Valve opening and closing state real-time tracking and abnormity diagnosis method

Abstract

The invention provides a valve opening and closing state real-time tracking and abnormality diagnosis method, which comprises the following steps: comparing the impedance growth rate of each branch with increased roughness with a preset flow distribution rated value to obtain the flow contribution variation of each branch, and determining the interference level of each branch according to the flow contribution variation; the abnormal type of the abnormal valve and the flow of the branch to which the abnormal valve belongs are analyzed, the influence weight between the abnormal type and the flow distribution amount is evaluated by combining the interference level, and the accurate coordinate of the abnormal position is determined; the actual flow at the accurate coordinates of the abnormal position is obtained, the floating range between the expected flow and the actual flow is determined, and the tracking judgment boundary value of the opening and closing state of the valve is determined according to the floating range; the roughness change rate and the opening reading which are collected in real time are processed by tracking and judging boundary values of the opening and closing state of the valve, and whether the opening and closing state of the valve is normal or not is evaluated.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and in particular to a valve opening and closing state real-time tracking and abnormal diagnosis method. BACKGROUND

[0002] In the field of industrial pipe network operation management, real-time tracking and abnormal diagnosis of valve opening and closing state is crucial, as it is directly related to the safety and efficiency of system operation. Whether it is water treatment, heating or chemical production, as the core component of flow control, the accurate judgment of the state of the valve plays a decisive role in ensuring the stable operation of the entire pipe network. Any misjudgment or abnormality of the valve state can lead to uncontrolled flow, even causing system failure and affecting production safety. However, current valve state monitoring relies too much on single valve opening data to infer its flow contribution, ignoring changes in external conditions that can lead to inaccurate judgments. In particular, when the resistance characteristics of some pipes in the pipe network change, even if the opening of a valve remains unchanged, the actual flow passing through it can fluctuate significantly, making the traditional judgment method unreliable. The deeper technical difficulty lies in the dynamic changes of pipe network impedance distribution, a core factor. Impedance distribution refers to the resistance characteristics of each pipe in the pipe network, which is affected by factors such as pipe aging and sediment accumulation and changes constantly. When the impedance distribution changes, the logic of estimating flow based on fixed opening no longer applies, as the same opening will affect the actual flow controlled by the valve due to the redistribution of the entire pipe network resistance. This change is not only difficult to predict, but also leads to deviations in valve state judgment. For example, in a multi-branch parallel heating pipe network, a branch pipe becomes rough due to long-term use, increasing resistance, causing the flow of other branches to increase significantly even if the valve opening does not change, completely invalidating the original flow estimation benchmark. Therefore, how to accurately determine the real flow contribution and opening and closing state of each valve in a complex environment where the impedance distribution of the pipe network is constantly changing has become a key problem that needs to be solved in this research. SUMMARY

[0003] The present application provides a valve opening and closing state real-time tracking and abnormal diagnosis method, comprising: Obtaining real-time opening values of each valve, branch flow distribution and pipe roughness change information, performing timestamp alignment and correlation processing to obtain a pipe network impedance distribution matrix, a valve opening and flow mapping function; Identifying the impedance growth rate of each branch according to the pipe network impedance distribution matrix, calculating the correlation coefficient between the impedance growth rate and the historical data sequence of roughness to identify the roughness increasing branch; Calculating the flow contribution variation by the impedance growth rate of the roughness increasing branch and the preset flow distribution rated value, and determining the interference level of each branch according to the flow contribution variation; An expected flow value under a target valve opening is calculated based on the valve opening-flow mapping function, and a flow difference value is obtained by comparing the actual flow data, to identify a flow abnormal valve; An abnormal valve abnormal type identifier, branch flow data, and interference level are obtained, an influence weight matrix is constructed, and an abnormal position accurate coordinate is determined in combination with a valve associated node set; Actual flow data and an expected flow value at the abnormal position accurate coordinate are obtained, a deviation rate is calculated, and a valve opening and closing state tracking judgment boundary value is determined; The valve opening and closing state tracking judgment boundary value is used to correct and compare real-time roughness change information and opening reading, and the valve opening and closing state is evaluated.

[0004] Further, the real-time opening value of each valve, the branch flow distribution amount, and the pipeline roughness change information are obtained, time stamp alignment and association processing are performed, a pipeline impedance distribution matrix and a valve opening-flow mapping function are obtained, including: The valve opening value, branch flow monitoring data, pipeline roughness coefficient change amount, and node pressure value are obtained by sensors, and the data are aligned according to the time stamp to obtain a synchronous pipeline operation state data set; The pressure difference value of adjacent nodes is calculated according to the synchronous pipeline operation state data set, and the Darcy-Weisbach formula is used to calculate the impedance coefficient of each pipe section to obtain the pipeline impedance distribution matrix by using the pressure difference value, flow, and roughness coefficient change amount; The valve opening-flow mapping function is established by using the least square method to fit the pipeline impedance distribution matrix and the real-time opening value of the valve.

[0005] Further, the impedance growth rate of each branch is identified according to the pipeline impedance distribution matrix, and the correlation coefficient is calculated between the impedance growth rate and the roughness historical data sequence to identify a roughness increasing branch, including: The impedance value sequence of each branch at consecutive time points in the pipeline impedance distribution matrix is obtained, and linear regression fitting is performed with time as the independent variable to obtain the impedance growth rate; The roughness historical data sequence of the corresponding branch is extracted, and the roughness change rate of adjacent time points is calculated; The Pearson correlation coefficient is calculated by the roughness change rate and the impedance growth rate to determine the roughness increasing branch.

[0006] Further, the flow contribution variation amount is calculated by the impedance growth rate of the roughness increasing branch and the preset flow distribution rated value, and the interference level of each branch is determined according to the flow contribution variation amount, including: The flow deviation coefficient is calculated by the impedance growth rate of the roughness increasing branch and the preset flow distribution rated value, and the flow contribution variation amount is determined according to the flow deviation coefficient and the current actual flow; The flow contribution variable change amount and the total flow of the pipe network are used to calculate an influence weight value, and the low, medium, and high interference levels are determined according to the influence weight value.

