Water supply pipeline leakage loss effect prediction method based on survival analysis

Abstract

The invention belongs to the technical field of urban water supply pipe networks, and provides a water supply pipeline leakage loss effect prediction method based on survival analysis, which comprises the following steps: step 1, collecting and processing pipeline damage point data and undamaged pipeline data of a water supply pipeline; 2, selecting the pipe diameter, the pipe material, the pipe length, the pressure, the road type and the burial depth of the pipeline as input variables, and defining the pipe age as survival time; 3, constructing a pipeline leakage loss effect prediction model by using a random survival forest algorithm; 4, evaluating the performance of the constructed pipeline leakage loss effect prediction model through the consistency index; 5, providing a survival probability curve for the active service pipeline of the specific research area according to the prediction model; according to the method, a survival analysis method is introduced, the survival time characteristic and various influence factors of the pipeline data are comprehensively considered, and the complexity and uncertainty of the data are processed by using the random survival forest algorithm, so that a more accurate pipeline leakage loss effect prediction model is constructed.

Description

Technical Field

[0001] The present invention belongs to the technical field of urban water supply pipe networks, and specifically relates to a prediction method for leakage failure of water supply pipelines based on survival analysis. Background Art

[0002] In the urban water supply pipe network system, the leakage failure of pipelines is a long-existing problem that is difficult to completely solve. Traditional prediction methods for pipeline leakage failure often rely on empirical judgment or simple statistical analysis. These methods cannot comprehensively consider various influencing factors of pipelines and their complex relationships, resulting in inaccurate prediction results. With the continuous expansion of the scale of urban water supply pipe networks and the increasing severity of aging problems, the frequency and impact of pipeline leakage failure are also increasing, bringing serious troubles to urban water supply and residents' lives.

[0003] Currently, although some advanced prediction methods have been proposed, such as prediction models based on machine learning, etc., these methods still have deficiencies in dealing with the complexity and uncertainty of pipeline data. In particular, these methods often ignore the survival analysis characteristics of pipeline data, that is, the time process from the pipeline being put into use to the occurrence of leakage failure, which affects the accuracy and reliability of the prediction model to a certain extent.

[0004] Therefore, those skilled in the art have proposed a prediction method for leakage failure of water supply pipelines based on survival analysis to solve the problems raised in the background art. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention provides a prediction method for leakage failure of water supply pipelines based on survival analysis to solve the deficiencies still existing in the prior art in dealing with the complexity and uncertainty of pipeline data. In particular, these methods often ignore the survival analysis characteristics of pipeline data, that is, the time process from the pipeline being put into use to the occurrence of leakage failure, which affects the accuracy and reliability of the prediction model to a certain extent, etc.

[0006] A prediction method for leakage failure of water supply pipelines based on survival analysis includes:

[0007] Step 1: Collect and process the pipeline break point data and unbroken pipeline data (censored data) of the water supply pipeline;

[0008] Step 2: Select the pipe diameter, pipe material, pipe length, pressure, road type, and burial depth of the pipeline as input variables, and define the pipe age as the survival time;

[0009] Step 3: Use the Random Survival Forest (RSF) algorithm to construct a prediction model for pipeline leakage failure;

[0010] Step 4: Evaluate the performance of the constructed pipeline leakage failure prediction model through the concordance index;

[0011] Step 5: Provide a survival probability curve for the in-service pipelines in a specific study area according to the prediction model;

[0012] Step 6: Calculate and output the expected remaining service life of the pipeline by using the mathematical integration method.

[0013] Preferably, in step S1, the pipeline break point data and unbroken pipeline data are sourced from the urban water supply network management system or on-site survey data.

[0014] Preferably, in step S1, when processing the pipeline break point data and unbroken pipeline data of the water supply pipeline, LASSO (Least Absolute Shrinkage and Selection Operator) regression is introduced for feature selection, and features that have a significant impact on the survival time are automatically selected through the penalty term.

[0015] Preferably, in step S3, when constructing the prediction model using the random survival forest algorithm, the complexity and uncertainty of the pipeline data are considered, improving the prediction accuracy of the model.

[0016] Preferably, in step S4, the algorithm formula of the concordance index is as follows:

[0017]

[0018] where mum represents the number of all pairs, d i = 1 indicates that the pipeline has experienced a leakage event, y i and y j represent the actual values of pipelines i and j respectively, I(·) represents the indicator function, and represent the predicted values of pipelines i and j.

