CN110543719A - A method for predicting leakage of water supply pipeline based on spatial metering model - Google Patents

Abstract

the invention provides a water supply pipeline leakage prediction method based on a spatial metering model, and belongs to the technical field of water supply pipeline leakage. The method determines an index system influencing leakage of the water supply pipeline from pipeline properties, environmental factors and operation factors. Based on the established index system, EPANET software is utilized to simulate the index value of a water supply pipeline in a certain city in a certain day, the pipeline data of the water supply area A with the most serious leakage condition in the city are obtained, the pipeline data are divided into different spaces according to the serial number of the pipeline, the spatial correlation analysis is carried out on the water supply pipeline, a spatial weight matrix is defined, a spatial metering model is determined, a water supply pipeline leakage evaluation system is established, and the leakage condition of the pipeline is predicted.

Description

A method for predicting leakage of water supply pipeline based on spatial metering model

technical field

The invention relates to a method for predicting leakage of water supply pipelines based on a spatial metering model, and belongs to the technical field of leakage of water supply pipelines.

Background technique

With the development of society and the continuous advancement of urbanization, the leakage of water supply pipes is becoming one of the important problems faced by the water supply system. For a long time, the leakage rate of pipelines in my country's urban water supply system has remained high, which not only hinders the operation and development of water supply enterprises, but also seriously affects the efficiency of urban water supply. Therefore, how to explore the factors that cause the leakage of water supply pipelines, how to propose effective methods and countermeasures to reduce pipeline leakage, how to accurately predict pipeline leakage, and how to accurately locate the leakage points of water supply pipelines to reduce pipeline leakage The occurrence of accidents is an urgent problem that needs to be solved at present.

SUMMARY OF THE INVENTION

In order to reduce the occurrence of pipeline leakage accidents, the present invention proposes a method for predicting leakage of water supply pipelines based on a spatial metering model, and the adopted technical solutions are as follows:

A method for predicting leakage of water supply pipelines based on a spatial metering model, the method for predicting leakage of water supply pipelines includes:

The first step is to obtain prediction objects, spaces, data and factors through EPANET software combined with dynamic hydraulic model to analyze its influence on pipeline leakage;

The second step is to determine the spatial correlation between the pipelines through spatial correlation analysis;

The third step is to use the jplv toolbox in the MATLAB software to perform LM inspection on the pipeline leakage data of the water supply area A to determine the spatial measurement model;

The fourth step is to use the model determined in the third step to calculate the estimated coefficients of each parameter, and then use the estimated coefficients of each parameter to explain each factor through the direct effect and indirect effect, and obtain the estimated results of the direct effect and indirect effect; The estimation results of direct effect and indirect effect analyze the influence of various factors on pipeline leakage.

Further, the process of acquiring prediction objects, spaces, data and factors in the first step includes:

Step 1. Using the EPANET software, use the dynamic hydraulic model of the water supply pipeline to simulate the data of 24 time periods in a day with one hour as a time node, and obtain the pipeline data of the water supply area A with the most serious leakage in the city as the prediction object ;

Step 2. Divide into multiple spaces according to the difference of the pipeline, and each pipeline is used as a space to complete the establishment, analysis and prediction of the model;

Step 3. Model the data of the first 20 time periods in the 24 time periods described in step 1 to explore the influencing factors of the leakage rate of the water supply pipeline. The data of the last 4 time periods are used for prediction to verify the reliability of the model. sex;

Step 4: Select 7 factors of pipe diameter, pipe age, pipe material, pipe length, thickness of covering soil, pipeline operating pressure and pipeline water flow velocity as explanatory variables to analyze the influence degree of the 7 factors on pipeline leakage.

Further, the specific number of the spaces described in step 2 is 39.

Further, the process of determining the spatial correlation between the pipelines through the spatial correlation analysis described in the second step includes:

Step 1. Use a method based on a combination of geographic location and leakage rate to establish a spatial weight matrix, where the spatial weight matrix is:

in, and are the average leakage rates of pipeline i and pipeline j in the first 20 time periods;

Step 2. Use GeoDa software to perform spatial autocorrelation analysis, and use the average data of each index data of water supply area A in the first 20 time periods to obtain the Moran's I index value and Moran scatter plot of the leakage of each pipeline in this water supply area;

Step 3. Obtain the spatial correlation between the pipelines according to Moran's I index value and Moran's scatter plot.