[0007] Further, the roughness increase branch current impedance growth rate and the preset flow distribution rated value are used to calculate a flow attenuation ratio, the rough layer thickness data and the pressure loss degree are collected, and the delivery difference of the roughness increase branch is determined, and specifically includes: The roughness increase branch current impedance growth rate and the preset flow distribution rated value are used to calculate a flow attenuation ratio, and the rough layer thickness data and the pressure loss degree are collected; According to the flow attenuation ratio, the rough layer thickness data, and the pressure loss degree, an influence coefficient is calculated using the Hazen-Williams formula, the influence coefficient is multiplied by the original flow capacity of the pipeline to obtain the actual flow capacity; The actual flow capacity and the rated demand flow of the downstream water supply area are compared to obtain a downstream flow shortage value, and the delivery difference of the roughness increase branch is determined according to the downstream flow shortage value and the flow attenuation ratio.

[0008] Further, the target valve opening degree and the expected flow value are calculated based on the mapping function of the valve opening degree and the flow, and the flow difference value is obtained by comparing the actual flow data, and the abnormal valve is identified, including: The expected flow value is obtained by linear interpolation of the target valve current opening value according to the mapping function of the valve opening degree and the flow, and the flow difference value is obtained by comparing the actual flow data read by the flow sensor; According to the positive and negative of the flow difference value, the valve blockage type abnormal valve or the valve leakage type abnormal valve is identified.

[0009] Further, the abnormal valve abnormal type identifier, the branch flow data, and the interference level are obtained, an influence weight matrix is constructed, and the abnormal position accurate coordinates are determined in combination with the valve associated node set, including: The abnormal valve abnormal type identifier and the branch real-time flow data are obtained, the upstream and downstream node numbers and their spatial coordinates are extracted according to the pipe network topology structure, and a valve associated node set is constructed; A judgment matrix is constructed by the abnormal type identifier, the branch flow data, and the interference level, the influence weight matrix is calculated by calculating the weight of each factor; According to the influence weight matrix, the influence degree of the abnormality to the adjacent nodes is analyzed to decrease, and the node with a weight greater than a preset threshold is determined to obtain an abnormal influence distribution vector; The spatial coordinates of the three nodes with the largest weight in the abnormal influence distribution vector are calculated by a weighted average method to calculate the abnormal center position, and the abnormal position accurate coordinates are determined.

[0010] Further, the method comprises the following steps: Obtaining actual flow data and expected flow value corresponding to the abnormal position accurate coordinate, and calculating a deviation rate. Based on the statistical distribution of historical operation data, the deviation rate interval corresponding to the normal floating range is determined, and the valve opening and closing state tracking judgment boundary value is obtained.

[0011] Further, the method comprises the following steps: Obtaining real-time roughness change information to calculate a roughness influence factor, and multiplying the opening degree reading to obtain a corrected opening degree value. Comparing the corrected opening degree value with the valve opening and closing state tracking judgment boundary value to determine whether the valve opening and closing state is normal or abnormal.

[0012] The technical scheme provided by the embodiment of the application can have the following beneficial effects: The application discloses a valve opening and closing state real-time tracking and abnormal diagnosis method. BRIEF DESCRIPTION OF DRAWINGS

[0013] Fig. 1 The flowchart of the valve opening and closing state real-time tracking and abnormal diagnosis method.

[0014] Fig. 2 The schematic diagram of the valve opening and closing state real-time tracking and abnormal diagnosis method.

[0015] Fig. 3 The schematic diagram of the valve opening and closing state real-time tracking and abnormal diagnosis method. DETAILED DESCRIPTION

[0016] In order for those skilled in the art to better understand the technical solutions in the specification, the technical solutions in the specification will be clearly and completely described below in combination with the drawings in the specification. Obviously, the described embodiments are only part of the embodiments of the specification, not all. Based on the embodiments in the specification, all other embodiments obtained by those of ordinary skill in the art without creative labor should belong to the protection scope of the specification.

[0017] As Figs. 1-3 The valve opening and closing state real-time tracking and abnormal diagnosis method can specifically include: S101, acquiring real-time opening degree and branch flow distribution of each valve, acquiring roughness change rate of pipeline at the same time, integrating and correlating, obtaining initial pipeline impedance distribution and mapping relationship between valve opening degree and flow.

[0018] The real-time opening degree value of each valve and the corresponding branch flow monitoring data are obtained by the sensor, and the roughness coefficient change amount of the inner wall of the pipeline and the node pressure value are collected at the same time. According to the time stamp of the opening degree value and the flow data, the data is aligned to obtain the synchronized pipeline operation state data set. According to the branch flow and node pressure value in the pipeline operation state data set, the pressure difference value between adjacent nodes is calculated, the Darcy-Weisbach formula is used, the pressure difference value, flow and roughness coefficient change amount are used as input parameters to calculate the impedance coefficient of each pipe section, if the roughness coefficient exceeds the preset threshold, the impedance coefficient is corrected, and the pipeline impedance distribution matrix is obtained. Through the pipeline impedance distribution matrix and the real-time opening degree value of the valve, the least square method is used to fit the functional relationship between the valve opening degree and the control branch flow, the mapping function of the valve opening degree and the flow is established, and the flow contribution value of each valve at different opening degrees is determined.

[0019] Specifically, in an embodiment, the pipeline monitoring system collects valve rotation angle data in real time through the opening degree sensor deployed at each valve position, and converts the angle value to percentage opening degree value. At the same time, ultrasonic flow meters are installed in each branch pipeline, and instantaneous flow values are recorded every five seconds. The roughness coefficient of the inner wall of the pipeline is indirectly measured by differential pressure method, that is, pressure sensors are arranged at both ends of the pipe section, and the roughness degree is inversely calculated according to the relationship between pressure loss and flow rate. The time stamp alignment uses interpolation algorithm to unify the data of different sampling frequencies to the same time reference.

