Water supply network corrosion pipe section identification and diagnosis method based on water quality inversion

Abstract

The invention provides a water supply network corrosion pipe section identification and diagnosis method based on water quality inversion, and the method comprises the steps: collecting multi-point water quality data and pipe section attribute information in a water supply network area, judging a topological water flow direction based on residual chlorine attenuation characteristics, determining a suitable diagnosis node and a target independent pipe section, and carrying out the recognition and diagnosis of the corrosion pipe section of the water supply network. And monitoring water quality parameter change characteristics in the pipe section, and realizing target positioning of the corroded pipe section and rapid inversion diagnosis of the corrosion state in combination with a quantitative evaluation model of the water quality variation and the corrosion degree of the pipe section. An input layer of the model is water quality change characteristics and pipe section information between pipe network diagnosis nodes, an operation layer adopts an algorithm based on a mapping relation between water quality decay characteristics and corrosion states, an output layer is a pipe section corrosion state score, and when a diagnosis result is lower than a preset safety threshold value, early warning information and maintenance suggestions are generated. According to the method, the diagnosis effect that the decision coefficient exceeds 0.95 can be achieved, and the safety and reliability of operation management of the water supply network are effectively improved.

Description

Technical Field

[0001] The present invention belongs to the field of water supply network detection, and relates to a method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion. Background Art

[0002] Drinking water safety is crucial to people's well-being and socioeconomic development, and ensuring water quality is crucial. Water supply networks are a crucial component of urban water supply systems, their primary function being to deliver treated drinking water to consumers. However, with the increasing age of these networks and the long-term effects of complex water quality environments, corrosion within water supply pipes is becoming increasingly prominent. Internal corrosion within pipe walls not only deteriorates water quality and can lead to secondary pollution problems such as excessive heavy metals, increased turbidity, and increased color, seriously impacting the safety and quality of tap water, but the accumulation of corrosion products also impairs hydraulic performance, increases energy consumption and network operating costs, shortens pipe life, and impacts the stable operation of urban water supply systems.

[0003] Internal corrosion detection, pipeline cleaning, maintenance, and renovation of water supply networks are currently important operational contents for water supply companies, but they usually have problems such as high costs, low diagnostic efficiency, and untimely maintenance. In non-excavation pipeline detection, related technologies such as a pipeline internal corrosion detection device for pressure pipelines (publication number CN119246686A) and a water pipeline inner wall inspection equipment and detection method (authorization announcement number CN118998514B) use ultrasonic detection, optical detection and other sensor technologies to detect the internal conditions of the pipeline. However, there are problems such as high equipment costs and inability to achieve continuous monitoring. The pipeline internal detection method and system based on multi-sensor fusion (authorization announcement number CN119150175B) predicts the pipeline status by fusing multi-sensor data and analyzing historical data, but it relies on a large amount of historical data, has high maintenance costs, and cannot accurately judge the pipeline status. A method and device for monitoring corrosion in supercritical steam pipelines (publication number CN 119848752A) only monitors and simulates water quality at the inlet and outlet of the steam pipelines. This method is unable to rapidly locate and diagnose the corrosion status of target pipe sections in water supply networks with multiple nodes and complex hydraulic conditions. Therefore, a method that can quickly and accurately quantify the corrosion status of water supply networks and provide real-time dynamic early warning is urgently needed to ensure the safety and reliability of network operation. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for rapid inversion diagnosis of the corrosion state of the inner wall of a water supply network by monitoring the decay characteristics of water flow and water quality parameters in multiple nodes in a water supply network area, combining the network operation data to construct a corrosion degree evaluation model, quickly locating and accurately quantifying the corrosion state of the target pipe section.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion comprises the following steps:

[0007] Step S1: Obtain water quality data and pipe segment attribute information for multiple nodes within the water supply network area;

[0008] Step S2: Determine the topological water flow direction by the direction of residual chlorine concentration attenuation change, identify appropriate diagnostic nodes based on the diagnostic target, eliminate branch pipe interference, and determine the target independent pipe section;

[0009] Step S3: extracting features from the water quality data at two points upstream and downstream of the target independent pipe section to obtain water quality decay features that reflect the corrosion status of the water supply pipeline;

[0010] Step S4: Using the water quality decay characteristics and the pipe section information as input data for the input layer, constructing a regression model in the computation layer to characterize the mapping relationship between the water quality decay characteristics and the corrosion state, performing an inverse calculation on the corrosion state of the inner wall of the water supply pipe, obtaining an evaluation score for the degree of corrosion, and outputting the evaluation score through the output layer;

[0011] Step S5: quickly diagnose the corrosion degree of the water supply pipeline according to the evaluation score, and make early warning prompts or maintenance decisions.