Further, the process of determining the spatial econometric model by the LM test described in the third step includes:

Step a: Use the OLS regression method to obtain the random effect model of the OLS regression;

Step b: Two LM (Lagrange multiplier test) to judge LM-error and LM-lag;

Step c: If the judgment result is that both LM-error and LM-lag tests are not significant, keep the OLS result; if the judgment result is that both test results are significant, perform step d; if the judgment result is that there is only one significant , then, when it is judged that the LM-error is significant, the SEM model (Spatial Error Model, SEM) is used; when the LM-lag is judged to be significant, the SAR model (spatial autoregressive model, Spatial Auto Regressive model, SEM) is used. SAR);

Step d: further test RLM-lag and RLM-error, and judge the test results of RLM-lag and RLM-error; if the judgment result is that both test results are not significant, select the OLS regression model; if the judgment result is RLM If the -error is significant, the SEM model (Spatial Error Model, SEM) is selected; if the judgment result is that the RLM-lag is significant, the SAR model (Spatial Auto Regressive model, SAR) is selected.

Among them, LM-error, LM-lag, RLM-lag and RLM-error respectively represent four different statistics, namely variable spatial correlation test (LM-lag), error term spatial correlation test (LM-error), Robust variable spatial correlation test (RLM-lag), robust error term spatial correlation test (RLM-error).

Beneficial effects of the present invention:

Based on the spatial measurement model, the present invention proposes a method for predicting the leakage of urban water supply pipe network sections, explores the spatial distribution of pipeline leakage in the water supply area A, establishes a spatial autoregressive model based on the physical topological connection of pipe sections, and analyzes the ability to realize urban water supply. The importance ordering and prediction of leakage at the section level of the pipe network provide a new strategy and technical means for the reconstruction of the urban water supply pipe network.

Description of drawings

Fig. 1 is the LM inspection flow chart of the present invention;

Fig. 2 is the Moran scatter diagram of the water supply area A of the present invention;

FIG. 3 is a comparison diagram of the predicted value and the actual value of the method according to the present invention from 20:00 to 21:00.

Detailed ways

The present invention will be further described below in conjunction with specific embodiments, but the present invention is not limited by the embodiments.

Example 1:

A method for predicting leakage of water supply pipelines based on a spatial metering model, the method for predicting leakage of water supply pipelines includes:

The first step is to obtain prediction objects, spaces, data and factors through EPANET software combined with dynamic hydraulic model to analyze its influence on pipeline leakage;

The second step is to determine the spatial correlation between the pipelines through spatial correlation analysis;

The third step is to use the jplv toolbox in the MATLAB software to perform LM inspection on the pipeline leakage data of the water supply area A to determine the spatial measurement model;

The fourth step is to use the model determined in the third step to calculate the estimated coefficients of each parameter, and then use the estimated coefficients of each parameter to explain each factor through the direct effect and indirect effect, and obtain the estimated results of the direct effect and indirect effect; The estimation results of direct effect and indirect effect analyze the influence of various factors on pipeline leakage.

Wherein, the process of obtaining prediction objects, spaces, data and factors in the first step includes:

Step 1. Using the EPANET software, use the dynamic hydraulic model of the water supply pipeline to simulate the data of 24 time periods in a day with one hour as a time node, and obtain the pipeline data of the water supply area A with the most serious leakage in the city as the prediction object ;

Step 2: Divide into multiple spaces according to different pipelines, and each pipeline is used as a space to complete the establishment, analysis and prediction of the model; wherein, the specific number of the spaces is 39;

Step 3. Model the data of the first 20 time periods in the 24 time periods described in step 1 to explore the influencing factors of the leakage rate of the water supply pipeline. The data of the last 4 time periods are used for prediction to verify the reliability of the model. sex;

Step 4: Select 7 factors of pipe diameter, pipe age, pipe material, pipe length, thickness of covering soil, pipeline operating pressure and pipeline water flow velocity as explanatory variables to analyze the influence degree of the 7 factors on pipeline leakage.

