A method and system for predicting the risk of pipe burst in water supply network

Abstract

The present invention provides a method and system for predicting the risk of pipe burst in a water supply network. The method includes: using the principal component analysis method to perform principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set to determine the index factor set; using the hierarchical analysis method to determine the index weights corresponding to each of the indicators in the index factor set; establishing a pipe burst risk prediction model; determining the pipe burst risk prediction model parameters; and calculating the pipe burst rate of each pipe section and the total number of pipe bursts in the region in the future time period through the pipe burst risk prediction model with the determined parameters. The present invention can solve the problem that the existing prediction methods do not fully consider the factors affecting the pipe burst of the water supply network and do not consider the correlation between the factors while ensuring that data information is not lost as much as possible, thereby improving the accuracy of predicting the risk of pipe burst in the regional water supply network, and thus being able to reasonably avoid the danger caused by the pipe burst of the water supply network.

Description

Technical Field

[0001] The present invention relates to the technical field of risk prediction, and in particular to a method and system for predicting the risk of pipe burst in a water supply network. Background Art

[0002] The urban water supply network system is responsible for delivering and distributing water to users, and is an indispensable infrastructure to ensure the normal operation of urban production and life. In recent years, with the continuous acceleration of my country's urbanization process, urban water demand has soared, the scale of water supply networks has gradually expanded, and the network system has become increasingly complex. At the same time, my country is in the "aging" stage of underground pipe networks. Due to the aging of the network and insufficient investment in network construction, pipe burst accidents occur frequently, which have caused varying degrees of impact on urban production and residents' lives, and at the same time caused a large amount of waste of clean water resources. Therefore, the use of scientific means to predict the risk of pipe bursts in water supply networks, identify high-risk pipe sections, and take targeted corresponding operation and maintenance measures can effectively prevent the occurrence of pipe burst accidents and ensure the safe operation of urban water supply networks.

[0003] At present, common water supply pipe burst risk prediction models mainly include physical models, statistical models and pipe burst risk prediction models based on intelligent algorithms. Among them, the physical model predicts pipeline accidents by trying to analyze the load acting on the pipeline, the ability of the pipeline to resist loads, and the degree and scope of corrosion inside and outside the pipeline. However, the construction of the physical model faces problems such as complex physical mechanisms of pipe bursts and difficulty in obtaining required data. The intelligent algorithm model can obtain complex and nonlinear correlations between different variables in the system through data learning, but a large amount of historical data on pipe bursts is required to train the model, and the model simulation process is complicated. The statistical model is based on the historical pipe burst data of the pipe network operation, and uses statistical methods to quantify the laws of pipe burst accidents. The model simulation process is simple and highly executable.

[0004] The water supply pipe burst risk prediction model based on Cox survival analysis is a statistical model. It applies the survival analysis theory to the estimation of pipe survival time, and adds many risk factors that affect pipe burst, such as pipe characteristics and environmental factors. It can simultaneously analyze the impact of multiple factors on pipe burst and realize the scientific prediction of pipe burst risk. At present, the existing water supply pipe burst risk prediction model based on the Cox proportional risk model usually only considers two risk factors, namely pipe diameter and pipe material. However, the pipe network will be affected by multiple factors during operation, and the causes of pipe burst are complex. Only by comprehensively considering various factors can a scientific and reasonable pipe burst risk prediction model be constructed. At the same time, there will be correlation between multiple pipe burst factors. Existing studies directly substitute the values ​​of pipe burst influencing factors as risk factors into the model for prediction, without considering the correlation between multiple factors. The data information overlaps, which will lead to the inconsistency between the pipe burst risk prediction results and the actual situation, thus failing to ensure the safe operation of the urban water supply network. Summary of the invention

[0005] Based on this, the purpose of the present invention is to provide a method and system for predicting the risk of pipe burst in a water supply network, so as to improve the accuracy of the prediction results of the risk of pipe burst in a water supply network.

[0006] To achieve the above object, the present invention provides a method for predicting the risk of pipe burst in a water supply network, the method comprising:

[0007] Step S1: construct a pipe burst factor set;

[0008] Step S2: using the principal component analysis method, performing principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set to determine an indicator factor set;

[0009] Step S3: using the hierarchical analysis method to determine the indicator weight corresponding to each indicator in the indicator factor set, and the indicator with the highest indicator weight is taken as the most important indicator;

[0010] Step S4: establishing a pipe burst rate benchmark function according to the most important indicator; taking the remaining indicators as risk factors, establishing a pipe burst risk related conditional strength function according to the risk factors; and establishing a pipe burst risk prediction model according to the pipe burst rate benchmark function and the pipe burst risk related conditional strength function based on the survival analysis theory;

[0011] Step S5: determining pipe burst risk prediction model parameters;

[0012] Step S6: Calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in the future time period through the pipe burst risk prediction model with determined parameters.

