KR20110086529A - Management system for water distribution network employing logistic regression - Google Patents

Abstract

PURPOSE: An optimal management system for a water distribution system using a logistic regression analysis is provided to utilize methods as information for managing water distribution system through a score evaluation method, a fuzzy technique, an ELECTRE(Elimination Et Choice Translating Reality) technique, logistic regression analysis and a Bayes method. CONSTITUTION: Pipe damage and water leakage record are inputted. Data is processed by using a GIS. Pipe state indexes and surrounding environment indexes are obtained. A logistic regression analysis is processed. Model fitness is analyzed and a model coefficient is determined. A pipe damage model is determined. A pipe damage classification standard is determined. A flow monitoring system monitors a block flow rate/pressure.

Description

MANAGEMENT SYSTEM FOR WATER DISTRIBUTION NETWORK EMPLOYING LOGISTIC REGRESSION}

The present invention relates to a method and system for analyzing and assessing the soundness of a water supply network for improving the flow rate, and a block flow monitoring system using the same.

Water pipes underground or above the ground are divided into water pipes and sewer pipes. Here, the water pipe is a passage of purified and disinfected high water supply, and a large pressure must be applied to supply the water to a remotely located consumer. Therefore, the water pipe must have sufficient rigidity to support the pressure.

For this purpose, the water pipe is made of steel and the diameter is made as large as possible to minimize the amount of water pressure per unit area of the water pipe. Therefore, the water pressure received by the water pipe can be reduced, and the material of the water pipe itself performs the task as a water pipe without major problems.

By the way, since such a water pipe cannot be integrated from a water source to a consumer, a plurality of water pipes are piped afterwards. Thus, a joint is formed between the water pipe and the water pipe. However, a gap may be generated in such a joint and a constant flowing at high water pressure may leak through the gap.

Of course, such leakage is compensated to some extent by packing with synthetic resin covering the boundary of the joint where the gap occurs or by processing the shape so that the connection between the water pipes is close to each other, but it does not completely block the leakage, As the joint between the water pipe and the water pipe opens, the gap becomes large, so that the leakage caused by the gap formed in the joint does not completely be eliminated.

On the other hand, if a leak occurs in the joint of the underground water pipe, it is not easy to recognize. In other words, if a leak occurs in the underground water pipe, it is not possible to quickly cope with it. Therefore, the water is unnecessarily consumed by continuous water leakage, and the cause of the water leak becomes worse, and eventually the water pipe needs to be replaced. will be.

In addition, the maintenance of water supply facilities is becoming increasingly important as the demand for water and quantitative quality increases due to the improvement of citizens' living and economic standards. In order to reduce the leakage rate, each government is investing a lot of budget and manpower through projects such as maintenance of old pipes, construction of drainage basin for stable water pressure, leak detection and night minimum flow measurement. In particular, because pipelines are a major component of water supply facilities, a lot of investment is being made for the maintenance and replacement of defective pipelines. However, the project costs a lot of money and effort due to cutting and road excavation. Therefore, in order to efficiently execute such projects, a more rational and easy-to-use approach is needed to predict the priorities for the maintenance and replacement of bad pipelines.

The water supply network is buried underground and complicatedly enclosed in a very large area, so that physical direct diagnosis is very difficult. Therefore, it is necessary to reduce the investigation target area to the vulnerable area through indirect area diagnosis and to undertake the direct diagnosis or improvement of area through special intensive investigation on the vulnerable areas only. For this purpose, the soundness evaluation of the water supply network is most important, and it is an object of the present invention to more efficiently manage the water supply network through various soundness evaluation methods.

The present invention for achieving the above object is utilized as information for water supply network management through five different soundness evaluation methods (score evaluation method, fuzzy method, ELECTRE method, logistic regression analysis, Bayesian method).