[0019] Preferably, the specific description of S5 includes: According to the survival function S(t) obtained from the prediction model, the survival probability curve of the pipeline can be plotted.

[0020] Preferably, in step S5, in order to achieve a fine drawing of the survival probability curve, the kernel density estimation (KDE) algorithm is introduced to draw a smoother and more accurate survival probability curve.

[0021] Preferably, the specific description of S6 includes: According to the survival function S(t) , the expected remaining service life of the pipeline can be calculated.

[0022] Preferably, in step S6, for the confidence interval estimation of the expected remaining service life, the Bootstrap resampling algorithm is used to estimate the confidence interval of the expected remaining service life to reflect the uncertainty of the prediction result;

[0023] Meanwhile, the Bayesian Cox proportional hazards model is introduced to quantify the uncertainty of the expected remaining service life through the posterior distribution.

[0024] A water supply pipeline leakage failure prediction system based on survival analysis, using the above-mentioned water supply pipeline leakage failure prediction method based on survival analysis, includes:

[0025] A data collection and processing module, which is used to collect the pipeline break point data and unbroken pipeline data (censored data) of the water supply pipeline, and introduce LASSO (Least Absolute Shrinkage and Selection Operator) regression for feature selection, and automatically select the features that have a significant impact on the survival time through the penalty term;

[0026] An input variable definition module, which is used to select the pipe diameter, pipe material, pipe length, pressure, road type and burial depth of the pipeline as input variables, and define the pipe age as the survival time;

[0027] A model construction module, which uses the Random Survival Forest (RSF) algorithm to construct a pipeline leakage failure prediction model. The random survival forest algorithm takes into account the complexity and uncertainty of the pipeline data when constructing the prediction model to improve the prediction accuracy of the model;

[0028] A performance evaluation module, which evaluates the performance of the constructed pipeline leakage failure prediction model through the concordance index;

[0029] A survival probability curve providing module, which draws the survival probability curve of the pipeline according to the survival function obtained from the prediction model. In order to achieve the fine drawing of the survival probability curve, the kernel density estimation (KDE) algorithm is introduced to draw a smoother and more accurate survival probability curve;

[0030] An expected remaining service life calculation module, which calculates and outputs the expected remaining service life of the pipeline according to the survival function. For the confidence interval estimation of the expected remaining service life, the Bootstrap resampling algorithm is used to estimate the confidence interval of the expected remaining service life to reflect the uncertainty of the prediction result. Meanwhile, the Bayesian Cox proportional hazards model is introduced to quantify the uncertainty of the expected remaining service life through the posterior distribution;

[0031] A user interface module, which is used to display the survival probability curve and the expected remaining service life to the user.

[0032] A processor configured to execute a prediction method for water supply pipeline leakage failure based on survival analysis as described above.

[0033] A computer-readable storage medium storing a computer program which, when executed by a processor, implements a prediction method for water supply pipeline leakage failure based on survival analysis as described above.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. By introducing the method of survival analysis, the present invention comprehensively considers the survival time characteristics of pipeline data and various influencing factors, and constructs a more accurate prediction model for pipeline leakage failure; using the random survival forest algorithm, the present invention can handle the complexity and uncertainty of pipeline data and improve the accuracy of prediction.

[0036] 2. The present invention introduces LASSO regression for feature selection, automatically selects features that have a significant impact on the survival time through the penalty term, effectively reduces the interference of noise data, and improves the stability and generalization ability of the model.

[0037] 3. The present invention not only provides the prediction result of pipeline leakage failure, but also can draw the survival probability curve of the pipeline according to the prediction model, and calculate and output the expected remaining service life of the pipeline; this provides more intuitive and comprehensive information support for the maintenance and management of urban water supply networks.

[0038] 4. By using the Bootstrap resampling algorithm and the Bayesian Cox proportional hazards model, the present invention estimates the confidence interval of the expected remaining service life to reflect the uncertainty of the prediction result; this helps decision-makers better evaluate the reliability and risk of the prediction result. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a flowchart of the prediction method for water supply pipeline leakage failure based on survival analysis of the present invention;

[0040] Figure 2 is a framework diagram of the prediction system for water supply pipeline leakage failure based on survival analysis of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] The following further describes in detail the embodiments of the present invention with reference to the drawings and examples. The following examples are used to illustrate the present invention, but cannot be used to limit the scope of the present invention.