[0020] Specifically, the application process of Darcy-Weisbach formula in pipe network impedance calculation is as follows: firstly, the pressure values of nodes at both ends of the pipe section are obtained, and the pressure difference is calculated as the driving head; then the pressure difference is divided by the measured flow of the pipe section to obtain the initial impedance coefficient; the impedance coefficient reflects the resistance characteristics of the pipe to the fluid, and is closely related to the length, diameter and roughness of the inner wall of the pipe. When the inner wall of the pipe is deposited or corroded due to long-term use, the roughness coefficient will increase, and a correction factor needs to be introduced. The correction process is as follows: the ratio of the measured roughness coefficient to the standard roughness coefficient is calculated, and if the ratio exceeds the preset threshold value of 1.2, the initial impedance coefficient is multiplied by the square root of the ratio as the corrected impedance coefficient. By calculating all the pipe sections in the pipe network as described above, an impedance matrix reflecting the resistance distribution of the entire pipe network is formed.

[0021] In a possible implementation, the process of establishing the mapping function of valve opening and flow by using the least square method includes: collecting the opening values of the same valve at different times and the flow values of the corresponding branches to form a data point set; selecting a quadratic polynomial as the fitting function form, and determining the polynomial coefficients by minimizing the sum of squares of errors between the measured flow and the fitted flow; the mapping function can predict the corresponding flow contribution value according to any opening value.

[0022] It should be noted that the flow contribution value represents the influence degree of a single valve on the flow of its controlled branch, and the value is dynamically adjusted with the impedance changes of other parts of the pipe network, so as to accurately track the valve state.

[0023] S102, according to the initial pipe network impedance distribution, identify the impedance growth rate of each branch, and evaluate whether the impedance growth rate exceeds the set impedance growth limit value, if it exceeds, compare the impedance growth rate with the roughness historical data to identify the branch with increased roughness.

[0024] Obtain the impedance value sequence of each branch in the pipe network impedance distribution matrix at consecutive time points, and perform linear regression fitting with time as the independent variable and impedance value as the dependent variable, and calculate the slope of the fitted straight line as the impedance growth rate of each branch. According to the comparison between the impedance growth rate and the preset impedance growth limit value, if the limit value is exceeded, the roughness historical data sequence of the corresponding branch is extracted from the pipe network historical monitoring database, and the roughness change rate is calculated by dividing the difference between the roughness of adjacent time points by the time interval. The Pearson correlation coefficient is calculated by the roughness change rate and the impedance growth rate, and the correlation coefficient of the two is obtained. When the correlation coefficient is greater than the preset threshold value, it is determined that the branch has the problem of increased roughness.

[0025] Specifically, in one embodiment, the pipe network monitoring system records the impedance value of each branch once an hour, forming a time series dataset. For each branch, the impedance value sequence of the last seven days is extracted, and a time-impedance coordinate system is constructed, where time is in units of hours. Linear regression fitting is performed using the least squares method, and the slope of the fitted straight line is the impedance growth rate, which reflects the trend of changes in pipe resistance characteristics.

[0026] It should be noted that the preset impedance growth limit value is determined according to the design life and material characteristics of the pipe network. For cast iron pipes, the limit value is set to not more than two percent of the initial value per month; for cement pipes, the limit value is appropriately relaxed to three percent. When the impedance growth rate of a branch exceeds the limit value, the system automatically retrieves the roughness monitoring records of the branch from the historical database for the past three months.

[0027] Specifically, the calculation process of the Pearson correlation coefficient involves statistical correlation analysis of the roughness change rate sequence and the impedance growth rate sequence. First, the two sequences are standardized to eliminate the influence of dimensions; then the covariance of the two sequences is calculated, and divided by the product of the standard deviations to obtain the correlation coefficient. The coefficient takes a value ranging from -1 to 1, and when the coefficient is greater than 0.7, it indicates that the increase in roughness is the main cause of the impedance growth. In a heating pipe network, the deposition of scale on the inner wall of the pipe will cause both roughness and impedance to increase, and the two are strongly positively correlated; while in a chemical pipe network, the accumulation of corrosion products will also have a similar effect. Through correlation judgment, it can be distinguished whether the impedance growth is caused by roughness change or other factors such as pipe diameter deformation, valve failure.

[0028] Preferably, for the case where the correlation coefficient is between 0.5 and 0.7, the system will further analyze the time distribution characteristics of the roughness change, and if the change shows a sustained growth trend rather than random fluctuations, it will still determine that there is a roughness increase problem in the branch.

[0029] In one possible implementation, the preset threshold value is adjusted according to different pipe network types, with a threshold value of 0.75 for domestic water supply pipe networks and a threshold value of 0.65 for industrial circulating water pipe networks, to adapt to the monitoring needs of different application scenarios.

[0030] S103, by comparing the impedance growth rates of each branch with the preset roughness increase, the flow contribution variation of each branch is obtained, and the interference level of each branch is determined according to the flow contribution variation.

[0031] The impedance growth rate of each branch with increased roughness is obtained, and the ratio of this rate to the preset flow distribution rating of each branch is used to calculate the flow deviation coefficient. The product of this flow deviation coefficient and the current actual flow rate of the branch is used to determine the flow contribution variation of each branch. The ratio of this flow contribution variation to the total flow rate of the pipeline network is used to calculate the influence weight of each branch on the pipeline network operation, obtaining the influence weight value for each branch. This influence weight value is compared with a preset interference level threshold. If the influence weight is less than the lower threshold, it is determined to be a low interference level; if it is between the lower and upper thresholds, it is determined to be a medium interference level; and if it exceeds the upper threshold, it is determined to be a high interference level.

[0032] Specifically, in one implementation, the flow deviation coefficient characterizes the degree to which the actual operating state of a branch deviates from the design conditions by measuring the ratio of the impedance growth rate to a preset flow distribution rating. The preset flow distribution rating is determined during the pipeline network design phase based on the water demand of each area and the pipeline's carrying capacity, serving as a benchmark for ideal operating conditions. When impedance growth causes a change in flow distribution, the deviation coefficient reflects the magnitude of this change.