Specifically, the present invention is a soundness evaluation method for water supply network management, comprising the steps of inputting the water supply GIS small block DB, leakage accident and recovery data; Analyzing the leakage accidents for each small block and selecting performance indicators based on the water supply GIS small block DB and leakage accidents and recovery data; Determining a small block rank for each performance indicator; Classifying grades by performance indicators; Adding weights of performance indicators to the classified performance indicator ratings; And determining a small block comprehensive ranking.

According to another feature of the invention, the water supply GIS small block-specific DB includes the age, the water valve and the hydrant.

According to another feature of the invention, the soundness evaluation method for water supply network management, comprising: inputting the performance evaluation data of the pipeline; Grading the pipeline status by indicator; Determining a pipeline aging degree by adding weight to the graded indicator pipeline status; Determining a fuzzy aging degree; Training by applying the fuzzy aging degree to a basic pipeline aging degree prediction model; Predicting pipeline aging through the training; Calculating a likelihood of failure through the predicted pipe aging; Combining damage results predicted through the calculated damage probability, fuzzy rule-set, and damage size according to the diameter; A soundness evaluation method is disclosed, comprising: predicting a risk of breakage of a pipeline by applying a fuzzy algorithm.

According to another feature of the present invention, the performance evaluation data of the pipeline includes a diameter, species and age.

According to still another aspect of the present invention, a soundness evaluation method for water supply network management, comprising: inputting water-related data and leakage accident recovery data on a GIS; Selecting performance indicators; Determining and normalizing ranks by blocks according to the performance indicators; Determining a consistency index and an inconsistency index; Determining strong rank preferences and weak rank preferences; Determining a forward rank and a reverse rank; Determining and applying a threshold; Determining a model for each R model; Applying an average absolute error, a root mean square error, a correlation coefficient, and a rank matching rate to the model for each R model; A soundness evaluation method is disclosed, comprising: evaluating a model to determine an optimal model.

According to still another aspect of the present invention, a method for evaluating soundness for water supply network management, comprising: inputting a pipe break and a leak record; Processing data using a GIS; Obtaining pipe condition indicators and environmental indicators; Performing logistic regression; Analyzing model fit and determining model coefficients; Determining a tube failure model; And determining the tube breakage classification criteria is disclosed.

According to still another aspect of the present invention, a soundness evaluation method for water supply network management, comprising: inputting a leak accident data; Determining a state factor; Combining the individual factors; Calculating a conditional leak probability and an average leak probability; Applying the conditional leak occurrence probability and the average leak occurrence probability to all the constant pipeline data; Determining a conduit length ratio of the conditional leak occurrence probability that is greater than the average leak occurrence probability of each block; A soundness evaluation method is disclosed, including: determining the probability of occurrence of leaks for each block.

According to another feature of the invention, the step of inputting the leakage accident data; Determining a state factor; Combining the individual factors; Calculating a conditional leak probability and an average leak probability; Applying the conditional leak occurrence probability and the average leak occurrence probability to all the constant pipeline data; And determining the probability of occurrence of leaks for individual pipelines.

According to another feature of the invention, the GIS is interlocked from the step of applying to the entire water pipe data.

According to another feature of the invention, the state factors include diameter, species and age.

In accordance with another aspect of the present invention, a computer readable medium comprising a program for executing the above-described soundness evaluation method is disclosed.

In accordance with another aspect of the present invention, a health assessment analysis system is disclosed that includes such computer readable media.

According to another feature of the invention, such a soundness evaluation analysis system; Web pages; A system linker connected to the pressurized system DB; Situation board system; A real-time alert transmission system that measures the flow rate or pressure of the facility and transmits an alert and a leak warning about an abnormal condition; And a flow rate monitoring system for monitoring block flow rate / pressure.

According to the present invention, the management practitioners of the water supply network can obtain a more rational and accessible method and system that can predict the priority of the maintenance and replacement of the bad pipe.