[0013] Optionally, the burst factor set includes basic attribute data of the water supply network, water supply network operation data, surrounding environment and facility data, and the number of bursts in the network. The basic attribute data of the water supply network includes: burial depth, pipe diameter, pipe material and pipe age; the water supply network operation data includes: network operation pressure; the surrounding environment and facility data includes: temperature difference, road grade, soil corrosion index, pH value, redox point, drainage conditions, location, pipeline identification, preventive measures and public education.

[0014] Optionally, the indicator factor set specifically includes: burial depth, pipe diameter, pipe pressure, pipe material, temperature difference, road load, pipe age, historical accidents, soil environment and regional attributes.

[0015] Optionally, determining the pipe burst risk prediction model parameters specifically includes:

[0016] Step S51: group the pipe burst records in year y according to the pipe burst age, with a group interval of 10 years, and calculate the annual average number of pipe bursts;

[0017] Step S52: calculating the pipe burst rate of the pipe network according to the annual average number of pipe bursts;

[0018] Step S53: fitting quadratic parabola regression parameters with the median of the pipe age group and the pipe burst rate of the pipe network as dependent variables to calculate pipe burst risk prediction model parameters, wherein the pipe burst risk prediction model parameters include the fitting coefficient of the pipe burst rate benchmark function and the regression coefficient corresponding to the dth risk factor.

[0019] Optionally, the pipe burst rate of each pipe section and the total number of pipe bursts in the region in a future time period are calculated by the pipe burst risk prediction model with determined parameters, specifically including:

[0020] Step S61: Calculate various indicators corresponding to each pipe section by principal component analysis method, and substitute them into the pipe burst risk prediction model with determined parameters to obtain the pipe burst rate per unit pipe length;

[0021] Step S62: multiplying the pipe burst rate per unit pipe length by the pipe length corresponding to each pipe section to obtain the pipe burst rate corresponding to each pipe section;

[0022] Step S63: Integrate the pipe burst rate corresponding to each pipe segment with respect to time to obtain the number of pipe bursts of each pipe segment in the future time period;

[0023] Step S64: summing the number of pipe bursts of each pipe section in the future time period to obtain the total number of pipe bursts in the region in the future time period.

[0024] The present invention also provides a water supply pipe network burst risk prediction system, the system comprising:

[0025] A pipe burst factor set construction module, used to construct a pipe burst factor set;

[0026] An indicator factor set construction module is used to perform principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set by using the principal component analysis method to determine the indicator factor set;

[0027] An indicator weight determination module is used to determine the indicator weight corresponding to each indicator in the indicator factor set by using a hierarchical analysis method, and the indicator with the highest indicator weight is taken as the most important indicator;

[0028] A pipe burst risk prediction model building module is used to establish a pipe burst rate benchmark function based on the most important indicator; take the remaining indicators as risk factors, and establish a pipe burst risk related conditional strength function based on the risk factors; based on the survival analysis theory, establish a pipe burst risk prediction model based on the pipe burst rate benchmark function and the pipe burst risk related conditional strength function;

[0029] A parameter determination module, used to determine the pipe burst risk prediction model parameters;

[0030] The prediction module is used to calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in a future time period through the pipe burst risk prediction model with determined parameters.

[0031] Optionally, the burst factor set includes basic attribute data of the water supply network, water supply network operation data, surrounding environment and facility data, and the number of bursts in the network. The basic attribute data of the water supply network includes: burial depth, pipe diameter, pipe material and pipe age; the water supply network operation data includes: network operation pressure; the surrounding environment and facility data includes: temperature difference, road grade, soil corrosion index, pH value, redox point, drainage conditions, location, pipeline identification, preventive measures and public education.

[0032] Optionally, the indicator factor set specifically includes: burial depth, pipe diameter, pipe pressure, pipe material, temperature difference, road load, pipe age, historical accidents, soil environment and regional attributes.

[0033] Optionally, the parameter determination module specifically includes:

[0034] The unit for calculating the average number of pipe bursts per year is used to group the pipe burst records in year y by the age of the pipe bursts, with a group interval of 10 years, and calculate the average number of pipe bursts per year;

[0035] A pipe network burst rate calculation unit, used to calculate the pipe network burst rate according to the annual average number of pipe bursts;

[0036] The fitting unit is used to perform quadratic parabola regression parameter fitting with the median of the pipe age group and the pipe burst rate of the pipe network as dependent variables to calculate the pipe burst risk prediction model parameters, wherein the pipe burst risk prediction model parameters include the fitting coefficient of the pipe burst rate benchmark function and the regression coefficient corresponding to the dth risk factor.

[0037] Optionally, the prediction module specifically includes:

[0038] A pipe burst rate calculation unit per pipe length is used to calculate various indicators corresponding to each pipe section by principal component analysis, and substitute the indicators into the pipe burst risk prediction model with determined parameters to obtain the pipe burst rate per pipe length;

[0039] A pipe section burst rate calculation unit is used to multiply the pipe burst rate per unit pipe length by the corresponding pipe length of each pipe section to obtain the pipe section burst rate corresponding to each pipe section;

[0040] A pipe burst frequency calculation unit in a future time period is used to integrate the pipe burst rate corresponding to each pipe section with respect to time to obtain the pipe burst frequency of each pipe section in the future time period;

[0041] The regional total pipe burst times calculation unit is used to sum the pipe burst times of each pipe section in the future time period to obtain the regional total pipe burst times in the future time period.