[0033] Specifically, the calculation of the flow contribution variation involves a comprehensive consideration of the deviation coefficient and the actual flow rate. The deviation coefficient reflects the degree of deviation of the impedance growth rate from the preset flow allocation rating. The flow contribution variation is calculated using the formula ΔQ = (deviation coefficient - 1) × Q. actual Calculate, where Q actual This represents the current actual flow rate of the branch. If the deviation coefficient of a branch is 1.2 and the actual flow rate is 50 liters per second, then the change in flow contribution is (1.2-1)×50=10 liters per second, indicating the change in flow rate from the rated value due to increased roughness. A deviation coefficient greater than 1 indicates that the flow distribution deviates from the design conditions due to increased impedance. The larger the absolute value of the change, the more significant the impact on the flow distribution of the pipeline network.

[0034] It should be noted that the influence weight is used to quantify the impact of each branch on the overall operation by measuring the proportion of the change in flow contribution to the total flow of the pipeline network. In heating pipeline networks, the influence weight of the main pipeline is usually larger, while the influence weight of the branch pipeline is relatively smaller.

[0035] Preferably, the interference level classification adopts a three-level threshold mechanism, with a lower threshold set at 0.05 and an upper threshold set at 0.15. When the impact weight is less than 0.05, it indicates that the roughness change of the branch has a minor impact on the pipeline network operation, and is classified as a low interference level, requiring only routine monitoring; between 0.05 and 0.15 is a medium interference level, requiring increased monitoring frequency and the development of a maintenance plan; and above 0.15 is a high interference level, indicating that the branch has seriously affected the normal operation of the pipeline network, requiring immediate cleaning or replacement measures. This classification mechanism enables maintenance personnel to rationally allocate maintenance resources according to different interference levels, prioritizing high-interference branches and achieving precise management.

[0036] In one possible implementation, for a parallel pipeline network structure, the interference levels of each branch will also affect each other. When a branch is classified as having a high interference level, the flow of the branch connected in parallel with it will increase accordingly, and its interference level assessment needs to be dynamically adjusted.

[0037] The current impedance growth rate of the branch with increased roughness is obtained and compared with the preset flow distribution rating of the branch with increased roughness to obtain the flow attenuation ratio. At the same time, the thickness of the rough layer on the inner wall of the branch with increased roughness and the pressure loss when the fluid passes through are collected. The actual flow capacity attenuation of the branch with increased roughness after impedance growth is analyzed, the degree of flow shortage in the downstream water supply area caused by the flow capacity attenuation is assessed, and the transport difference of the branch with increased roughness is determined.

[0038] The current impedance growth rate of the branch with increased roughness is obtained and compared with the preset flow distribution rating of that branch. The flow attenuation ratio is calculated by the ratio of the two. Simultaneously, data on the thickness of the rough layer on the inner wall of the branch pipe and the pressure loss during fluid flow are collected. Based on the flow attenuation ratio, rough layer thickness, and pressure loss, the Hayzen-Williams formula is used to calculate the influence coefficient of pipe roughness on flow. This influence coefficient is multiplied by the original flow capacity of the pipe to obtain the actual flow capacity after impedance increase. This actual flow capacity is compared with the rated demand flow determined during the pipeline network design of the downstream water supply area. If the actual flow capacity is lower than the rated demand flow, the difference is calculated to obtain the downstream flow shortage value. Based on the downstream flow shortage value and the flow attenuation ratio, the reduction in transport capacity caused by increased roughness in this branch is calculated and determined as the transport difference of the branch with increased roughness.

[0039] Specifically, in one implementation, the flow attenuation ratio is calculated based on the ratio of the impedance growth rate to a preset flow allocation rating. This preset flow allocation rating is determined during the initial design phase of the pipeline network based on the water demand, pipe specifications, and water pressure distribution of each area, representing the flow share the pipeline should ideally handle. When the inner wall of the pipe becomes rougher due to long-term operation, the impedance growth rate deviates from the normal range; the ratio of these two values ​​directly reflects the severity of the flow attenuation.

[0040] It should be noted that the thickness of the rough layer on the inner wall of the pipe is measured non-contactly using an ultrasonic thickness gauge. The ultrasonic waves emitted by the probe are reflected at the interface between the rough layer and the pipe wall substrate, and the thickness of the rough layer is calculated based on the echo time difference. The pressure loss is obtained by high-precision pressure sensors installed at both ends of the pipe section, which record the pressure change of the fluid before and after passing through the pipe section in real time.

[0041] Specifically, the application of the Hayzen-Williams formula in pipeline hydraulic calculations involves the comprehensive consideration of several key parameters. This formula expresses the functional relationship between flow rate and pipe roughness coefficient, pipe diameter, and hydraulic gradient, where the roughness coefficient is the core parameter reflecting the condition of the pipe's inner wall. For new cast iron pipes, the roughness coefficient is typically 130; after ten years of use, it drops to around 100; for severely corroded or scaled pipes, the roughness coefficient may drop below 60. The current roughness coefficient value can be estimated using the measured thickness of the roughened layer. Substituting the flow rate attenuation ratio, the measured roughness coefficient, and the degree of pressure loss into the Hayzen-Williams formula, the influence coefficient of pipe roughness on flow rate is calculated. This influence coefficient reflects the degree to which increased roughness weakens the pipe's water-carrying capacity, typically ranging from 0.6 to 1.0. Multiplying the influence coefficient by the original flow capacity of the pipe yields the actual flow capacity under the current impedance growth state. This actual flow capacity value directly determines the actual amount of water that the branch can supply downstream, serving as the basis for subsequent flow shortage assessments.

[0042] Preferably, the effect of temperature on water viscosity is also considered when calculating the actual flow capacity. For every 10-degree Celsius increase in water temperature, the viscosity decreases by about 20%, and the corresponding flow resistance also decreases. Therefore, in heating pipe networks, the actual flow capacity of high-temperature water will be slightly higher than that of room-temperature water.

[0043] In one possible implementation, the rated flow demand of the downstream water supply area is determined comprehensively based on the water-using population, industrial water consumption, and fire-fighting reserve requirements. Domestic water consumption is calculated at 200 liters per person per day, industrial water consumption is determined according to production process requirements, and fire-fighting reserves are set according to building area and fire risk level.