[0042] Example: The present invention provides a prediction method for water supply pipeline leakage failure based on survival analysis, as Figure 1 shown, including:

[0043] Step 1: Collect and process the pipeline break point data and unbroken pipeline data (censored data) of the water supply pipeline;

[0044] Step 2: Select the pipe diameter, pipe material, pipe length, pressure, road type, and burial depth of the pipeline as input variables, and define the pipe age as the survival time;

[0045] Step 3: Use the Random Survival Forest (RSF) algorithm to construct a pipeline leakage failure prediction model;

[0046] Step 4: Evaluate the performance of the constructed pipeline leakage failure prediction model through the concordance index;

[0047] Step 5: Provide a survival probability curve for the in-service pipelines in a specific study area according to the prediction model;

[0048] Step 6: Adopt a mathematical integration method to calculate and output the expected remaining service life of the pipeline.

[0049] As can be seen from the above, by introducing the method of survival analysis, the survival time characteristics and various influencing factors of pipeline data are comprehensively considered, and the complexity and uncertainty of data are processed by using the random survival forest algorithm, thus constructing a more accurate pipeline leakage failure prediction model; in addition, this method can also draw the survival probability curve of the pipeline and calculate the expected remaining service life, providing intuitive and comprehensive information support for the maintenance and management of urban water supply networks. At the same time, by estimating the confidence interval of the expected remaining service life, the uncertainty of the prediction result is reflected, which helps decision-makers better evaluate the reliability and risk of the prediction result.

[0050] Further, in step S1, the pipeline break point data and unbroken pipeline data are sourced from the urban water supply network management system or on-site survey data.

[0051] Further, in step S1, when processing the pipeline break point data and unbroken pipeline data of the water supply pipeline, LASSO (Least Absolute Shrinkage and Selection Operator) regression is introduced for feature selection, and features that have a significant impact on the survival time are automatically selected through the penalty term. The formula of the LASSO regression includes:

[0052]

[0053] where λ is the penalty parameter that controls the sparsity of features.

[0054] As can be seen from the above, when processing the data of pipeline break points and intact pipeline data in the water supply pipeline, LASSO regression is introduced for feature selection. By means of the penalty term, features that have a significant impact on the survival time are automatically selected. It can effectively reduce the interference of noise data, improve the stability and generalization ability of the model; the penalty parameter λ of LASSO regression can control the sparsity of features, so that some features with less impact on the survival time are automatically eliminated, thus retaining the most critical feature information, which helps to construct a more accurate and efficient pipeline leakage failure prediction model.

[0055] Furthermore, in step S3, when the random survival forest algorithm constructs the prediction model, it takes into account the complexity and uncertainty of pipeline data, improving the prediction accuracy of the model. The formula of the random survival forest algorithm is as follows:

[0056]

[0057] Wherein, is the predicted value of the survival function of the kth tree, and K is the total number of trees.

[0058] As can be seen from the above, when using the random survival forest algorithm to construct the pipeline leakage failure prediction model, the algorithm fully considers the complexity and uncertainty of pipeline data, and this characteristic greatly improves the prediction accuracy of the model; by combining the prediction results of multiple decision trees, the random survival forest algorithm can comprehensively capture the complex relationships and non-linear features in pipeline data, effectively coping with the noise and outliers in the data; in its formula, the predicted values of the survival functions of each tree jointly determine the final prediction result, and the total number of trees reflects the ensemble learning idea of the algorithm. By increasing the number of trees, the stability and prediction accuracy of the model can be further improved; this modeling method considering data complexity and uncertainty enables the prediction model to more accurately reflect the leakage failure situation of the pipeline, providing more reliable decision-making support for the maintenance and management of urban water supply networks.

[0059] Furthermore, in step S4, the algorithm formula of the concordance index is as follows:

[0060]

[0061] Wherein, mum represents the number of all pairs, d i = 1 indicates that a pipeline leakage event has occurred, y i and y j respectively represent the actual values of pipelines i and j, I(·) represents the indicator function, and represent the predicted values of pipelines i and j.