[0012] Preferably, in step S1, the water quality data includes conventional indicators and corrosion index parameters, and the acquisition method includes collecting samples from fire hydrants and user taps for testing and regularly uploading updated water quality monitoring data to the GIS system;

[0013] The conventional indicators include turbidity, total chlorine concentration, free chlorine concentration, pH value, ammonia nitrogen concentration, iron ion concentration, manganese ion concentration, total organic carbon (TOC), total nitrogen concentration (TN), and total trihalomethane concentration (THMs);

[0014] The corrosion indicators include calcium hardness, bicarbonate concentration, chloride ion concentration, sulfate concentration, Langelier saturation index, Larson index, and calcium carbonate precipitation potential (CCPP);

[0015] Among them, the conventional indicators adopt the indicators specified in the "Sanitary Standard for Drinking Water" GB5749; the corrosion indicators adopt indicators that assist in evaluating the potential of water bodies to cause metal corrosion and scale formation.

[0016] Preferably, in step S1, the attribute information of the target independent pipe section is obtained through a water supply network GIS system, including pipe age, pipe diameter, pipe section length, water flow velocity, and hydraulic retention time.

[0017] Preferably, in step S2, the total chlorine concentration of each node in the water supply network topology satisfies the following equation: in is the total chlorine concentration vector, is the total chlorine concentration change vector between nodes, is the node water supply vector, The flow direction of the water supply network topology is determined based on the direction of the residual chlorine concentration attenuation. Nodes with high residual chlorine concentrations on the same pipe segment are considered upstream, while nodes with low residual chlorine concentrations are considered downstream.

[0018] Preferably, in step S2, in order to achieve higher judgment accuracy, the distance range between the target pipe segment and the corresponding diagnosis nodes must meet the following conditions:

[0019] There should be no branch confluence on the connection path between two nodes to limit the maximum distance between nodes. The difference in total chlorine concentration between two nodes must satisfy ΔC>n×LOD, where LOD is the detection limit of total chlorine concentration and n is a coefficient determined according to the diagnostic accuracy requirements to limit the minimum distance between nodes.

[0020] Preferably, in step S3, the water quality decay characteristic refers to the water quality change per unit hydraulic residence time or per unit pipe length, which is calculated by the water quality index parameters of the diagnosis node and the attribute parameters of the corresponding target independent pipe segment, including the time decay variable X=(C2-C1) / T, the process decay variable X=(C2-C1) / L, and the space decay variable X=4·(C2-C1) / L·π·D 2 , where C2 is the downstream water quality parameter, C1 is the upstream water quality parameter, L is the pipe length, T is the hydraulic retention time, and D is the pipe diameter.

[0021] Preferably, in step S4, the algorithm feature of the operation layer of the inversion calculation is Y=f(X), where f uses a regression model to establish a mapping relationship between the input independent variable and the output dependent variable to achieve the inversion calculation of the corrosion state of the target independent pipe section; specific fitting methods include polynomial regression model, multiple linear regression model, random forest regression model, and neural network regression model; among the fitting methods, the multiple linear regression model is the preferred regression model: f(X)=β0+β1X1+β2X2+…+β p X p , where X is the input feature variable vector, β0 is the intercept term, β1,β2,…,β p is the regression coefficient; the regression coefficient is calculated using the least squares method Estimate, where X is the input sample matrix and Y is the column vector of observations corresponding to the sample.

[0022] Preferably, the corrosion status evaluation score of the output layer is used to reflect the physical and chemical properties and corrosion characteristics of the inner surface of the pipeline, including the corrosion rate, the composition of corrosion products, the remaining life of the pipeline and its impact on water quality.