[0044] For example, in the circulating water network of a chemical industrial park, the rated demand flow of a main branch is 500 cubic meters per hour. Calculations show that the actual flow capacity of this branch is only 420 cubic meters per hour, resulting in a flow shortage of 80 cubic meters per hour. This means that the downstream cooling tower cannot receive enough circulating water, potentially leading to a decrease in heat exchange efficiency. Furthermore, the determination of this flow difference comprehensively considers both the flow shortage value and the flow attenuation ratio. The flow shortage value directly reveals the downstream water supply gap, while the flow attenuation ratio allows us to trace how much of this gap is directly caused by increased roughness.

[0045] For example, if the flow rate reduction ratio is 0.8, it means that the actual flow rate is only 80% of the rated value, and the conveying capacity has decreased by 20%. Multiplying this reduction ratio by the flow rate shortage value yields the conveying difference caused by increased roughness. This conveying difference is an important basis for developing pipeline cleaning or replacement plans.

[0046] Understandably, accurately determining the transport difference helps maintenance personnel assess the urgency and cost-effectiveness of pipeline maintenance. When the transport difference exceeds 30% of the pipeline's transport capacity, it is generally considered that immediate cleaning or replacement of the pipeline's inner wall is necessary.

[0047] S104. Based on the mapping relationship between valve opening and flow rate, analyze the expected flow rate under the target valve opening, analyze the difference between the actual flow rate and the expected flow rate under the target valve opening, obtain the flow rate drop value, evaluate whether the flow rate drop value exceeds the set limit, and identify abnormal valves with abnormal flow rates.

[0048] Based on the mapping relationship between valve opening and flow rate, the current opening value of the target valve is obtained. Linear interpolation is performed on adjacent data points in the mapping relationship to calculate the expected flow rate corresponding to this opening. Simultaneously, the actual flow rate data of the target valve control branch is read from the flow sensor. The difference between the expected flow rate and the actual flow rate data is calculated to obtain the flow rate difference value. The absolute value of the flow rate difference value is compared with a preset limit. If it exceeds the preset limit, the valve is determined to have an abnormal flow rate. For the flow rate difference value, its positive or negative sign determines the type of abnormality. A positive flow rate difference value indicates that the actual flow rate is less than the expected flow rate, identifying it as a valve blockage-type abnormality. A negative flow rate difference value indicates that the actual flow rate is greater than the expected flow rate, identifying it as a valve leakage-type abnormality.

[0049] Specifically, in one implementation, the mapping relationship between valve opening and flow rate is established using historical operating data, forming a discrete set of data points for valve opening and flow rate. When it is necessary to calculate the expected flow rate for a specific opening, a linear interpolation method is used to estimate between adjacent data points.

[0050] Specifically, the linear interpolation process involves finding adjacent data points of the target valve opening value in the mapping dataset. Assuming the target valve opening is 65%, and the mapping dataset stores discrete points corresponding to a flow rate of 40 liters per second for a 60% opening and 45 liters per second for a 70% opening, then the expected flow rate corresponding to a 65% opening is calculated to be 42.5 liters per second through linear interpolation. Actual flow data is collected in real-time by an electromagnetic flowmeter installed downstream of the valve, which updates its measurement value every second. The flow difference is obtained by subtracting the actual flow rate from the expected flow rate; this difference directly reflects the deviation between the valve's actual operating state and its ideal state. A positive flow difference indicates that the actual flow rate is less than the expected flow rate, suggesting potential blockage in the valve-controlled pipeline, such as scale buildup on the pipeline wall, valve core jamming, or filter clogging. A negative flow difference indicates that the actual flow rate is greater than the expected flow rate, suggesting potential problems such as poor valve sealing, valve body damage, or leakage in the bypass pipeline.

[0051] It should be noted that the preset limit is determined based on the normal operating fluctuation range of the pipeline network, and is usually set at 10% of the expected flow rate. Exceeding this limit indicates that the valve's operating status has deviated from the normal range.

[0052] Preferably, the preset limit can be adjusted for different types of valves. The limit for ball valves is set to 8%, the limit for butterfly valves is set to 12%, and the limit for gate valves is set to 15%.

[0053] In one possible implementation, after anomaly type identification, the system records the location number of the abnormal valve, the anomaly type, the flow drop value, and the occurrence time, forming an anomaly file for maintenance personnel to refer to.

[0054] For example, when the regulating valve of a water supply network is opened to 50%, the expected flow rate is 30 liters per second, but the actual measured flow rate is only 22 liters per second. The flow rate difference is 8 liters, which exceeds the preset limit of 3 liters. This is judged as a blockage-type abnormality, indicating that pipeline cleaning and maintenance are required.

[0055] S105. Analyze the abnormality type of the abnormal valve and the flow rate of its branch, evaluate the influence weight between the abnormality type and the flow distribution in combination with the interference level, and determine the precise coordinates of the abnormal location.

[0056] Obtain the anomaly type identifier and real-time flow data of the branch to which the abnormal valve belongs. Extract the upstream and downstream node numbers and their spatial coordinates of the valve based on the pipeline topology, and construct a set of valve-associated nodes. Using the anomaly type identifier and branch flow data, combined with a preset interference level, construct a judgment matrix using the severity of the anomaly type, the degree of flow deviation, and the interference level as judgment factors, and calculate the weight of each factor to obtain an influence weight matrix. Based on the influence weight matrix and the set of valve-associated nodes, analyze the decreasing influence of the anomaly from the valve to adjacent nodes, identify nodes with weights greater than a preset threshold within the influence range, and obtain the anomaly influence distribution vector. Using the spatial coordinates of the three nodes with the largest weights in the anomaly influence distribution vector, calculate the anomaly center location using a weighted average method to determine the precise coordinates of the anomaly location.

[0057] Specifically, in one implementation, the pipeline topology is stored in the form of an adjacency matrix, with each node recording its unique number, spatial coordinates, and connection relationships. When an abnormal valve is detected, the system extracts the valve's location information from the topology database, including its directly connected upstream water supply nodes and downstream water user nodes. S104 identifies abnormal valves through branch flow data and valve status monitoring, including blockage, leakage, and jamming anomalies. Anomaly type identification uses numerical coding, where 1 represents a blockage anomaly, 2 represents a leakage anomaly, and 3 represents a valve jamming anomaly. The branch flow data is collected in real time by an electromagnetic flowmeter installed on the pipeline, with a sampling frequency of once per second.