[0062] As can be seen from the above, the higher the consistency index, the more consistent the prediction results of the model are with the actual situation, and the higher the prediction accuracy. This not only provides a clear direction for model optimization, but also ensures that the constructed pipeline leakage failure prediction model can provide more reliable and accurate prediction results in practical applications, providing strong support for the maintenance and management of urban water supply networks.

[0063] Furthermore, the specific description of S5 includes: According to the survival function S(t) obtained from the prediction model, the survival probability curve of the pipeline can be drawn, and its formula specifically includes:

[0064]

[0065] where h(s) is the hazard function.

[0066] As can be seen from the above, drawing the survival probability curve of the pipeline according to the survival function obtained from the prediction model realizes the intuitive display of the pipeline leakage failure risk. The survival probability curve can clearly reflect the failure probability of the pipeline at different time points, providing an important reference basis for the maintenance and management of urban water supply networks. Decision-makers can intuitively understand the health status and potential risks of the pipeline according to the survival probability curve, so as to formulate more scientific and reasonable maintenance plans and management strategies, effectively extend the service life of the pipeline, reduce the occurrence of leakage failures, and ensure the stability and safety of urban water supply.

[0067] Furthermore, in step S5, in order to achieve the fine drawing of the survival probability curve, the kernel density estimation (KDE) algorithm is introduced to draw a smoother and more accurate survival probability curve. The formula of the kernel density estimation algorithm is as follows:

[0068] f h (x)=(1 / n)*ΣK h (x - x i );

[0069] where K h is the kernel function, h is the bandwidth parameter, and x i is the sample data.

[0070] As can be seen from the above, the kernel density estimation algorithm can more precisely capture the distribution characteristics of pipeline leakage failure data. Especially in the case of sparse data or the presence of noise, it can still effectively smooth the curve, reduce fluctuations, and make the survival probability curve closer to the actual situation. This not only improves the readability of the curve, but also provides more accurate and reliable pipeline failure risk information for decision-makers, helping the urban water supply network management department to make more accurate maintenance decisions, further optimizing the operation efficiency of the network, and reducing the leakage risk.

[0071] Furthermore, the specific description of S6 includes: According to the survival functionS(t) , the expected remaining service life of the pipeline can be calculated, and its formula specifically includes:

[0072]

[0073] where T is the failure time, t is the current time.

[0074] As can be seen from the above, the accurate calculation of the expected remaining service life can directly reflect the current health status and remaining service potential of the pipeline, enabling decision-makers to clearly understand the specific maintenance urgency and priority of each pipeline; this not only helps to plan maintenance resources in advance and reasonably arrange maintenance plans, but also effectively avoids risks such as water supply interruption or water quality pollution caused by pipeline aging or sudden failure, thereby ensuring the continuity and safety of urban water supply and improving the overall operation efficiency and economic benefits of the water supply network.

[0075] Furthermore, in step S6, for the confidence interval estimation of the expected remaining service life, by using the Bootstrap resampling algorithm, the confidence interval of the expected remaining service life is estimated to reflect the uncertainty of the prediction result. The formula of the Bootstrap resampling algorithm includes:

[0076] θ * = θ(x 1 , x 2 ,..., x n );

[0077] where θ * is the estimated value after resampling, and x 1 , x 2 ,..., x n are the Bootstrap samples randomly drawn from the original sample;

[0078] At the same time, the Bayesian Cox proportional hazards model is introduced to quantify the uncertainty of the expected remaining service life through the posterior distribution. The formula of the Bayesian Cox proportional hazards model includes:

[0079] h(t|X) = h 0 (t)exp(Xβ);

[0080] where h(tX) is the hazard function, = h 0 (t) is the baseline hazard function, Xβ is the linear predictor, and β is the parameter estimated by the Bayesian method.

[0081] As can be seen from the above, the present invention adopts a method combining the Bootstrap resampling algorithm and the Bayesian Cox proportional hazards model; by using the Bootstrap resampling algorithm, multiple Bootstrap samples can be randomly drawn from the original sample, and then the confidence interval of the expected remaining service life can be estimated, which effectively reflects the uncertainty of the prediction result; at the same time, the Bayesian Cox proportional hazards model is introduced to quantify the uncertainty of the expected remaining service life through the posterior distribution, providing a more accurate and comprehensive risk assessment; the beneficial effect of this combined method is that it can not only give the point estimate value of the expected remaining service life, but also provide its confidence interval, enabling decision-makers to more comprehensively understand the uncertainty range of the prediction result, and thus make more robust and reliable decisions; this helps to reduce the risk brought by prediction errors and improve the scientificity and effectiveness of the maintenance and management of urban water supply networks.