1 is a flowchart of a waterworks block system evaluation program using the scoring method.
2 is a flowchart illustrating a water pipe failure risk prediction program using a fuzzy technique.
3 is a flowchart illustrating a water supply block system evaluation program using a multi-criteria decision-making method.
4 is a flow chart illustrating a logistic regression analysis, that is, a water pipe failure prediction program through statistical analysis.
5 is a flowchart illustrating a leak frequency spatial analysis program using the Bayes technique.
6 is a diagram showing the configuration of a block flow monitoring system network according to an embodiment of the present invention.
7 illustrates a block flow monitoring system according to an embodiment of the present invention.
8 is a diagram illustrating an integrated management system according to an embodiment of the present invention.

As described above, the soundness evaluation of the water supply network of the present invention is performed through the following five soundness evaluation methods (score evaluation method, fuzzy method, ELECTRE method, logistic regression analysis, Bayesian method).

1. Scoring Method

A small block database was constructed using relevant data on water supply GIS, leakage accidents and recovery data in the study area, and the current status and characteristics were analyzed. By selecting performance indicators such as pipeline ratio, ratio of non-corrosion resistant pipe, leakage accident and stability, we can identify the blocks that need improvement of the constant pipeline through the small-block soundness evaluation. It can be easily accessed by practitioners because the amount of data to be acquired is relatively small and the procedure does not require much background knowledge. It can be used as a part to prioritize the selection of replacement areas. It is expected that progress can be made more effective when performance improvement investment projects.

This scoring method is a program for evaluating the soundness of block systems using water related data on GIS, leakage accident and recovery data, and performance indicators for small block evaluation. The data can be used as is and can be easily used by anyone as a decision tool for rehabilitation and replacement prioritization of small blocks.

1 is a flowchart of a waterworks block system evaluation program using the scoring method.

First, input the water supply GIS small block DB, leakage accident and recovery data. Then, based on the GIS small block DB, leakage accidents, and recovery data, the leakage accidents are analyzed and the performance indicators are selected. Based on this, the small block rank for each performance indicator is determined and the grade for each performance indicator is classified. The soundness of the water supply network is evaluated by determining the overall ranking of the small blocks by adding the weights of the performance indicators to the classifications of the performance indicators.

The water supply GIS small block DB may include, but is not limited to, a pipe age, a water valve and a fire hydrant.

2. Fuzzy Techniques

Although water pipes are introduced with various types of pipes with a quantitative increase, the service life of water pipes can be significantly shorter. In addition, due to the weak durability and functionality of the water pipes not only problems such as the generation of rust of tap water, but also the situation that causes economic losses due to leakage. To reduce this leak rate, many governments are implementing policies such as replacing old water pipes.

It is difficult to evaluate and predict the aging of the water supply pipe because it is buried underground, it is difficult to visually check the aging status, and there is not enough data to evaluate the old age. . Therefore, a more effective and accurate methodology for estimating or predicting the aging of water pipes is required. The method of evaluating the aging degree of a water supply pipe is a direct method using data that directly examines the condition of the pipe, such as pipe thickness, coating thickness, tensile strength, yield strength, and electrical conductivity, and indirectly such as pipe type, diameter, laying year, and soil type. There is an indirect way of using survey data.

In the present invention, the aging predictive model of the water supply pipe was studied using the indirect evaluation factors for the large waterworks of the global waterworks. In the case of large water pipes, the damage is very serious when leakage occurs due to aging or damage. Therefore, it is appropriate to approach rehabilitation and replacement of old water pipes in a preventive manner rather than symptomatically. do.

At present, there are many cases where there is a lack of data on the factors affecting the aging of water pipes, and there are a lot of subjective or ambiguous cases. In particular, it is difficult to define the degree of deterioration because the definition of the aging constant pipeline is also variously expressed. One of the more objective and quantitative methods of expressing the subjective and ambiguous data and the meaning of the old water supply pipeline is the fuzzy technique, and if the fuzzy technique is used in the aging evaluation model of the water supply pipeline, To overcome some of the limitations of the evaluation model, the Markov process is a technique that predicts the future state by using the current state and provides information for the water supply network by predicting the aging of the water pipe. do.