[0058] It should be noted that the valve-related node set includes not only directly connected nodes, but also second-level adjacent nodes, that is, other nodes connected to the directly connected nodes. The spatial coordinates of each node are represented in a geodetic coordinate system, including three dimensions: longitude, latitude, and elevation.

[0059] Specifically, the construction of the judgment matrix involves pairwise comparisons of three key judgment factors. The first factor is the severity of the anomaly type, determined by the anomaly type identifier: blockage anomalies are assigned a value of 9, leakage anomalies are assigned a value of 7, and stagnation anomalies are assigned a value of 5. The second factor is the degree of flow deviation, determined by the percentage deviation between the actual flow and the rated flow: deviation exceeding 30% is assigned a value of 9, deviation between 20% and 30% is assigned a value of 7, deviation between 10% and 20% is assigned a value of 5, and deviation less than 10% is assigned a value of 3. The third factor is the interference level, directly using the high, medium, and low interference levels determined in step S103, assigned values ​​of 9, 5, and 3 respectively. A 3×3 judgment matrix A is constructed, with matrix element a. ij This represents the ratio of the importance of the i-th factor to the j-th factor. The weights are calculated using the eigenvalue method of the Analytic Hierarchy Process (AHP): First, solve the equation (A-λ). max·I)·W=0, thus obtaining the largest eigenvalue λ max The corresponding eigenvector W; then the eigenvector is normalized so that Σw i =1, obtaining the weight values ​​w=[w1,w2,w3] for each factor. These three weight values ​​constitute the influence weight matrix, reflecting the comprehensive impact of anomaly type, flow deviation, and disturbance level on pipeline operation. Consistency testing is performed by calculating the consistency ratio CR=(λ). max The matrix is ​​calculated as (n) / (n-1) / RI, where n=3 is the matrix order and RI is the average random consistency index (RI=0.58 when n=3). When CR<0.1, the matrix is ​​considered to have satisfactory consistency. The weights of the three factors are used to form an influence weight matrix, which reflects the comprehensive impact of anomalies on pipeline operation. This multi-factor comprehensive evaluation method can comprehensively consider the different dimensions of anomalies, avoiding the one-sidedness of single-indicator judgment.

[0060] Preferably, when calculating the consistency of the judgment matrix, if the consistency ratio is greater than 0.1, the comparison values ​​in the judgment matrix need to be adjusted until the consistency requirements are met.

[0061] In one possible implementation, the distribution vector of anomaly impacts is obtained based on the law of diminishing impact. Starting from the node where the anomaly valve is located, the degree of impact decreases inversely proportional to the square of the distance. This is calculated using the following formula: Node Impact Weight W node =W base ×(k / d 2 ), where W base The basic influence weight is calculated using the judgment matrix, where k is a constant of 0.8, and d is the shortest path distance from the node to the anomalous valve (in terms of pipe segments). For directly connected nodes (d=1), the node influence weight is 0.8 times the basic weight; for second-order adjacent nodes (d=2), the node influence weight is 0.2 times the basic weight (0.8 / 4=0.2). Considering the actual complexity of the pipeline network, for nodes with d=2, the system adjusts the influence weight to 0.3 times the basic weight to improve the model's sensitivity to indirect influences. For example, a regulating valve in a chemical pipeline network experiences a blockage. The upstream node connected to this valve is numbered N101, and the downstream nodes are numbered N102 and N103. The basic influence weight W is calculated using the judgment matrix. base=0.75. Therefore, the influence weight of the directly connected node N101 is 0.75 × 0.8 = 0.6, and the influence weights of N102 and N103, as direct downstream nodes, are also 0.6. The influence weight of the second-level adjacent node N201 connected to N101 is 0.75 × 0.3 = 0.225. Furthermore, the abnormal influence distribution vector contains the numbers of all affected nodes and their corresponding influence weight values. By setting a weight threshold of 0.2, nodes with weights greater than this threshold are selected as key influencing nodes.

[0062] For example, the weighted average method for calculating the anomaly center location is as follows: Select the three nodes with the largest influence weights, multiply the spatial coordinates of each node by its normalized weight, and then sum the three weighted coordinates to obtain the precise coordinates of the anomaly location. These coordinates represent the geometric center of the anomaly's influence, facilitating rapid location and handling by maintenance personnel.

[0063] It is understandable that when an abnormal valve is located at the end of the pipeline network, the influence weight of only one or two nodes may exceed the threshold. In this case, the weighted average coordinates of these nodes are used as the abnormal location.

[0064] In one embodiment, a branch valve of a heating pipeline experienced a leakage anomaly. The coordinates of the anomaly location were calculated using the above method to be 116.3874 degrees east longitude and 39.9042 degrees north latitude. This location was exactly near the intersection of two pipelines, with a deviation of less than 5 meters from the actual leak point, thus achieving precise location of the anomaly.

[0065] S106. Obtain the actual flow rate at the precise coordinates of the abnormal location, determine the fluctuation range between the expected flow rate and the actual flow rate, and determine the tracking and judgment boundary value of the valve opening and closing status based on the fluctuation range.

[0066] The precise coordinates of the pipeline node corresponding to the abnormal location are obtained. The actual flow data of that node is read from the flow monitoring device, and the expected flow value for that location is extracted from the pipeline design parameters. The difference between the actual and expected flow is calculated, and the difference is divided by the expected flow to obtain the deviation rate. Based on the statistical distribution of historical operating data, the normal fluctuation range is determined as the flow value within a preset percentage range of the deviation rate. Using the normal fluctuation range in conjunction with the deviation rate, when the deviation rate exceeds the upper percentage limit, it is determined that the valve is in an abnormal closed state; when the deviation rate is below the lower percentage limit, it is determined that the valve is in an abnormal open state, thus obtaining the tracking and judgment boundary values ​​for the valve opening and closing states.