The data analyzed in this example comes from the water supply pipeline in M city in northern my country. The city has built a dynamic hydraulic model for the city's water supply pipeline, which enables real-time simulation of pipeline conditions. Therefore, through the EPANET software, using the dynamic hydraulic model of the water supply pipeline, taking one hour as a time node, simulating the data of 24 time periods in a day, and obtaining the pipeline data of the water supply area A with the most serious leakage in the city as the research object. It is divided into 39 spaces according to the different pipelines, and each pipeline is used as a space to complete the establishment, analysis and prediction of the model. In this example, the data of the first 20 time periods are used for modeling, to explore the influencing factors of the leakage rate of the water supply pipeline, and the data of the last 4 time periods are used to predict and verify the reliability of the model.

In this example, seven factors, such as pipe diameter, pipe age, pipe material, pipe length, thickness of covering soil, pipeline operating pressure and pipeline water flow velocity, are selected as explanatory variables, and the degree of influence of these factors on pipeline leakage is analyzed. Table 1 is a description of the selected relevant indicators.

Table 1 Pipeline leakage related indicators

Since the variables include continuous variables and discrete variables, for the convenience of research, the variables need to be preprocessed first. For continuous variables, logarithmization is required in order to eliminate the bias of the model estimation results caused by different dimensions; for discrete variables, it needs to be converted into 0-1 variables. That is to take the logarithm of the variables such as leakage rate, pipe diameter, pipe age, pipe length, covering soil thickness, pipe pressure and pipe water flow velocity, convert the variable pipe material into a 0-1 variable, and split it into 4 variables, respectively. Whether it is an ordinary cast iron pipe (isA), whether it is a steel pipe (isB), whether it is a PVC pipe (isC) and whether it is a ductile iron pipe (isD), some data after processing are shown in Table 2, and the meaning of each letter in the table is shown in Table 1 shown.

Table 2 Part of the data after preprocessing

The process of determining the spatial correlation between the pipelines through the spatial correlation analysis in the second step includes:

Step 1. Use a method based on a combination of geographic location and leakage rate to establish a spatial weight matrix, where the spatial weight matrix is:

in, and are the average leakage rates of pipeline i and pipeline j in the first 20 time periods;

Step 2. Use GeoDa software to perform spatial autocorrelation analysis, and use the average data of each index data of water supply area A in the first 20 time periods to obtain the Moran's I index value and Moran scatter diagram of the leakage of each pipeline in the water supply area, as shown in the figure. As shown in Figure 2;

Step 3. Obtain the spatial correlation between the pipelines according to Moran's I index value and Moran's scatter plot.

As shown in Figure 2, the Moran's I index value is 0.468, and the corresponding p value is 0.001. When the significance is 0.01, it can be considered that there is a significant positive spatial autocorrelation between pipeline leakage; Moran scatter plot The abscissa is the explained variable, and the ordinate is the product of the explained variable and the spatial weight matrix, which is often used to study spatial characteristics. From the Moran scatter plot shown in Figure 2, we can see the distribution of each pipeline. The pipelines are basically distributed in the first and third quadrants, indicating that there is a significant spatial correlation between the pipelines, and it has a positive impact. That is, there is the same variation in the leakage rate between adjacent pipes. It is further explained that the pipeline adjacent to the pipeline with a high leakage rate also has a higher leakage rate, and similarly, the leakage rate of the pipeline adjacent to the pipeline with a low leakage rate will not be high.

As shown in Figure 1, the process of determining the spatial econometric model by the LM test described in the third step includes:

Step a: Obtain the random effect model of OLS regression by using the OLS regression method; wherein, the specific form of the random effect model is as follows:

Among them, R _it is the explained variable, namely the leakage rate, ρ is the spatial autoregressive coefficient, W is the spatial weight matrix, D _it ,age _it ,…,V _it is the explanatory variable, β ₁ ,β ₂ ,…β ₁₁ The coefficient is estimated for the explanatory variable, reflecting the degree of influence of the explanatory variable on the explained variable, α is the random effect model term, and ε _it is the random error vector.

Step b: Two LM (Lagrange multiplier test) to judge LM-error and LM-lag;

Step c: If the judgment result is that both LM-error and LM-lag tests are not significant, keep the OLS result; if the judgment result is that both test results are significant, perform step d; if the judgment result is that there is only one significant , then, when it is judged that the LM-error is significant, the SEM model (Spatial Error Model, SEM) is used; when the LM-lag is judged to be significant, the SAR model (spatial autoregressive model, Spatial Auto Regressive model, SEM) is used. SAR);

Step d: further test RLM-lag and RLM-error, and judge the test results of RLM-lag and RLM-error; if the judgment result is that both test results are not significant, select the OLS regression model; if the judgment result is RLM If the -error is significant, the SEM model (Spatial Error Model, SEM) is selected; if the judgment result is that the RLM-lag is significant, the SAR model (Spatial Auto Regressive model, SAR) is selected.