Fuzzy technique is a program that predicts the aging and damage risk of water pipeline by using fuzzy technique which is useful for performance evaluation data and uncertainty about buried water pipeline. In the case of buried water pipes, it is not easy to grasp the age and condition of the pipes, and most of the subjective and ambiguous data in the old water pipes. It is a program that can perform the rehabilitation and replacement of water pipelines by using fuzzy techniques that objectively and quantitatively express these data to predict the aging and damage risks of water pipelines more precisely and accurately.

2 is a flowchart illustrating a water pipe failure risk prediction program using a fuzzy technique.

First, as can be seen in the figure, the performance evaluation data of the pipeline is input. Performance evaluation data of such pipelines may include, but are not limited to, diameter, species, age. Next, based on the performance evaluation data of these pipelines, the pipeline status for each indicator is graded. The pipe age is determined by adding weights to the graded indicator pipe states, and the fuzzy age is determined. The determined fuzzy aging is applied to the basic pipeline aging prediction model for training. Through the training step, the pipe age is predicted to calculate the probability of failure. Combination of a) failure probability, b) fuzzy rule-set, and c) damage results predicted by the damage size according to the diameter, and the fuzzy algorithm is applied to the combined results to damage the pipeline. Assess health by predicting risk.

ELECTRE (ELimination Et Choice Translating REality) technique

By arranging and analyzing historical pipe breakage and leakage accident data and extracting the available data for each block through GIS, the soundness evaluation for each block is conducted to determine rational and systematic replacement and rehabilitation priorities for water pipe improvement projects. The purpose of this is to ensure that economic and efficient planning is achieved in the prior planning.

The Electre technique solved the problem by converting the result into a normal score to solve the problem that the results of the PI (Performance Indicators) for each small block are distributed over a wide range and there are extreme values. Survey the experts to obtain the weight, and weight the survey results by the AHP (Analytic Hierarchy Process) method, and replace the rational and systematic rehabilitation priorities with rational and systematic block water pipes using the weights and result values for each PI. The priority is determined using ELECTRE, a multi-criteria decision making method. Furthermore, in order to determine the priority of rehabilitation and rehabilitation of each pipeline in the block, the pipe failure probability function is derived by logistic regression using the pipe breakage and leakage accident data, and the ELECTRE and logistic regression analysis are used. In order to determine the priority of replacement and rehabilitation by small blocks and pipelines, this study uses relevant data on GIS, leakage accidents and recovery data, and there are many PI items for small block health assessment. Five categories of average pipeline age, pipeline ratio over 21 years, ratio of non-corrosive pipe, leakage accident, and stability can be evaluated based on 12 pipes and selected as 12 PIs. In addition, the method for ranking and comparing rehabilitation and rehabilitation by pipeline is summarized and used as the data of water supply network.

The ELECTRE technique is a program for determining the soundness ranking of block systems through relative comparison between block systems using water related data on GIS, leakage and recovery data, and performance indicators for small block evaluation. This technique evaluates the block system through relative comparison between block systems, and therefore is a program that can more accurately and logically determine regeneration and replacement priorities between block systems. This program can be used to more systematically establish the maintenance and rehabilitation of the entire block system.

3 is a flowchart illustrating a water supply block system evaluation program using a multi-criteria decision-making method.

As shown in Figure 3, first, the water supply related data and leakage accident recovery data on the GIS is input. Performance indicators are selected based on these data, and ranking and normalization are performed for each block according to the performance indicators. Thus, the consistency index and inconsistency index are determined, strong rank preferences and weak rank preferences are determined, and then forward rank and reverse rank are determined. The threshold value is determined and applied to the calculated data to determine the model for each R model. Then, the model is evaluated by applying the mean absolute error, the root mean square error, the correlation coefficient, and the rank matching rate to the R model.