[0067] Specifically, in one implementation, the precise coordinates of the abnormal location have been determined using the aforementioned weighted average method, and the corresponding pipeline node is equipped with an electromagnetic flowmeter and a pressure sensor. The flow monitoring equipment collects the instantaneous flow rate value once per second, and the average of ten consecutive sampled values ​​is taken as the actual flow rate data for that node. The expected flow rate value is predetermined based on the pipeline hydraulic calculation model, taking into account factors such as pipe diameter, design pressure, and water demand.

[0068] It should be noted that the deviation rate reflects the degree to which the actual operating condition deviates from the design condition. The formula for calculating the deviation rate is R = (Qactual - Qexpected) / Qexpected, where R is the deviation rate, Qactual is the actual flow rate, and Qexpected is the expected flow rate. A positive deviation rate indicates that the actual flow rate is greater than the expected flow rate; a negative deviation rate indicates that the actual flow rate is less than the expected flow rate.

[0069] Specifically, the determination of the normal fluctuation range is based on a statistical analysis of historical operating data. Flow data from the past three months of normal operation at the node is collected, and after removing outliers, the distribution characteristics of the flow deviation rate are calculated. A normal distribution is used to fit the deviation rate data, and the mean and standard deviation are calculated. The mean plus or minus twice the standard deviation is used as the upper and lower limits of the normal fluctuation range, which covers 95% of normal operating conditions. In water supply networks, due to daily and seasonal variations in water demand, the normal deviation rate typically fluctuates between -15% and +15%. When the network operation is stable, the deviation rate concentrates in the range of -5% to +5%. The normal fluctuation range determined through statistical analysis provides a scientific basis for subsequent anomaly judgments, avoiding misjudgments caused by normal fluctuations.

[0070] Preferably, a time-segmented statistical method is used for flow data in different time periods. The day is divided into peak water usage periods, off-peak periods, and low-water periods, and the normal fluctuation range of each period is statistically analyzed.

[0071] In one possible implementation, the setting of the tracking and judgment boundary values ​​takes into account the mechanical characteristics of the valve. When the deviation rate exceeds 20% of the upper limit of the normal floating range, it is determined that the valve is abnormally open; when the deviation rate is lower than 20% of the lower limit of the normal floating range, it is determined that the valve is abnormally closed.

[0072] For example, the normal deviation rate range for a certain chemical pipeline node is from -10% to +10%. When the measured deviation rate reaches +12%, exceeding the upper limit, the system determines that the valve at that location may be blocked or not properly closed. Furthermore, the tracking and judgment boundary value is dynamically adjusted according to the pipeline's operating conditions, appropriately widening the boundary range during maintenance and tightening it during critical production periods to achieve precise monitoring of valve status.

[0073] S107. By tracking and judging the boundary value of the valve opening and closing status, the roughness change rate and opening reading collected in real time are processed to evaluate whether the valve opening and closing status is normal.

[0074] By tracking and judging the boundary values ​​of the valve's opening and closing status, the real-time roughness change rate and valve opening reading are obtained. The ratio of the roughness change rate to a preset change threshold is used as the roughness influence factor. The roughness influence factor is multiplied by the opening reading to obtain a corrected opening value. This corrected opening value is compared with the tracking and judging boundary values ​​to determine if it is within the normal range. If the corrected opening value is within the boundary value range, the valve's opening and closing status is assessed as normal; if it exceeds the boundary value range, the valve's opening and closing status is assessed as abnormal.

[0075] Specifically, in one implementation, boundary values ​​are tracked and used as a benchmark for valve condition assessment. These values ​​are pre-determined based on historical valve failure data and engineering experiments and stored in the monitoring system. Roughness variation information is obtained through periodic inspection of the pipe inner wall, using an ultrasonic thickness gauge to measure the roughness layer thickness weekly and recording the thickness value for each measurement. Valve opening readings are acquired in real-time using a rotary encoder mounted on the valve stem, achieving an accuracy of 0.1%.

[0076] Specifically, the calculation process of the roughness influence factor involves a comparative analysis of the roughness change rate and a preset change threshold. First, the difference in roughness layer thickness between two adjacent measurements is calculated and divided by the measurement time interval (7 days) to obtain the weekly roughness change rate (unit: mm / week). The weekly change rate is multiplied by 4 to convert it to a monthly roughness change rate (unit: mm / month). The preset change threshold is determined based on the pipe material and service life: the threshold for new pipes is set at no more than 0.1 mm per month, and for pipes used for more than five years, the threshold is appropriately relaxed to 0.2 mm. The measured monthly roughness change rate is divided by the preset change threshold to obtain the dimensionless roughness influence factor. The calculation formula is: Roughness influence factor = Monthly roughness change rate / Preset change threshold. When the influence factor is greater than 1, it indicates that the roughness growth rate exceeds the normal level; when it is less than 1, it indicates that the roughness growth is within an acceptable range. This influence factor directly reflects the degree of impact of roughness changes on the valve's flow control capability. In heating pipe networks, due to the presence of minerals in the water, scale easily forms on the inner walls of the pipes, and the roughness influence factor typically varies between 0.8 and 1.5.

[0077] It should be noted that the opening value is corrected using a multiplicative correction method. The original opening reading is multiplied by the roughness influence factor to obtain the corrected opening value that takes into account the roughness effect. This correction method is based on fluid mechanics principles, where increased roughness reduces the effective flow area of ​​the valve.

[0078] Preferably, when the roughness influence factor exceeds 1.3, the system will issue a warning signal, indicating that pipeline cleaning and maintenance are required.

[0079] In one possible implementation, the comparison between the corrected opening value and the tracking judgment boundary value uses an interval judgment method. The boundary value includes two thresholds: an upper limit and a lower limit. If the corrected opening value falls within this interval, it is determined to be in a normal state.

[0080] For example, if the initial opening reading of a valve is 60% and the roughness influence factor is 1.2, then the corrected opening value is 72%. If the tracking and judgment boundary value range is between 50% and 80%, then the valve's condition is assessed as normal. Furthermore, abnormal conditions can be further subdivided into minor and severe abnormalities, each corresponding to different maintenance strategies, achieving refined management of the valve's operating status.

