선택한 단어 수는 입니다.
최소 단어 이상 선택하여야 합니다.
최소 단어 이상 선택하여야 합니다.
선택한 단어 수는 30입니다.
최대 10 단어까지만 선택 가능합니다.
최대 10 단어까지만 선택 가능합니다.
제안 방법
- (1) Above all, explanatory variables and the category of explanatory variables that was selected through the analysis of the characteristics of large-water-pipeline bursting were used. As for the variable selection, a logistic-regression model was developed using the forward selection method.
- Data on pipe bursting and leakage cases were used to improve the large water pipeline management system and to facilitate pipe replacement projects. As a result, the factors that influence pipe bursting the most were identified, and each pipe bursting and leakage case was analyzed by studying the characteristics of the leakage cases.
- Data Collection and Scope of Research: In this study, various data were analyzed(including GIS) to investigate and examine the documents or records of the aged water pipes of South Korea and their repair records, as well as the geologic conditions of the areas where water mains are installed.
- Determination of the standard for the categorization of the pipe bursting risk : In this study, the observed values were analyzed by assigning T' for bursting and “0” for no bursting. The errors in the predicted variable found through the functional formula can be demonstrated in two ways.
- In this study, a new system for leakage maintenance was developed, and the statistical percentage of water pipe bursting was determined by identifying the factor that influences pipe bursting the most. The related accidents were also analyzed by case, using the data obtained from the research that was conducted on the different cases of damage and leakage of water mains that occurred over seven years, along with the information on the present condition of the water utilities throughout South Korea.
- In this study, models for leakage management and statistical bursting probability were developed. Data on pipe bursting and leakage cases were used to improve the large water pipeline management system and to facilitate pipe replacement projects.
- The time of leakage was categorized into different seasons, months, and days and night. Moreover, the pipe burst probability model was designed by selecting the factors that influence pipe bursting.
- It was also found that there are various reasons for pipe bursting. Pipe bursting was classified and analyzed according to the material, diameters, installation environment, water pressure, age, seasonal or monthly changes, causes of bursting, and damaged parts.
- Research procedure- Various research data were used in this study: the report on the present condition of the wide pipe networks throughout South Korea, the detailed records of the water mains that were broken and that leaked from 2001 to 2007, and the data obtained from an analytical study on the present condition of the water services on GIS(the environmental factors in the areas where water pipes are installed). Research was conducted using the method shown in Fig.
- This amount was determined by dividing the service expense for the regular inspection of the pipes as preventive measures by the average number of pipe bursting cases. The amount of loss incurred by the errors was computed according to the classification standard, by multiplying the amount of loss in one case with the misclassification error rate. The point at which the total amount of loss in all the cases was minimal was chosen as the classification standard.
대상 데이터
- The total number of pipe bursting and leakage cases over the aforementioned seven-year period was 377; from these, damage artificially occurred in 11 cases. The specific characteristics of the pipe bursting were studied based on 326 cases, excluding the 11 cases where damage artificially occurred. The primary causes of pipe bursting were found to be aging(151 cases), defective materials(56 cases), external weight pressure(49 cases), pipe expansion and contraction caused by temperature changes(24 cases), water pressure within the pipes(21 cases), and others(25 cases).
데이터처리
- Logistic regression was performed using the variable shown above, and forward selection, a stepwise regression, was chosen as the logistic-regression coefficient. As for the significance of the regression coefficient, using Wald statistics, a coefficient that was significant within the significance level of 5% was chosen.
- It is believed that correctly identifying the pipes that have been damaged is crucial, for which reason the prior mistake was designated as type I error. To identify the optimal criteria of pipe bursting probability for pipe replacement and rehabilitation, the amounts of loss incurred by the two aforementioned errors were determined through approximate calculation. That is, for the type I error, as the pipes that have actually been broken were categorized into pipes that have never been broken, the amount of leaking water caused by the neglect of the pipe bursting was regarded as the amount of loss.
이론/모형
- bursting were used. As for the variable selection, a logistic-regression model was developed using the forward selection method. As a result of such development, the null hypothesis that “the coefficient of the independent variables included in the models is 0” was rejected.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.