4. Logistic Regression

Factors affecting pipe damage are extracted by using various data such as the current status of pipeline facilities in the metropolitan waterworks, 7-year history of pipe breakage and leak recovery, buried environment, water pressure, and the number of years of buried. We analyzed the failure probability function using the logistic regression model. In addition, it analyzes the failure model by subdividing it into a group with characteristics to enhance the reliability of the probability function with accurate and significant analysis results.In case of the management and maintenance, rehabilitation and replacement of damage and leakage of large water pipes It can be used as an indirect judgment basis of statistical techniques in setting priorities, and through this, it is possible to reduce the frequency of pipeline breakage and improve flow rate by more proactive network management by establishing and coping with accident prevention plans such as inspection and maintenance in advance.

Logistic regression is a program that predicts current or future tube breaks using past breakage data. This technique is a useful program for predicting pipe failure through statistical analysis when the history of pipe breakage is secured in the past, and it is easy and useful as a decision tool in pipe rehabilitation and replacement business by determining the probability of failure and replacement criteria for each pipeline. It is a program that can be utilized.

4 is a flow chart illustrating a logistic regression analysis, that is, a water pipe failure prediction program through statistical analysis.

Again, as shown, this program is initiated by entering a record of tube breaks and leaks. These records are processed using GIS to obtain pipe condition indicators and environmental indicators. Logistic regression is performed on these indicators to analyze model fit and determine model coefficients. Then, the tube failure model is determined based on the determined model coefficients. Finally, tube breakage classification criteria are determined to assess soundness.

5. Bayes Technique

Leakage is caused by complex interactions of various factors such as pipe material, construction technology, internal works, other works, natural disasters, laying environment, etc., and it is difficult to predict it, and even if the leak is recovered through leak detection, If it does, it has a recovery characteristic in which a leaking area that has not been found previously grows. Therefore, more rational and economical planning is needed considering the scientific forecasting method of leakage accidents and the nature of leakage restoration. Analyze the characteristics of leakage accidents in the target area to calculate the probability of future leakage for a combination of significant influence factors, and find the increase in natural non-reported leaks and target leakage reduction for small blocks that can be measured in real time. Calculate the economic leakage level and the frequency of leak detection, and find the combinations that show the high probability of leakage among the combinations of the factors affecting the leakage accidents, and determine the total pipeline extension for each small block and the percentage of pipeline extension with those combinations. In order to determine the priority of each small block, the priorities for improvement, replacement and leakage detection of small blocks including the leak bundle pipeline, and to calculate the economic leak level and the frequency of leak detection of small blocks selected as vulnerable areas Can be utilized to establish a more rational leak plan so that It is expected to contribute to the improvement.

Bayes 'technique is a water leak channel and drainage block frequency determination program using leak data and Bayes' theorem. According to this technique, it is possible to calculate the leak probability more accurately by determining the conditional probability with the given leak data and calculating the posterior probability of the constant pipeline by using it. This can be used to effectively plan rehabilitation and replacement of blocks and water pipes.

5 is a flowchart illustrating a leak frequency spatial analysis program using the Bayes technique.

As described in the figure, first, leakage accident data is input and a state factor is determined. Status factors may include, but are not limited to, diameter, species and age. Then, the individual factors are combined and the probability of conditional leakage and the average leakage probability are calculated. The conditional leak probability and the average leak probability are applied to the whole constant pipeline data, and the ratio of the length of the conditional leak is greater than the average leak probability of each block. Finally, the probability of leak occurrence for each block is determined.

Alternatively, after applying conditional leakage probability and average leakage probability to all constant pipeline data, the probability of leakage of individual pipelines is determined.

From the step of applying to the entire water pipe data can be linked to GIS.