Among them, LM-error, LM-lag, RLM-lag and RLM-error respectively represent four different statistics, namely variable spatial correlation test (LM-lag), error term spatial correlation test (LM-error), Robust variable spatial correlation test (RLM-lag), robust error term spatial correlation test (RLM-error).

The specific process of the interpretation analysis described in the fourth step is as follows:

According to the initial setting of the model shown in formula (1), using MATLAB software, the estimated coefficients of each parameter are calculated, and the estimated values of each coefficient are shown in the second column of Table 3. For the spatial panel autoregressive model, due to the existence of spatial effects, the estimated coefficients of each parameter of the model cannot directly represent the degree of influence of each factor, and each factor needs to be explained through direct and indirect effects. The estimation results of direct effect and indirect effect are shown in Table 3. By estimating direct effect and indirect effect, the influence of various factors on pipeline leakage can be analyzed.

(1) Spatial factor: In the estimation results of the spatial autoregressive model, the spatial lag coefficient ρ=0.408, and it is significant at the level of 0.01, indicating that the leakage of the pipeline is largely affected by the leakage of its adjacent pipelines. As the leakage of the adjacent pipeline increases, the leakage of the own pipeline will also become more serious.

(2) Pipe properties: First, at a significant level of 0.1, the direct effects of three 0-1 variables of pipe diameter and pipe material (whether ordinary cast iron pipe, whether PVC pipe and whether ductile iron pipe are not) passed the test, and All three coefficients of 0-1 variables for the pipe are positive. Among the four 0-1 variables of the pipe, the largest coefficient is "whether it is a ductile iron pipe", the value is 0.2636, and the smallest coefficient is "whether it is a steel pipe", the value is 0.0369, the difference between the two is 0.2267, indicating that if the pipe is made from ductile iron, the value is 0.2267. If the cast iron pipe is replaced by a steel pipe, the leakage rate can be reduced by 0.2267 units; the pipe diameter has passed the significance test at the 0.1 level, and its regression coefficient is negative, that is, as the pipe diameter increases, the leakage loss will decrease; The length did not pass the significance test, that is, the effect of the pipe length on the leakage rate of the pipeline was not significant; the pipe age did not pass the significance test, probably because the data used in this chapter for the study, the pipe age data distribution is uneven, basically all 49 years, there are only a few 7 years and 29 years, so that the pipe age is not significant for pipeline leakage. In the estimation results of indirect effects, only the pipe diameter passed the significance test of 0.1, indicating that the size of the pipe diameter will affect the leakage of the adjacent pipes, and other factors in the pipe properties will not affect the adjacent pipes.

(3) Environmental factors: the coefficient of the thickness of the covering soil is negative, which is significant at the level of 0.05. The greater the thickness of the covering soil, the less the pipeline is affected by the ground load, and the smaller the leakage of the pipeline; and its direct and indirect effects are significant. , indicating that the thickness of the covering soil will not only affect the leakage of its own pipeline, but also affect the leakage of the adjacent pipeline. The leakage of the pipeline will be affected by the ground bearing of the own pipeline and the adjacent pipeline at the same time.

(4) Operational factors: both the pipeline operating pressure and the pipeline water flow velocity have passed the significance test at the 0.05 level, and the regression coefficients are 0.346 and 0.182, respectively, and the direct and indirect effects are both significant, indicating that with the pipeline pressure and pipeline water flow As the speed increases, the leakage of the pipeline will increase; excessive pressure and flow velocity will not only cause leakage of the own pipeline, but also pose a threat to the adjacent pipeline.

According to the determined equation (1), the predicted value of the leakage rate of each pipeline in the water supply area A is calculated and checked by using the pipe diameter, pipe material, pipe operating pressure, pipe water flow velocity and covering soil thickness values in the following four time periods. predict the effect. The following table only lists the prediction results for the time period from 20:00 to 21:00. The relative error of the prediction is in the range of [0.0051, 0.0364], and the prediction effect is better.