[0023] Preferably, the regression model needs to meet the following conditions:

[0024] Coefficient of determination (R 2 )>0.9, the residual distribution should meet the normality assumption and have no obvious heteroscedasticity, the residual mean should be close to zero, the calculated variance inflation factor VIF<10, and the Durbin-Watson test (DW value) should be between 1.5 and 2.5; among the candidate models that meet the above conditions, the determination coefficient (R 2 ) is the highest, the Durbin-Watson test (DW value) is close to 2, and the root mean square error (RMSE) and mean absolute error (MAE) are the smallest models as the optimal regression model.

[0025] Preferably, in step S5, the warning limit of the corrosion status evaluation score is 0, a score higher than 0 indicates that the pipeline is in good condition, a score lower than 0 indicates that there is obvious corrosion on the inner wall of the pipeline, and the lower the evaluation score, the more serious the corrosion.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] (1) Reduce manpower and resource waste: The present invention uses water quality monitoring and real-time data analysis to quickly identify corroded pipe sections based on water quality change inversion, prioritize high corrosion risk areas for maintenance, and achieve efficient resource allocation;

[0028] (2) Rapid diagnosis and accurate early warning: This invention adopts a regression model based on optimal parameters, combined with the network attributes and water quality decay characteristics, to accurately determine the topology and water flow direction of the water supply network, locate the target corrosion pipe section, diagnose the pipeline corrosion status and issue a timely early warning to ensure the safe operation of the network;

[0029] (3) Intelligent monitoring and management: This invention establishes an algorithm to invert the pipeline status based on the change in water quality parameters, thereby realizing intelligent perception, dynamic updating and refined management of the corrosion status of the water supply network, and improving the efficiency of corrosion prevention and control of the pipeline network;

[0030] (4) Wide applicability and scalability: The method of the present invention can be integrated into the existing water supply network management system and flexibly adjusted to different water quality conditions and network structures, and has good applicability and scalability.

[0031] The principle of this invention is to collect water quality data and pipe segment attribute information from multiple points within a water supply network area. Based on the residual chlorine decay characteristics, suitable diagnostic nodes and target independent pipe segments are identified. A quantitative evaluation model is then constructed to measure water quality changes and the degree of corrosion in the target pipe segments. The model's input layer is the water quality variation characteristics and pipe segment information between the network's diagnostic nodes. The computational layer uses an algorithm based on the mapping between water quality decay characteristics and corrosion status. The output layer is the pipeline corrosion status score. When the diagnostic result falls below a preset safety threshold, early warning information and maintenance recommendations are generated. The constructed model can achieve a diagnostic effect with a coefficient of determination exceeding 0.95, and provides dynamic assessment and precise early warning capabilities for the corrosion status of the water supply network, effectively improving the safety and reliability of water supply network operation and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A schematic flow chart of a method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion provided by an embodiment of the present invention;

[0033] Figure 2 The embodiment of the present invention provides a method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion. The method is implemented with a preferred multiple linear regression model and the coefficient of determination R is used. 2 As the evaluation index. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0035] like Figure 1 As shown in FIG, a method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion is provided in this embodiment, including:

[0036] Step S1: Acquire water quality data and pipe segment attribute information of multiple nodes in the water supply network area; pipe segment attribute information includes pipe age, pipe diameter, water flow velocity, and hydraulic retention time;

[0037] Step S2: Determine the topological water flow direction by the direction of residual chlorine concentration attenuation change, identify appropriate diagnostic nodes based on the diagnostic target, eliminate branch pipe interference, and determine the target independent pipe section;

[0038] Step S3: The variables are the attribute information of the target pipe section, such as pipe age and pipe diameter, as well as the water quality decay characteristics obtained by solving the water quality data of the upstream and downstream points of the target independent pipe section and the pipe section attribute information;

[0039] Step S4: Operation layer: Use the optimal multiple linear regression model

[0040] Y=β0+β1X1+β2X2+…+βp X p

[0041] The model can achieve the coefficient of determination (R 2 )>0.9, the variance inflation factor (VIF) of each variable is less than 10, and the F test and Durbin-Watson test (DW value) are within the range of 1.5 to 2.5; the output variable is the pipeline corrosion status evaluation score.