[0082] Furthermore, a comparison of the effects is made between the water supply pipeline leakage failure prediction method based on survival analysis in the embodiment and the currently existing water supply pipeline leakage failure prediction methods (comparative examples), and the following table is obtained:

[0083]

[0084]

[0085] As can be seen from the above table, the water supply pipeline leakage failure prediction method based on survival analysis in the embodiment is superior to the currently existing prediction methods in many aspects, especially in terms of data processing ability, model accuracy, result presentation form, and uncertainty assessment, providing more scientific and effective support for the maintenance and management of urban water supply networks.

[0086] A water supply pipeline leakage failure prediction system based on survival analysis, as Figure 2 shown, uses the above-mentioned water supply pipeline leakage failure prediction method based on survival analysis, and includes:

[0087] A data collection and processing module, which is used to collect the pipeline break point data and unbroken pipeline data (censored data) of the water supply pipeline, and introduce LASSO (Least Absolute Shrinkage and Selection Operator) regression for feature selection, and automatically select the features that have a significant impact on the survival time through the penalty term;

[0088] An input variable definition module, which is used to select the pipe diameter, pipe material, pipe length, pressure, road type, and burial depth of the pipeline as input variables, and define the pipe age as the survival time;

[0089] A model construction module that constructs a pipeline leakage failure prediction model using the Random Survival Forest (RSF) algorithm. The random survival forest algorithm takes into account the complexity and uncertainty of pipeline data when constructing the prediction model to improve the prediction accuracy of the model;

[0090] A performance evaluation module that evaluates the performance of the constructed pipeline leakage failure prediction model through the concordance index;

[0091] A survival probability curve providing module that draws the survival probability curve of the pipeline based on the survival function obtained from the prediction model. To achieve a more refined drawing of the survival probability curve, the Kernel Density Estimation (KDE) algorithm is introduced to draw a smoother and more accurate survival probability curve;

[0092] An expected remaining service life calculation module that calculates and outputs the expected remaining service life of the pipeline based on the survival function. For the confidence interval estimation of the expected remaining service life, the Bootstrap resampling algorithm is used to estimate the confidence interval of the expected remaining service life to reflect the uncertainty of the prediction result. At the same time, the Bayesian Cox proportional hazards model is introduced to quantify the uncertainty of the expected remaining service life through the posterior distribution;

[0093] A user interface module for displaying the survival probability curve and the expected remaining service life to the user.

[0094] Working principle: First, collect and process the pipeline break point data and non - broken pipeline data of the water supply pipeline. Through the introduction of LASSO regression for feature selection, automatically screen out the feature variables that have a significant impact on pipeline leakage failure; then, select the pipe diameter, pipe material, pipe length, pressure, road type, and burial depth of the pipeline as input variables, define the pipe age as the survival time, and use the random survival forest algorithm to construct a pipeline leakage failure prediction model. This algorithm fully considers the complexity and uncertainty of pipeline data and improves the prediction accuracy of the model; then, evaluate the performance of the model through the concordance index to ensure the reliability of the model; subsequently, draw the survival probability curve of the pipeline based on the survival function obtained from the prediction model, and introduce the kernel density estimation algorithm to make the curve smoother and more accurate; finally, calculate the expected remaining service life of the pipeline according to the survival function, and use a method combining the Bootstrap resampling algorithm and the Bayesian Cox proportional hazards model to estimate the confidence interval of the expected remaining service life to comprehensively reflect the uncertainty of the prediction result and provide a scientific basis for the maintenance and management of the urban water supply network.

[0095] An embodiment of the present application provides an electronic device applicable to the above - mentioned water supply pipeline leakage failure prediction method based on survival analysis, including:

[0096] A memory for protecting computer programs and data;

[0097] A processor for running system programs.