[0081] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A valve opening and closing state real-time tracking and abnormality diagnosis method, characterized by, The method comprises the following steps: Obtain real-time valve opening value, branch flow distribution amount and pipeline roughness change information, perform timestamp alignment and correlation processing, obtain pipeline network impedance distribution matrix, valve opening and flow mapping function; Identify the impedance growth rate of each branch according to the pipeline network impedance distribution matrix, calculate the correlation coefficient between the impedance growth rate and the roughness historical data sequence, and identify the roughness increasing branch; Calculate the flow contribution variation by the impedance growth rate of the roughness increasing branch and the preset flow distribution rated value, and determine the interference level of each branch according to the flow contribution variation; Calculate the expected flow value under the target valve opening based on the valve opening and flow mapping function, compare the expected flow value with the actual flow data to obtain the flow difference value, and identify the flow abnormal valve; Obtain the abnormal type identification of the abnormal valve, the branch flow data and the interference level, construct an influence weight matrix, and determine the accurate coordinates of the abnormal position combined with the valve correlation node set; Obtain the actual flow data and the expected flow value at the accurate coordinates of the abnormal position, calculate the deviation rate and determine the valve opening and closing state tracking judgment boundary value; Compare and correct the real-time roughness change information and opening reading by the valve opening and closing state tracking judgment boundary value, and evaluate the valve opening and closing state.

2. The method of claim 1, wherein The method comprises the following steps: Obtain valve opening value, branch flow monitoring data, pipeline roughness coefficient change and node pressure value through sensors, align the data according to timestamp to obtain a synchronous pipeline network operation state data set; Calculate the pressure difference value of adjacent nodes according to the synchronous pipeline network operation state data set, calculate the impedance coefficient of each pipe section by the pressure difference value, flow and roughness coefficient change to obtain the pipeline network impedance distribution matrix by using Darcy-Weisbach formula; Establish the mapping function of valve opening and flow by using least square method fitting based on the pipeline network impedance distribution matrix and real-time valve opening value.

3. The method of claim 1, wherein the method further comprises: The method comprises the following steps: Obtain the impedance value sequence of each branch at consecutive time points in the pipeline network impedance distribution matrix, perform linear regression fitting with time as the independent variable to obtain the impedance growth rate; Extract the roughness historical data sequence of the corresponding branch, calculate the roughness change rate of adjacent time points; Calculate the Pearson correlation coefficient by the roughness change rate and the impedance growth rate to determine the roughness increasing branch.

4. The method of claim 1, wherein the method further comprises: The method comprises the following steps: Calculate the flow deviation coefficient by the impedance growth rate of the roughness increasing branch and the preset flow distribution rated value, determine the flow contribution variation according to the flow deviation coefficient and the current actual flow; Calculate the influence weight value by the flow contribution variation and the total flow of the pipeline network, and determine the low, medium and high interference levels according to the influence weight value.

5. The method of claim 4, wherein the method further comprises: Also include the acquisition of roughness increase branch current impedance growth rate and preset flow distribution rating value calculation flow attenuation ratio, collecting the rough layer thickness data and pressure loss degree, determine the roughness increase branch delivery difference, specifically including: Acquire roughness increase branch current impedance growth rate and preset flow distribution rating value calculation flow attenuation ratio, collecting the rough layer thickness data and pressure loss degree; According to the flow attenuation ratio, rough layer thickness data and pressure loss degree, the influence coefficient is calculated by using the Hazen-Williams formula, and the actual flow capacity is obtained by multiplying the influence coefficient and the original flow capacity of the pipeline; By comparing the actual flow capacity with the rated demand flow of the downstream water supply area, the downstream flow shortage value is obtained, and the delivery difference of the roughness increase branch is determined according to the downstream flow shortage value and the flow attenuation ratio.

6. The method of claim 1, wherein the method further comprises: The target valve opening degree is calculated based on the mapping function of the valve opening degree and the flow, and the expected flow value is compared with the actual flow data to obtain the flow difference value, and the abnormal valve is identified, including: According to the mapping function of the valve opening degree and the flow, the expected flow value is obtained by linear interpolation of the current opening value of the target valve, and the flow difference value is obtained by comparing the actual flow data read by the flow sensor; According to the positive and negative of the flow difference value, the valve blockage type abnormal valve or the valve leakage type abnormal valve is identified.

7. The method of claim 1, wherein the method further comprises: The abnormal type identification of the abnormal valve, the branch flow data and the interference level are obtained, the influence weight matrix is constructed, and the accurate coordinates of the abnormal position are determined by combining the valve associated node set, including: The abnormal type identification of the abnormal valve, the branch flow data and the interference level are obtained, the influence weight matrix is constructed, and the accurate coordinates of the abnormal position are determined by combining the valve associated node set, including: The judgment matrix is constructed by the abnormal type identification, the branch flow data and the interference level, the weight of each factor is calculated to obtain the influence weight matrix; According to the influence weight matrix, the influence degree of abnormality to adjacent nodes is analyzed, and the nodes with weight greater than the preset threshold are determined to obtain the abnormal influence distribution vector; The spatial coordinates of the three nodes with the largest weight in the abnormal influence distribution vector are calculated by weighted average method to obtain the accurate coordinates of the abnormal position.

8. The method of claim 1, wherein the method further comprises: The actual flow data and the expected flow value at the accurate coordinates of the abnormal position are obtained, the deviation rate is calculated, and the valve opening and closing state tracking judgment boundary value is determined, including: The actual flow data and the expected flow value of the pipeline network node corresponding to the accurate coordinates of the abnormal position are obtained, and the deviation rate is calculated; Based on the statistical distribution of historical operation data, the deviation rate interval corresponding to the normal floating range is determined, and the valve opening and closing state tracking judgment boundary value is obtained.

9. The method of claim 1, wherein the method further comprises: The valve opening and closing state is evaluated by comparing the real-time roughness change information and the opening reading with the valve opening and closing state tracking judgment boundary value, including: The roughness influence factor is calculated by acquiring the real-time roughness change information, and the modified opening value is obtained by multiplying the opening reading; The modified opening value is compared with the valve opening and closing state tracking judgment boundary value to determine whether the valve opening and closing state is normal or abnormal.

Images