[0042] Step S5: Rapidly diagnose the degree of corrosion of the water supply network based on the evaluation score, and make early warning prompts or maintenance decisions.

[0043] In steps S2 through S5, water quality monitoring and real-time data analysis rapidly identify target pipe sections for diagnosis and assess the corrosion status of the network. After data collection is complete, the computational layer automatically executes the entire process of feature calculation, model inference, and early warning determination, encompassing steps S2 through S5. This system enables rapid inverse assessment and response decision-making of the corrosion status of the water supply network, enabling immediate identification and dynamic management of corrosion risks, significantly improving diagnostic efficiency and resource allocation accuracy.

[0044] Example 1

[0045] This embodiment is based on a water supply network in a certain place in the southeast region. Based on the model established in the embodiment, a water supply network with a pipe age ranging from 5 to 50 years is analyzed. Figure 1 The process is used to screen the diagnostic nodes and pipe sections and diagnose and evaluate the corrosion situation. The model written in the operation layer specifically selects the multivariate linear regression model. The conditions for verifying that the model meets the requirements are that the variance inflation factor (VIF) is less than 10, the Durbin-Watson test (DW value) is between 1.5 and 2.5, and the determination coefficient R 2 >0.9.

[0046] The data variables selected for input variables include the age, diameter, and length of the selected pipe section, as well as the turbidity between nodes, total chlorine concentration, free chlorine concentration, pH value, ammonia nitrogen concentration, calcium hardness, bicarbonate concentration, chloride ion concentration, sulfate concentration, Langelier saturation index (LSI), Larson index, calcium carbonate precipitation potential (CCPP), iron ion concentration, manganese ion concentration, total organic carbon (TOC), and total nitrogen (TN). Water quality parameter decay characteristics, where the decay characteristics are solved based on the decay of water quality parameters per unit length of the pipe section:

[0047]

[0048] The above-mentioned pipe section parameters and water quality decay parameters are used as input variable matrices and screened by stepwise regression method. One of the screening conditions is to use variance inflation factor (VIF) to detect multicollinearity between independent variables, and VIF must be less than 10:

[0049]

[0050] The second screening condition is to check the independence of the residuals through the Durbin-Watson test. The DW value is in the range of 1.5 to 2.5, which meets the requirements:

[0051]

[0052] Input the independent variables obtained by screening and use the multiple linear regression model operation:

[0053] Y=β0+β1X1+β2X2+…+β p X p

[0054] The determination coefficient (R 2 ) The calculation effect is shown in Figure 2 , calculated by comparing the actual value of the test dataset with the model calculation value:

[0055]

[0056] where y i is the actual value of the corrosion score, is the model calculation value. In this implementation of the embodiment, the determination coefficient (R 2 ) is 0.9545 ≥ 0.9, indicating that this embodiment can explain over 95% of the variance in the calculated pipe segment corrosion conditions, a very close match. Among the 18 pipe groups, 10 pipe segments were identified as corroded based on the scoring threshold, which is consistent with the actual situation. This shows that this embodiment has a 100% accuracy rate in diagnosing the presence of inner wall corrosion in all 18 pipe segments. It can accurately diagnose the corrosion status of a pipe segment based on the input pipe segment parameters and water quality decay characteristic parameters, avoiding the high costs and waste of resources caused by blind excavation.

[0057] Example 2

[0058] This example uses a water supply network in a southeastern region as the implementation context. Based on the system established in this example, an optimal multivariate linear regression model is selected, and the corrosion warning limit is set to 0. The pipeline parameters and water quality decay characteristics of a 22-year-old, 1400mm diameter ductile iron pipe segment in the southeastern region are substituted into the optimal regression model. Under this implementation, this example calculates a corrosion score of 2.09 for this segment, which exceeds the set corrosion warning limit. This indicates that the segment is in good condition and does not require excavation for maintenance or replacement.

[0059] The above-mentioned diagnosis and maintenance of corrosion in a single pipe segment does not require excavation. This embodiment only tests eight water quality parameters at two points on the pipe segment. Combined with the GIS system, the operating cost is approximately 1,000 yuan, and the diagnosis can be completed in 2 hours. This is far less than the cost required for excavation, saving the manpower and material resources that would be wasted in excavation and pipeline inspection.