[0042] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0043] The present invention provides a method and system for predicting the risk of pipe burst in a water supply network. The method includes: using the principal component analysis method to perform principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set to determine the index factor set; using the hierarchical analysis method to determine the index weights corresponding to each of the indicators in the index factor set; establishing a pipe burst risk prediction model; determining the pipe burst risk prediction model parameters; and calculating the pipe burst rate of each pipe section and the total number of pipe bursts in the region in the future time period through the pipe burst risk prediction model with the determined parameters. The present invention can solve the problem that the existing prediction methods do not fully consider the factors affecting the pipe burst of the water supply network and do not consider the correlation between the factors while ensuring that data information is not lost as much as possible, thereby improving the accuracy of predicting the risk of pipe burst in the regional water supply network, and thus being able to reasonably avoid the danger caused by the pipe burst of the water supply network. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0045] Figure 1 This is a flow chart of a method for predicting the risk of pipe burst in a water supply network provided in an embodiment of the present invention.

[0046] Figure 2This is a structural diagram of the water supply network burst risk prediction system provided by the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] The purpose of the present invention is to provide a method and system for predicting the risk of pipe burst in a water supply network, so as to improve the accuracy of the prediction results of the risk of pipe burst in a water supply network.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] like Figure 1 As shown, the present invention discloses a method for predicting the risk of pipe burst in a water supply network, the method comprising:

[0051] Step S1: Construct a pipe burst factor set.

[0052] Step S2: using the principal component analysis method, a principal component analysis is performed on the factors affecting the safety of the water supply network in the pipe burst factor set to determine an indicator factor set.

[0053] Step S3: using the hierarchical analysis method to determine the indicator weight corresponding to each indicator in the indicator factor set, and the indicator with the highest indicator weight is taken as the most important indicator.

[0054] Step S4: establishing a pipe burst rate benchmark function according to the most important indicator; taking the remaining indicators as risk factors, establishing a pipe burst risk-related conditional strength function according to the risk factors; and establishing a pipe burst risk prediction model according to the pipe burst rate benchmark function and the pipe burst risk-related conditional strength function based on the survival analysis theory.

[0055] Step S5: Determine pipe burst risk prediction model parameters.

[0056] Step S6: Calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in the future time period through the pipe burst risk prediction model with determined parameters.

[0057] The following is a detailed discussion of each step:

[0058] Step S1: Construct a pipe burst factor set, including:

[0059] Step S11: Collect historical data on pipe bursts in the water supply network.

[0060] Step S12: Analyze the historical data by importance, screen out the factors that affect the safety of the water supply network, and establish a pipe burst factor set. The pipe burst factor set includes basic attribute data of the water supply network, water supply network operation data, surrounding environment and facility data, and the number of pipe bursts in the network. The basic attribute data of the water supply network includes: burial depth, pipe diameter, pipe material and pipe age; the water supply network operation data includes: pipe network operation pressure; the surrounding environment and facility data includes: temperature difference, road grade, soil corrosion index, pH value, redox point, drainage conditions, location, pipeline identification, preventive measures and public education.

[0061] Step S2: using the principal component analysis method, the factors affecting the safety of the water supply network in the pipe burst factor set are analyzed to determine the index factor set, which specifically includes:

[0062] Step S21: The factors affecting the safety of the water supply network in the burst pipe factor set are divided into n samples, each sample has v indicator variables, forming an n×v data matrix. The specific formula is:

[0063]

[0064] Where X represents the n×v data matrix, X v represents the vth indicator variable in the pipe burst factor set, X nv Represents the vth indicator variable in the nth sample.

[0065] Step S22: In each sample, a correlation coefficient matrix is ​​established for v variables. The correlation coefficient matrix is:

[0066]

[0067] Among them, R represents the correlation coefficient matrix, r ij is the i-th variable X i and the jth variable X j The correlation coefficient, r ij =r ji , i, j = 1, 2,…, v.

[0068]

[0069] Among them, n is the number of samples.

[0070] Step S23: Calculate multiple eigenvalues ​​corresponding to the correlation coefficient matrix and eigenvectors corresponding to each eigenvalue, specifically:

[0071] Solve the characteristic equation |λI-R|=0, where I is the identity matrix, and find the eigenvalue λ i , and arranged in order of size:

[0072] λ1≥λ2≥…≥λv≥0;

[0073] Find the i-th eigenvalue λ respectively i The corresponding eigenvector l i (i=1,2,…,v), requiring ||l i ||=1, that is

[0074]

[0075] Among them, l ij represents the i-th eigenvector l i The jth component of .

[0076] Step S24: Calculate the contribution rate corresponding to each eigenvalue and the cumulative contribution rates corresponding to the first p eigenvalues ​​according to the plurality of eigenvalues.

[0077] Step S241: The specific formula for calculating the contribution rate corresponding to each characteristic value is:

[0078]

[0079] Among them, λ i represents the i-th eigenvalue, k represents the variable, and v represents the total number of eigenvalues.