As described above, in general, a method for evaluating the health of a water supply network for indirect area diagnosis and the like for underground water supply networks has been exemplified, but the present invention is not limited thereto and may include various modifications and changes.

The block flow monitoring system will now be described with reference to FIGS. 6 to 8. 6 is a diagram illustrating the configuration of a block flow monitoring system network. This system network is a sensor measurement system (SITE # 1 to #n) equipped with CDMA equipment for transmitting such data from instruments measuring flow rate information, e.g., instantaneous flow, integrated flow and pressure. And a flow rate monitoring web server including a wireless Internet network for receiving data transmitted from the system, and a data server for storing and processing the above-described measurement data, and an image system for control status and management. Water supply management headquarters, establishments, environment departments, related organizations or citizens may communicate with the above-described flow monitoring web server through a computer terminal or the like. In addition, field personnel and managers can also communicate with the above-described flow rate monitoring web server through a laptop or mobile phone, through which the water supply network can be monitored and managed.

7 illustrates a block flow monitoring system according to an embodiment of the present invention. The block flow monitoring system includes an integrity evaluation analysis system for evaluating and analyzing leakage health and managing a water supply network using the above-described soundness evaluation method; Web pages; A system linker connected to the pressurized system DB; Situation board system; A real-time alert transmission system that measures the flow rate or pressure of the facility and transmits an alert and a leak warning about an abnormal condition; And a flow rate monitoring system for monitoring the block flow rate / pressure. Such a system allows practitioners to easily assess the health of the water supply network, thereby effectively managing the water supply network. The health assessment analysis system described above may include a computer readable medium containing a program for executing the health assessment method described above.

8 is a diagram illustrating an integrated management system according to an embodiment of the present invention. 8 illustrates an integrated block monitoring system of Incheon city in Korea, which is web-based, and is composed of a system and an integrated system for each business establishment, and is composed of a map-based meshup information display and data analysis system. The operation menu includes real-time information, regional analysis, hourly analysis, alarm status and facility status. The log-in information is divided into integrated managers and managers of each business site, and differentiated information accessibility, management scope and authority. In the measurement information, information such as instantaneous / accumulated flow rate and instantaneous / maximum / minimum pressure is displayed. The weather information displays temperature, rainfall, and humidity in connection with external systems. In the measurement system report, graphs / tables were used to generate the report by daily / monthly / yearly. In such reports, for example, the flow rate (m ³ / h), the hydraulic pressure and the flow rate can be displayed in graphs or tables, for example, the maximum flow rate is indigo blue, the minimum flow rate is blue, and the maximum hydraulic pressure is red. The minimum flow rate can be marked in orange, the maximum flow rate in green, and the minimum flow rate in light green. Map-based measurement information is composed to show communication status and status by measurement data.

As described above, although the present invention has been described with reference to the limited embodiments and the drawings, the present invention is not limited to the above-described embodiments, and the present invention is not limited to the technical spirit of the present invention. Many changes and modifications are possible by those skilled in the art.

Through the present invention, economical and efficient planning is possible when establishing a preliminary plan for improving the performance of a bad pipeline of a water supply pipe network.

Claims (1)

Block flow monitoring system for evaluating and analyzing the health of water supply network management,
Health assessment analysis system;
Web pages;
A system linker connected to the pressurized system DB;
Situation board system;
A real-time alert transmission system that measures the flow rate or pressure of the facility and transmits an alert and a leak warning about an abnormal condition; And
A flow monitoring system for monitoring block flow / pressure;
The soundness evaluation analysis system,
Inputting a tube break and leak record;
Processing data using a GIS;
Obtaining pipe condition indicators and environmental indicators;
Performing logistic regression;
Analyzing model fit and determining model coefficients;
Determining a tube failure model; And
Determining a tube breakage classification criterion; and a computer readable medium having recorded thereon a program for executing the block flow classification criteria.

Images