[0060] Comparative Example 1

[0061] A method of directly performing pipeline maintenance based on pipe age was selected as a comparative example. Maintaining all other application conditions identical to those of the previous embodiment, the corrosion diagnosis process based on this embodiment was applied to the pipe segment data from Example 1. Using pipes older than 30 years as the screening criteria, calculations were performed using the optimal regression model, with 0 set as the corrosion warning limit. The calculation results are shown in Table 1. Of the six pipe segments, those older than 50 years were directly replaced, while those younger than 50 years were not. Three of these pipelines were treated inappropriately, resulting in not only the failure to promptly replace and maintain corroded pipes but also the replacement of unnecessary pipes, resulting in a waste of human and material resources.

[0062] Table 1 compares the actual treatment conditions and calculation results of pipe sections with an age of more than 30 years.

[0063] Table 1

[0064]

[0065]

[0066] Comparative Example 2

[0067] Two methods, endoscopic inspection and direct excavation for pipeline maintenance and replacement, were selected as comparative examples to compare the cost, time consumption, and scope of application of the three methods, as shown in Table 2. Maintaining the same application conditions as in the previous example, the same 480m long, DN1400 ductile iron pipe section in Example 2 was subjected to endoscopic inspection. Sediments within the pipe must be removed in advance using a high-pressure water gun or mechanical pig. The inspection equipment is lowered into the pipeline through a manhole or reserved opening. The probe advances at a constant speed and transmits data back, which is then marked and analyzed by engineers. Direct costs include equipment rental, labor, and material costs, while indirect costs include lane occupation, minor dust and noise. The construction period is approximately 3 days, the inspection cost is approximately 60,000 yuan / km, and it relies on the use of endoscopic equipment and professional operation. The direct costs of excavating the pipeline include equipment fees, labor costs, material costs, trench excavation fees, backfill compaction fees, and slag transportation fees generated during the excavation, transportation, and landfill processes. Indirect costs include social costs such as traffic congestion and damage to existing facilities, and environmental costs such as noise, dust, and rainwater erosion of the soil at the construction site. The construction period is approximately 60 days, and the total pipeline maintenance cost is approximately RMB 5 million per km. After actual excavation and replacement, it was found that the original pipeline had not suffered obvious corrosion, resulting in a large amount of resource waste.

[0068] Table 2 shows a cost comparison between this embodiment and currently commonly used water supply pipeline detection technologies.

[0069] Table 2

[0070]

[0071]

[0072] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0073] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0074] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion, characterized in that: The steps include: Step S1: Obtain water quality data and pipe segment attribute information for multiple nodes within the water supply network area; Step S2: Determine the topological water flow direction by the direction of residual chlorine concentration attenuation change, identify appropriate diagnostic nodes based on the diagnostic target, eliminate branch pipe interference, and determine the target independent pipe section; Step S3: extracting features from the water quality data at two points upstream and downstream of the target independent pipe section to obtain water quality decay features that reflect the corrosion status of the water supply pipe section; Step S4: Using the water quality decay characteristics and the pipe section information as input data for the input layer, a regression model is constructed in the computation layer to characterize the mapping relationship between the water quality decay characteristics and the corrosion state. The corrosion state of the inner wall of the target pipe section of the water supply network is inverted and calculated to obtain an evaluation score for the corrosion degree. The evaluation score is then outputted through the output layer. Step S5: quickly diagnose the corrosion degree of the water supply pipe section according to the evaluation score, and make early warning prompts or maintenance decisions.

2. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 1 is characterized in that: In step S1, the water quality data includes conventional indicators and corrosion index parameters, and the acquisition method includes collecting samples from fire hydrants and user taps for testing and regularly uploading updated water quality monitoring data into the GIS system; The conventional indicators include turbidity, total chlorine concentration, free chlorine concentration, pH value, ammonia nitrogen concentration, iron ion concentration, manganese ion concentration, total organic carbon (TOC), total nitrogen concentration (TN), and total trihalomethane concentration (THMs); The corrosion indicators include calcium hardness, bicarbonate concentration, chloride ion concentration, sulfate concentration, Langelier saturation index, Larson index, and calcium carbonate precipitation potential (CCPP); Among them, the conventional indicators adopt the indicators specified in the "Sanitary Standard for Drinking Water" GB5749; the corrosion indicators adopt indicators that assist in evaluating the potential of water bodies to cause metal corrosion and scale formation.

3. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 1 is characterized in that: In step S1, the attribute information of the target independent pipe section is obtained through the water supply network GIS system, including pipe age, pipe diameter, pipe section length, water flow velocity, and hydraulic retention time.

4. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 1 is characterized in that: In step S2, the total chlorine concentration of each node in the water supply network topology satisfies the following equation: in is the total chlorine concentration vector, is the total chlorine concentration change vector between nodes, is the node water supply vector, The flow direction of the water supply network topology is determined based on the direction of the residual chlorine concentration attenuation. Nodes with high residual chlorine concentrations on the same pipe segment are considered upstream, while nodes with low residual chlorine concentrations are considered downstream.

5. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 1 is characterized in that: In step S2, in order to achieve higher judgment accuracy, the distance range between the target pipe segment and the corresponding diagnosis nodes must meet the following conditions: There should be no branch confluence on the connection path between two nodes to limit the maximum distance between nodes. The difference in total chlorine concentration between two nodes must satisfy ΔC>n×LOD, where LOD is the detection limit of total chlorine concentration and n is a coefficient determined according to the diagnostic accuracy requirements to limit the minimum distance between nodes.

6. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 1 is characterized in that: In step S3, the water quality decay characteristic refers to the water quality change per unit hydraulic residence time or per unit pipe length, which is calculated by the water quality index parameters of the diagnosis node and the attribute parameters of the corresponding target independent pipe segment, including the time decay variable X = (C2-C1) / T, the process decay variable X = (C2-C1) / L, and the space decay variable X = 4·(C2-C1) / L·π·D 2 , where C2 is the downstream water quality parameter, C1 is the upstream water quality parameter, L is the pipe length, T is the hydraulic retention time, and D is the pipe diameter.

7. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 1 is characterized in that: In step S4, the algorithm feature of the operation layer of the inversion calculation is Y=f(X), where f uses a regression model to establish a mapping relationship between the input independent variable and the output dependent variable to achieve the inversion calculation of the corrosion state of the target independent pipe section; specific fitting methods include polynomial regression model, multiple linear regression model, random forest regression model, and neural network regression model; among the fitting methods, the multiple linear regression model is the preferred regression model: f(X)=β0+β1X1+β2X2+…+β p X p , where X is the input feature variable vector, β0 is the intercept term, β1,β2,…,β p is the regression coefficient; the regression coefficient is calculated using the least squares method Estimate, where X is the input sample matrix and Y is the column vector of observations corresponding to the sample.

8. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 7 is characterized in that: The corrosion status evaluation score of the output layer is used to reflect the physical and chemical properties and corrosion characteristics of the inner surface of the pipeline, including the corrosion rate, the composition of corrosion products, the remaining life of the pipeline and its impact on water quality.

9. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 7 is characterized in that: The regression model must meet the following conditions: Coefficient of determination (R 2 )>0.9, the residual distribution should meet the normality assumption and have no obvious heteroscedasticity, the residual mean should be close to zero, the calculated variance inflation factor VIF<10, and the Durbin-Watson test (DW value) should be between 1.5 and 2.5; among the candidate models that meet the above conditions, the determination coefficient (R 2 ) is the highest, the Durbin-Watson test (DW value) is close to 2, and the root mean square error (RMSE) and mean absolute error (MAE) are the smallest models as the optimal regression model.

10. The method for identifying and diagnosing corroded pipe sections in a water supply network based on water quality inversion according to claim 1, characterized in that: In step S5, the warning limit of the corrosion status evaluation score is 0. A score higher than 0 indicates that the pipe section is in good condition, and a score lower than 0 indicates that there is obvious corrosion on the inner wall of the pipe section. The lower the evaluation score, the more serious the corrosion.

Images