[0080] Step S242: The specific formula for calculating the cumulative contribution rate corresponding to the first p eigenvalues ​​is:

[0081]

[0082] Among them, G p Represents the cumulative contribution rate corresponding to the first p eigenvalues.

[0083] Step S25: Select the index corresponding to each characteristic value whose cumulative contribution rate is within the set range, and determine the index factor set based on the index. The specific formula is:

[0084]

[0085] Among them, z represents the indicator factor set, X v represents the vth indicator variable in the pipe burst factor set, X nv represents the vth indicator variable in the nth sample, l pv represents the pth eigenvector l p The vth component of g It represents the gth indicator in the indicator factor set, which is the factor affecting the safety of the water supply network. 1≤g≤p, p is the total number of indicators.

[0086] In this embodiment, the setting range is 85% to 95%, so the indicator factor set specifically includes:

[0087] Index z1 represents the buried depth, index z2 represents the pipe diameter, index z3 represents the pipe pressure, index z4 represents the pipe material, index z5 represents the temperature difference, index z6 represents the road load, index z7 represents the pipe age, index z8 represents historical accidents, index z9 represents the soil environment, index z1 represents the pipe material, index z1 represents the pipe material, index z2 represents the pipe material, index z3 represents the pipe pressure, index z4 represents the pipe material, index z5 represents the temperature difference, index z6 represents the road load, index z7 represents the pipe age, index z8 represents historical accidents, index z9 represents the soil environment, index z1 represents the pipe material ...2 represents the pipe material, index z3 represents the pipe material, index z4 represents the pipe material, index z5 represents the pipe material, index z6 represents the road load, index z7 represents the pipe age, index z8 represents historical accidents, index z9 represents the soil environment, index z 10 Represents the region attributes.

[0088] Step S3: using the analytic hierarchy process to determine the indicator weights corresponding to the indicators in the indicator factor set, and the indicator with the highest weight is taken as the most important indicator.

[0089] Step S31: Determine the importance of any two indicators in the indicator factor set; that is, fg Indicates the index z f Relative to index z g The importance of index z g Relative to index z f The importance of is 1 / A fg , z f 、z g are the fth and gth indicators in the indicator factor set respectively; p is the total number of indicators.

[0090] Step S32: constructing a judgment matrix according to the importance of any two indicators in the indicator factor set.

[0091] The importance is divided into intervals. If they are in the first set interval, the two indicators are said to have the same importance. If they are in the second set interval, the former is said to be slightly more important than the latter. If they are in the third set interval, the former is said to be obviously more important than the latter. If they are in the fourth set interval, the former is said to be strongly important than the latter. If they are in the fifth set interval, the former is said to be extremely important than the latter. See Table 1 for details.

[0092] Table 1

[0093] Scale value meaning 1 Indicates that the two indicators are equally important. 3 Indicates that compared with the two indicators, the former is slightly more important than the latter 5 Indicates that compared with the two meta-indicators, the former is significantly more important than the latter 7 Indicates that compared with the two indicators, the former is more important than the latter 9 Indicates that compared with the two indicators, the former is extremely important than the latter

[0094] Step S33: Calculating the maximum eigenvalue corresponding to the judgment matrix, specifically including:

[0095] (1) Multiply the indices of the judgment matrix row by row to obtain a first index set U = (U1, U2, ..., U f , …, U p ), the calculation formula is:

[0096]

[0097] Among them, U fis the fth index of the first index set U, f = 1, 2, ..., p, p is the total number of indexes, A fg It represents the scale value of the corresponding position between the f-th indicator and the g-th indicator in the judgment matrix. This scale value can be any value among 1, 3, 5, 7, and 9.

[0098] So: U1=A 11 ×A 12 ×A 13 ×…×A 1p ; U2=A 21 ×A 22 ×A 23 ×…×A 2p ;……;U p =A p1 ×A p2 ×A p3 ×…×A pp .

[0099] (2) Calculate the pth root of each indicator in the first indicator set to obtain the second indicator set. The calculation formula is:

[0100]

[0101] Wherein, w represents the second index set, w=(w1, w2, ..., w f ,…,w p ), w f is the fth index in the second index set, 1≤f≤p.

[0102] so:

[0103] (3) Normalize the second indicator set to obtain a weight set. The specific formula is:

[0104]

[0105] Among them, W is the weight set, W f is the fth index in the weight set, w f is the fth index in the second index set.

[0106] so:

[0107] (4) Calculate the maximum eigenvalue corresponding to the judgment matrix according to the weight set, and the calculation formula is:

[0108]

[0109] Among them, λ max is the maximum eigenvalue corresponding to the judgment matrix A, W fis the fth index in the weight set.

