consistency identifier, calculates and obtains the composite linkage response quantity, compares the composite linkage response quantity with the response tolerance threshold, filters the time point combination that meets the direction consistency and the response quantity exceeds the threshold, and obtains the linkage trigger time series. The behavior tagging submodule calls the linkage trigger time sequence, adds behavior tag labels to the linkage trigger time points in chronological order, counts the structural response value combinations corresponding to each tag event, summarizes them into a unified index sequence, and records the corresponding time points in the sequence to establish a multi-component linkage time sequence record.

5. The water conservancy facility operation monitoring system based on sensor networks according to claim 4, characterized in that, The specific formula for obtaining the composite linkage response quantity is as follows: ； in, Indicates the composite linkage response quantity. This represents the normalized value indicating the magnitude of structural displacement variation between adjacent sampling points for the gate and the diversion structure. This represents the normalized value of the orifice water pressure fluctuation between adjacent sampling points. This represents the normalized value of the instantaneous flow rate increase of the pump body during that time period. This represents the normalized value of the sampling duration for that segment within the sampling period.

6. The water conservancy facility operation monitoring system based on sensor networks according to claim 5, characterized in that, The aging response module includes: The actuation extraction submodule obtains the linkage time period in the linkage time sequence record of the multi-component linkage, retrieves the actuation record data of the gate hinge and the pump station shaft seal, extracts three types of index data of each type of structure within the time period, namely the number of actuations, the duration of actuation and the load change rate, and integrates each data item in segments according to the time sequence to generate an actuation feature index set. The response integration submodule, based on the set of actuation characteristic indicators, calls the actuation number, actuation duration, and load change rate corresponding to each actuation segment in each type of structure, and calculates the response integration intensity value by combining the material type and service life of the structure. The integration results under each structure number are accumulated in chronological order to obtain the sequence of actuation response integration quantities. The numbering and aggregation submodule calls the actuation response integral sequence, sorts it in descending order according to the actuation response integral, and summarizes the integral intensity sequence data corresponding to the structure based on the number field to which the structure belongs, generating an actuation cumulative intensity list.

7. The water conservancy facility operation monitoring system based on sensor networks according to claim 6, characterized in that, The specific formula for obtaining the integral intensity value of the response is as follows: ； in, This represents the integral strength value of the response. The normalized value representing the number of actions. The normalized value representing the rate of change of load. The normalized value representing the duration of the action. The normalized value representing the number of years of use. Indicates the material type grade factor. This indicates the intensity of the influence factor in high humidity and high temperature environments.

8. The water conservancy facility operation monitoring system based on sensor networks according to claim 7, characterized in that, The dynamic discrimination module includes: The high-load extraction submodule obtains the list of cumulative intensity of actuation, extracts the normalized value of the load change rate corresponding to each actuation record, calculates the average value of the load change rate, identifies the time period above the average value, classifies and summarizes the high-load time period by number, and generates a high-load actuation time group. The slope recognition submodule extracts the structural displacement trend curve and water pressure trend curve within the corresponding time period based on the high-load operation time group. Taking the continuous data interval in each curve segment as input, it performs linear segment fitting on the curve data, extracts the curve slope value of the fitted segment, filters the time interval where the difference between adjacent slopes is greater than the jump discrimination threshold, and establishes a slope jump segment set. The time overlap submodule calls the slope jump segment set and extracts all time values ​​within each segment. It calls the recorded actuation trigger time point sequence, compares the two sets of time sequences, counts the number of time overlaps between each slope jump segment and the actuation trigger point, calculates the time overlap ratio of each segment, compares it with the overlap degree identification threshold, filters the segment numbers and time points that meet the conditions, and generates a mutation-related time group.

9. The water conservancy facility operation monitoring system based on sensor networks according to claim 8, characterized in that, The system also includes: Based on the mutation-related time group and the cumulative intensity list of operations, the maintenance triggering module numbers the components with the highest frequency of operations, calculates the frequency growth ratio, and if the growth ratio is higher than the facility maintenance critical frequency ratio, it extracts the component's category and the corresponding water conservancy facility failure mode index number, retrieves the corresponding item in the maintenance response table, and generates a high-frequency operation component maintenance index list. The maintenance index list for high-frequency actuating components includes maintenance level numbers, fault mode label sets, and component number lists.

10. The water conservancy facility operation monitoring system based on sensor networks according to claim 9, characterized in that, The maintenance triggering module includes: The frequency extraction submodule, based on the mutation association time group and the cumulative intensity list of actions, identifies the component number with the most frequent actions according to the action frequency of each component number, extracts the number of actions of the number in the current period and the number of actions in the corresponding historical quarterly period, calculates the ratio of the two, and establishes a set of action frequency growth ratio values. The grade matching submodule identifies the number whose ratio is higher than the facility maintenance critical frequency ratio based on the set of actuation frequency growth ratios, extracts the structural classification code and failure mode index number corresponding to the number, locates the maintenance grade field that matches the number in the maintenance response table, extracts the maintenance response grade and pairs it with the component number to obtain a maintenance grade matching pair. The list generation submodule calls the maintenance level matching pair group and the cumulative intensity of operation list, arranges the component numbers with maintenance level records, establishes a correspondence table between component numbers and response levels, and summarizes the maintenance index list of high-frequency operation components.

Images