[0098] An embodiment of the present application provides a computer storage medium, which is applicable to the above-mentioned method for predicting leakage failure of water supply pipelines based on survival analysis, and performs hierarchical confidentiality management on the above-mentioned system and data according to the requirements of confidentiality management.

[0099] Those skilled in the art should understand that the embodiments of the present application can be provided as a system or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0100] The present application is described with reference to the flowcharts and / or block diagrams of the devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0101] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0102] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0103] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0104] Memory includes non-permanent memory in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0105] Computer readable media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0106] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, commodity or device including the elements.

[0107] The embodiments of the present invention are provided for the purpose of illustration and description. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations of the present invention. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for predicting water supply pipeline leakage failure based on survival analysis, characterized in that: include: Step 1: Collect and process the damaged point data and undamaged pipeline data of the water supply pipeline; Step 2: Select the pipeline diameter, pipe material, pipe length, pressure, road type and burial depth as input variables, and define the pipeline age as the survival time; Step 3: Use the random survival forest algorithm to build a pipeline leakage failure prediction model; Step 4: Evaluate the performance of the constructed pipeline leakage failure prediction model through the consistency index; Step 5: Provide a survival probability curve for the active pipelines in a specific study area based on the prediction model; Step 6: Use mathematical integration method to calculate and output the expected remaining service life of the pipeline.

2. A method for predicting water supply pipeline leakage failure based on survival analysis as claimed in claim 1, characterized in that: In step S1, the pipeline damaged point data and the undamaged pipeline data are derived from the urban water supply network management system or on-site survey data.

3. A method for predicting water supply pipeline leakage failure based on survival analysis as claimed in claim 1, characterized in that: In step S1, the pipeline damaged point data and the undamaged pipeline data of the processed water supply pipeline are introduced into LASSO regression for feature selection, and the features that have a significant impact on the survival time are automatically selected through the penalty term.

4. A method for predicting water supply pipeline leakage failure based on survival analysis as claimed in claim 1, characterized in that: In step S3, the random survival forest algorithm considers the complexity and uncertainty of pipeline data when constructing the prediction model.

5. A method for predicting water supply pipeline leakage failure based on survival analysis as claimed in claim 1, characterized in that: In step S4, the algorithm formula of the consistency index is as follows: Where mum represents the number of all pairs, d i =1 means the pipeline has experienced leakage, y i With y j represent the actual values ​​of pipelines i and j respectively, I(·) represents the indicator function, and Represents the predicted values ​​of pipelines i and j.

6. A method for predicting water supply pipeline leakage failure based on survival analysis as claimed in claim 1, characterized in that: The specific description of S5 includes: according to the survival function obtained by the prediction model, a pipeline survival probability curve can be drawn.

7. A method for predicting water supply pipeline leakage failure based on survival analysis as claimed in claim 1, characterized in that: In step S5, in order to achieve precise drawing of the survival probability curve, a kernel density estimation algorithm is introduced.

8. A method for predicting water supply pipeline leakage failure based on survival analysis as claimed in claim 1, characterized in that: The specific description of S6 includes: according to the survival function, the expected remaining service life of the pipeline can be calculated.

9. A water supply pipeline leakage failure prediction system based on survival analysis, characterized in that: A method for predicting water supply pipeline leakage failure based on survival analysis according to any one of claims 1 to 8, comprising: The data collection and processing module is used to collect the damaged point data and undamaged pipeline data of the water supply pipeline, and introduce LASSO regression for feature selection, and automatically select the features that have a significant impact on the survival time through the penalty term; The input variable definition module is used to select the pipeline diameter, pipe material, pipe length, pressure, road type and buried depth as input variables, and define the pipe age as survival time; A model building module, using a random survival forest algorithm to build a pipeline leakage failure prediction model, wherein the random survival forest algorithm takes into account the complexity and uncertainty of pipeline data when building the prediction model; The performance evaluation module evaluates the performance of the constructed pipeline leakage failure prediction model through the consistency index; The survival probability curve provides a module that draws the pipeline's survival probability curve based on the survival function obtained by the prediction model. In order to achieve precise drawing of the survival probability curve, the kernel density estimation algorithm is introduced; The expected remaining service life calculation module calculates and outputs the expected remaining service life of the pipeline according to the survival function, and estimates the confidence interval of the expected remaining service life; The user interface module is used to display the survival probability curve and the expected remaining useful life to the user.

Images