[0110] When all the comparison results are combined, it is inevitable that there will be a certain degree of inconsistency, so we need to perform a consistency check on the judgment matrix to finally determine whether it can be accepted. Therefore, it is necessary to perform a consistency check on the judgment matrix. The specific steps are as follows:

[0111] Step S34: performing a consistency check on the judgment matrix, specifically comprising:

[0112] (1) Calculate the consistency index based on the maximum characteristic root corresponding to the judgment matrix. The specific formula is:

[0113]

[0114] Where: CI represents the consistency index, λ max It represents the maximum eigenvalue corresponding to the judgment matrix A, and p represents the total number of indicators.

[0115] (2) Obtain the average random consistency index RI by looking up the table, as shown in Table 2.

[0116] Table 2

[0117] p 1 2 3 4 5 6 7 8 9 10 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49

[0118] (3) Calculate the consistency ratio based on the consistency index and the average random consistency index RI. The specific formula is:

[0119]

[0120] Among them: CR represents the consistency ratio, CI represents the consistency index, and RI represents the average random consistency index.

[0121] (4) Determine whether the consistency ratio is less than a set value; if the consistency ratio is less than the set value, it means that the consistency of the judgment matrix is ​​acceptable, otherwise return to "step S31" to make appropriate corrections to the judgment matrix. The set value is 0.1.

[0122] Step S35: taking each normalized index in the weight set after the consistency check as the index weight, among which the index weight corresponding to the pipe age is the largest, that is, the most important index is the pipe age.

[0123] Step S4: Establish a pipe burst rate benchmark function based on the most important indicator (pipe age); take the remaining indicators as risk factors, and establish a pipe burst risk related conditional strength function based on the risk factors; based on the survival analysis theory, establish a pipe burst risk prediction model based on the pipe burst rate benchmark function and the pipe burst risk related conditional strength function, specifically including:

[0124] Step S41: The burst frequency of the pipeline generally conforms to the "bathtub curve", that is, the burst rate first gradually decreases from large to small and reaches the mature stage, at which time the burst rate is the lowest, and finally increases greatly with the aging and degradation of the pipeline.

[0125] The burst rate benchmark function represents the change process of pipeline burst risk over time. The burst rate benchmark function is established according to the pipe age. The formula is as follows:

[0126] h0(t)=a+bt+ct 2

[0127] Wherein, h0(t) represents the pipe burst rate benchmark function, times / (year·km), a, b, c represent the fitting coefficients of the pipe burst rate benchmark function, and t represents the pipe age (year).

[0128] Step S42: Establish a pipe burst risk-related condition intensity function based on risk factors. The specific calculation formula is:

[0129]

[0130] Among them, h represents the conditional intensity function related to pipe burst risk, m represents the total number of risk factors, z d represents the d-th risk factor value, β d represents the regression coefficient corresponding to the d-th risk factor.

[0131] Step S43: using the survival analysis theory, a pipe burst risk model is established according to the pipe burst rate benchmark function and the pipe burst risk related condition intensity function. The specific formula is:

[0132]

[0133] Wherein, h(t,z) is the instantaneous burst rate of pipeline, i.e., the probability of a burst accident occurring at time t+△t for a pipeline section that survives to time t, times / (year·km), t is the age of the pipeline (year), m is the number of other indicators in the comprehensive index except the most important comprehensive index, i.e., the number of risk factors, z d is the dth risk factor value, a, b, c are the fitting coefficients of the pipe burst rate benchmark function, β d is the regression coefficient corresponding to the dth risk factor.

[0134] There is a very important parameter in the theory of survival analysis - survival time, which defines the time span from a certain starting event to a certain end event. Most survival analysis models are based on univariate analysis. When doing univariate analysis, the model only describes the relationship between the single variable and survival and ignores the influence of other variables; the Cox proportional hazard model based on the theory of survival analysis solves this problem and can analyze the impact of continuous variables on survival.

[0135] The present invention applies the Cox proportional hazard model to pipe burst risk prediction and constructs a pipe burst risk prediction model. The model takes survival outcome and survival time as dependent variables and can simultaneously analyze the impact of many factors such as pipeline characteristics and environmental factors on the risk of pipe burst.

[0136] Step S5: Determine the pipe burst risk prediction model parameters, specifically including:

[0137] Step S51: group the pipe burst records in year y according to the pipe burst age, with a group interval of 10 years, and calculate the annual average number of pipe bursts. The specific calculation formula is:

[0138]

[0139] Step S52: Calculate the pipe burst rate of the pipe network according to the annual average number of pipe bursts. The pipe burst rate of the pipe network is the annual average number of pipe bursts per unit length. The specific calculation formula is:

[0140]

[0141] Step S53: fitting the parameters of the pipe burst risk prediction model with the median of the pipe age group and the pipe burst rate as the dependent variable, and calculating the pipe burst risk prediction model parameters, wherein the pipe burst risk prediction model parameters include the fitting coefficients a, b and c of the pipe burst rate benchmark function, the regression coefficient β corresponding to the dth risk factor d .

[0142] Step S6: Calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in the future time period through the pipe burst risk prediction model with determined parameters, specifically including:

[0143] Step S61: Calculate various indicators corresponding to each pipe section by principal component analysis, and substitute them into the pipe burst risk prediction model with determined parameters to obtain the pipe burst rate per unit pipe length, in units of times / (year·km).