According to the prediction results of pipeline leakage in Table 4, it can be found that the leakage of each pipeline in water supply area A, the most serious leakage is the G01525 pipeline. trunk. The least leakage is the G01428 pipeline, which is in the branch line and has a lower water supply pressure. Figure 3 is a line graph of the predicted value and the actual value in this time period. It can be seen that the two broken lines almost overlap, and the fitting effect is good.

Table 3 Regression coefficient, direct effect and indirect effect estimation results

Note: *, ** and *** in the table represent significant levels at 10%, 5% and 1%, respectively

Table 4 Prediction of pipeline leakage

Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Anyone who is familiar with this technology can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, The protection scope of the present invention should be defined by the claims.

Claims (5)

1. a method for predicting leakage of water supply pipelines based on a spatial metering model, wherein the method for predicting leakage of water supply pipelines comprises: The first step is to obtain prediction objects, spaces, data and factors through EPANET software combined with dynamic hydraulic model to analyze its influence on pipeline leakage; The second step is to determine the spatial correlation between the pipelines through spatial correlation analysis; The third step is to perform LM test on the leakage data of the pipeline in the water supply area A to determine the spatial measurement model; The fourth step is to use the model determined in the third step to calculate the estimated coefficients of each parameter, and then use the estimated coefficients of each parameter to explain each factor through the direct effect and indirect effect, and obtain the estimated results of the direct effect and indirect effect; The estimation results of direct effect and indirect effect analyze the influence of various factors on pipeline leakage. 2 . The method for predicting leakage of water supply pipelines according to claim 1 , wherein the process of obtaining the predicted objects, spaces, data and factors in the first step comprises: 2 . Step 1. Using the EPANET software, use the dynamic hydraulic model of the water supply pipeline to simulate the data of 24 time periods in a day with one hour as a time node, and obtain the pipeline data of the water supply area A with the most serious leakage in the city as the prediction object ; Step 2. Divide into multiple spaces according to the difference of the pipeline, and each pipeline is used as a space to complete the establishment, analysis and prediction of the model; Step 3. Model the data of the first 20 time periods in the 24 time periods described in step 1 to explore the influencing factors of the leakage rate of the water supply pipeline. The data of the last 4 time periods are used for prediction to verify the reliability of the model. sex; Step 4: Select 7 factors of pipe diameter, pipe age, pipe material, pipe length, thickness of covering soil, pipeline operating pressure and pipeline water flow velocity as explanatory variables to analyze the influence degree of the 7 factors on pipeline leakage. 3 . The method for predicting leakage of water supply pipelines according to claim 2 , wherein the specific number of the spaces in step 2 is 39. 4 . 4. The method for predicting leakage of water supply pipelines according to claim 2, wherein the process of determining the spatial correlation existing between pipelines through spatial correlation analysis in the second step comprises: Step 1. Use a method based on a combination of geographic location and leakage rate to establish a spatial weight matrix, where the spatial weight matrix is: in, and are the average leakage rates of pipeline i and pipeline j in the first 20 time periods; Step 2. Use GeoDa software to perform spatial autocorrelation analysis, and use the average data of each index data of water supply area A in the first 20 time periods to obtain the Moran's I index value and Moran scatter plot of the leakage of each pipeline in this water supply area; Step 3. Obtain the spatial correlation between the pipelines according to Moran's I index value and Moran's scatter plot. 5. The method for predicting water supply pipeline leakage according to claim 1, wherein the process of determining the spatial measurement model by the LM test in the third step comprises: Step a: Use the OLS regression method to obtain the random effect model of the OLS regression; Step b: Two LMs judge LM-error and LM-lag; Step c: If the judgment result is that both LM-error and LM-lag tests are not significant, keep the OLS result; if the judgment result is that both test results are significant, perform step d; if the judgment result is that there is only one significant , then, when it is judged that the LM-error is significant, the SEM model is used; when it is judged that the LM-lag is significant, the SAR model is used; Step d: further test RLM-lag and RLM-error, and judge the test results of RLM-lag and RLM-error; if the judgment result is that both test results are not significant, select the OLS regression model; if the judgment result is RLM If the -error is significant, the SEM model is selected; if the judgment result is that the RLM-lag is significant, the SAR model is selected.