[0144] Step S62: multiplying the pipe burst rate per unit pipe length by the corresponding pipe length of each pipe section to obtain the pipe burst rate (times / year) corresponding to each pipe section.

[0145] Step S63: Integrate the pipe burst rate corresponding to each pipe segment with respect to time to obtain the number of pipe bursts of each pipe segment in the future time period. The specific calculation formula is:

[0146]

[0147] Among them, N s represents the number of pipe bursts in pipe section s in the future time period, h s Indicates the pipe burst rate per unit length, L srepresents the length of pipe section s, and t represents the calculation time.

[0148] Step S64: summing the number of pipe bursts of each pipe section in the future time period to obtain the total number of pipe bursts in the region in the future time period. The specific calculation formula is:

[0149]

[0150] Where N represents the total number of pipe bursts in the area, S represents the number of pipe sections in the area, and N s Represents the number of pipe bursts in the future for pipe segment s.

[0151] like Figure 2 As shown, the present invention also provides a water supply pipe network burst risk prediction system, the system comprising:

[0152] The pipe burst factor set construction module 1 is used to construct the pipe burst factor set.

[0153] The indicator factor set construction module 2 is used to use the principal component analysis method to perform principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set to determine the indicator factor set.

[0154] The indicator weight determination module 3 is used to determine the indicator weight corresponding to each indicator in the indicator factor set by using the hierarchical analysis method, and the indicator with the highest indicator weight is taken as the most important indicator.

[0155] The pipe burst risk prediction model construction module 4 is used to establish a pipe burst rate benchmark function based on the most important indicator; use the remaining indicators as risk factors and establish a pipe burst risk related conditional strength function based on the risk factors; based on the survival analysis theory, establish a pipe burst risk prediction model based on the pipe burst rate benchmark function and the pipe burst risk related conditional strength function.

[0156] The parameter determination module 5 is used to determine the pipe burst risk prediction model parameters.

[0157] The prediction module 6 is used to calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in a future time period through the pipe burst risk prediction model with determined parameters.

[0158] As an optional implementation, the pipe burst factor set of the present invention includes basic attribute data of the water supply network, water supply network operation data, surrounding environment and facility data, and the number of pipe bursts in the network. The basic attribute data of the water supply network includes: burial depth, pipe diameter, pipe material and pipe age; the water supply network operation data includes: pipe network operation pressure; the surrounding environment and facility data includes: temperature difference, road grade, soil corrosion index, pH value, redox point, drainage conditions, location, pipeline identification, preventive measures and public education.

[0159] As an optional implementation, the indicator factor set of the present invention specifically includes: burial depth, pipe diameter, pipe pressure, pipe material, temperature difference, road load, pipe age, historical accidents, soil environment and regional attributes.

[0160] As an optional implementation manner, the parameter determination module 5 of the present invention specifically includes:

[0161] The unit for calculating the average number of pipe bursts per year is used to group the pipe burst records in year y by the age of the pipe bursts, with a group interval of 10 years, and calculate the average number of pipe bursts per year.

[0162] The pipe network burst rate calculation unit is used to calculate the pipe network burst rate according to the annual average number of pipe bursts.

[0163] The fitting unit is used to perform quadratic parabola regression parameter fitting with the median of the pipe age group and the pipe burst rate of the pipe network as dependent variables to calculate the pipe burst risk prediction model parameters, wherein the pipe burst risk prediction model parameters include the fitting coefficient of the pipe burst rate benchmark function and the regression coefficient corresponding to the dth risk factor.

[0164] As an optional implementation, the prediction module 6 of the present invention specifically includes:

[0165] The pipe burst rate calculation unit per unit pipe length is used to calculate various indicators corresponding to each pipe section by principal component analysis method, and substitute the indicators into the pipe burst risk prediction model with determined parameters to obtain the pipe burst rate per unit pipe length.

[0166] The pipe section burst rate calculation unit is used to multiply the pipe burst rate per unit pipe length by the corresponding pipe length of each pipe section to obtain the pipe section burst rate corresponding to each pipe section.

[0167] The pipe burst frequency calculation unit in the future time period is used to integrate the pipe burst rate corresponding to each pipe section with respect to time to obtain the pipe burst frequency of each pipe section in the future time period.

[0168] The regional total pipe burst times calculation unit is used to sum the pipe burst times of each pipe section in the future time period to obtain the regional total pipe burst times in the future time period.

[0169] The present invention solves the problem that the existing prediction methods do not fully consider the factors affecting pipe bursts in water supply networks and do not consider the correlation between the factors, thereby improving the accuracy of the prediction results. The prediction method provided by the present invention can make full use of the historical data of pipe bursts in urban water supply networks, rationally screen the factors affecting pipe bursts, and the method is simple and versatile, which is conducive to water supply companies taking targeted measures according to the predicted pipe burst risks, effectively preventing and eliminating pipe burst accidents, and reducing the impact of pipe burst accidents on urban production and residents' lives.

[0170] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0171] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for predicting the risk of pipe burst in a water supply network, characterized in that: The method comprises: Step S1: construct a pipe burst factor set; The pipe burst factor set includes basic attribute data of the water supply network, water supply network operation data, surrounding environment and facility data, and the number of pipe bursts in the network. The basic attribute data of the water supply network includes: burial depth, pipe diameter, pipe material, and pipe age; the water supply network operation data includes: pipe network operation pressure; the surrounding environment and facility data includes: temperature difference, road grade, soil corrosion index, pH value, redox point, drainage conditions, location, pipeline identification, preventive measures, and public education; Step S2: using the principal component analysis method, performing principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set to determine an indicator factor set; The index factor set specifically includes: burial depth, pipe diameter, pipe pressure, pipe material, temperature difference, road load, pipe age, historical accidents, soil environment and regional attributes; The principal component analysis method is used to perform principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set to determine the index factor set, which specifically includes: Step S21: The factors affecting the safety of the water supply network in the burst pipe factor set are divided into n samples, each sample has v indicator variables, forming an n×v data matrix. The specific formula is: Where X represents the n×v data matrix, X v represents the vth indicator variable in the pipe burst factor set, X nv represents the vth indicator variable in the nth sample; Step S22: In each sample, a correlation coefficient matrix is ​​established for v variables. The correlation coefficient matrix is: Among them, R represents the correlation coefficient matrix, r ij is the i-th variable X i and the jth variable X j The correlation coefficient, r ij =r ji , i, j = 1, 2, ..., v; Where n is the number of samples; Step S23: Calculate multiple eigenvalues ​​corresponding to the correlation coefficient matrix and eigenvectors corresponding to each eigenvalue, specifically: Solve the characteristic equation |λI-R|=0, where I is the identity matrix, and find the eigenvalue λ i , and arranged in order of size: λ1≥λ2≥…≥λ V ≥0; Find the i-th eigenvalue λ respectively i The corresponding eigenvector l i (i=1,2,…,v), requiring ||l i ||=1, that is Among them, l ij represents the i-th eigenvector l i The jth component of Step S24: calculating the contribution rate corresponding to each eigenvalue and the cumulative contribution rate corresponding to the first p eigenvalues ​​according to the plurality of eigenvalues; Step S241: The specific formula for calculating the contribution rate corresponding to each characteristic value is: Among them, λ i represents the i-th eigenvalue, k represents the variable, and v represents the total number of eigenvalues; Step S242: The specific formula for calculating the cumulative contribution rate corresponding to the first p eigenvalues ​​is: Among them, G p represents the cumulative contribution rate corresponding to the first p eigenvalues; Step S25: Select the index corresponding to each characteristic value whose cumulative contribution rate is within the set range, and determine the index factor set based on the index. The specific formula is: Among them, z represents the indicator factor set, X v represents the vth indicator variable in the pipe burst factor set, X nv represents the vth indicator variable in the nth sample, l pv represents the pth eigenvector l p The vth component of g It represents the gth indicator in the indicator factor set, which is the factor affecting the safety of the water supply network, 1≤g≤p, and p is the total number of indicators; Step S3: using the hierarchical analysis method to determine the indicator weight corresponding to each indicator in the indicator factor set, and the indicator with the highest indicator weight is taken as the most important indicator; Step S31: Determine the importance of any two indicators in the indicator factor set; that is, fg Indicates the index z f Relative to index z g The importance of index z g Relative to index z f The importance of is 1 / A fg , z f 、z g are the fth and gth indicators of the indicator factor set respectively; p is the total number of indicators; Step S32: constructing a judgment matrix according to the importance of any two indicators in the indicator factor set; The importance is divided into intervals. If it is in the first set interval, the two indicators are said to be of equal importance. If it is in the second set interval, the former is said to be slightly more important than the latter. If it is in the third set interval, the former is said to be significantly more important than the latter. If it is in the fourth set interval, the former is said to be strongly more important than the latter. If it is in the fifth set interval, the former is said to be extremely more important than the latter. Step S33: Calculating the maximum eigenvalue corresponding to the judgment matrix, specifically including: (1) Multiply the indices of the judgment matrix row by row to obtain a first index set U = (U1, U2, ..., U f ,…,U p ), the calculation formula is: Among them, U f is the fth index of the first index set U, f = 1, 2, ..., p, p is the total number of indexes, A fg It represents the scale value of the corresponding position between the fth index and the gth index in the judgment matrix. This scale value can be any value among 1, 3, 5, 7, and 9. Reason: U1=A 11 ×A 12 ×A 13 ×…×A 1p ;U2=A 21 ×A 22 ×A 23 ×…×A 2p ;……;U p = A p1 ×A p2 ×A p3 ×…×A pp ; (2) Calculate the pth root of each indicator in the first indicator set to obtain the second indicator set. The calculation formula is: Wherein, w represents the second index set, w=(w1, w2, ..., w f ,…,w p ), w f is the fth index in the second index set, 1≤f≤p; so: (3) Normalize the second indicator set to obtain a weight set. The specific formula is: Among them, W is the weight set, W f is the fth index in the weight set, w f is the fth index in the second index set; so: (4) Calculate the maximum eigenvalue corresponding to the judgment matrix according to the weight set, and the calculation formula is: Among them, λ max is the maximum eigenvalue corresponding to the judgment matrix A, W f is the fth index in the weight set; When all comparison results are synthesized, it is necessary to perform consistency check on the judgment matrix. The specific steps are as follows: Step S34: performing a consistency check on the judgment matrix, specifically comprising: (1) Calculate the consistency index based on the maximum characteristic root corresponding to the judgment matrix. The specific formula is: Where: CI represents the consistency index, λ max represents the maximum eigenvalue corresponding to the judgment matrix A, and p represents the total number of indicators; (2) Obtain the average random consistency index RI by looking up the table; (3) Calculate the consistency ratio based on the consistency index and the average random consistency index RI. The specific formula is: Among them: CR represents the consistency ratio, CI represents the consistency index, and RI represents the average random consistency index; (4) Determine whether the consistency ratio is less than a set value; if the consistency ratio is less than the set value, it means that the consistency of the judgment matrix is ​​acceptable, otherwise return to "step S31" to make appropriate corrections to the judgment matrix; the set value is 0.1; Step S35: taking each normalized index in the weight set that has passed the consistency test as the index weight, wherein the index weight corresponding to the pipe age is the largest; Step S4: establishing a pipe burst rate benchmark function according to the most important indicator; taking the remaining indicators as risk factors, establishing a pipe burst risk related conditional strength function according to the risk factors; and establishing a pipe burst risk prediction model according to the pipe burst rate benchmark function and the pipe burst risk related conditional strength function based on the survival analysis theory; Step S5: determining pipe burst risk prediction model parameters; The step of determining the pipe burst risk prediction model parameters specifically includes: Step S51: group the pipe burst records in year y according to the pipe burst age, with a group interval of 10 years, and calculate the annual average number of pipe bursts; Step S52: calculating the pipe burst rate of the pipe network according to the annual average number of pipe bursts; Step S53: fitting quadratic parabola regression parameters with the median of the pipe age group and the pipe burst rate of the pipe network as dependent variables to calculate pipe burst risk prediction model parameters, wherein the pipe burst risk prediction model parameters include the fitting coefficient of the pipe burst rate benchmark function and the regression coefficient corresponding to the dth risk factor; Step S6: Calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in the future time period through the pipe burst risk prediction model with determined parameters; The pipe burst risk prediction model with determined parameters is used to calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in the future time period, specifically including: Step S61: Calculate various indicators corresponding to each pipe section by principal component analysis method, and substitute them into the pipe burst risk prediction model with determined parameters to obtain the pipe burst rate per unit pipe length; Step S62: multiplying the pipe burst rate per unit pipe length by the pipe length corresponding to each pipe section to obtain the pipe burst rate corresponding to each pipe section; Step S63: Integrate the pipe burst rate corresponding to each pipe segment with respect to time to obtain the number of pipe bursts of each pipe segment in the future time period; Step S64: summing the number of pipe bursts of each pipe section in the future time period to obtain the total number of pipe bursts in the region in the future time period.

2. A water supply network burst risk prediction system, using the water supply network burst risk prediction method according to claim 1, characterized in that: The system comprises: A pipe burst factor set construction module, used to construct a pipe burst factor set; An indicator factor set construction module is used to perform principal component analysis on the factors affecting the safety of the water supply network in the pipe burst factor set by using the principal component analysis method to determine the indicator factor set; An indicator weight determination module is used to determine the indicator weight corresponding to each indicator in the indicator factor set by using a hierarchical analysis method, and the indicator with the highest indicator weight is taken as the most important indicator; A pipe burst risk prediction model building module is used to establish a pipe burst rate benchmark function based on the most important indicator; take the remaining indicators as risk factors, and establish a pipe burst risk related conditional strength function based on the risk factors; based on the survival analysis theory, establish a pipe burst risk prediction model based on the pipe burst rate benchmark function and the pipe burst risk related conditional strength function; A parameter determination module, used to determine the pipe burst risk prediction model parameters; A prediction module, used to calculate the pipe burst rate of each pipe section and the total number of pipe bursts in the region in a future time period through the pipe burst risk prediction model with determined parameters; The prediction module specifically includes: A pipe burst rate calculation unit per pipe length is used to calculate various indicators corresponding to each pipe section by principal component analysis, and substitute the indicators into the pipe burst risk prediction model with determined parameters to obtain the pipe burst rate per pipe length; A pipe section burst rate calculation unit is used to multiply the pipe burst rate per unit pipe length by the corresponding pipe length of each pipe section to obtain the pipe section burst rate corresponding to each pipe section; A pipe burst frequency calculation unit in a future time period is used to integrate the pipe burst rate corresponding to each pipe section with respect to time to obtain the pipe burst frequency of each pipe section in the future time period; The regional total pipe burst times calculation unit is used to sum the pipe burst times of each pipe section in the future time period to obtain the regional total pipe burst times in the future time period.

Images