TA761 : Pipe failure prediction in water distribution networks using evolutionary polynomial regression (EPR)
Thesis > Central Library of Shahrood University > Civil & Architectural Engineering > MSc > 2023
Authors:
[Author], Ramezan Vagheei[Supervisor], [Supervisor]
Abstarct: The water distribution network is the last lixnk in the consumption chain in human life today. Incidents and events are considered an inevitable part of the operation of the distribution network. It is absolutely necessary to study and analyze the factors affecting the fractures, especially in the branches of the subscribers, in order to reduce the cost and prevent the wastage of water and increase the satisfaction of the people. In recent years, the use of statistical computing methods in analyzing the causes of failures and predicting fractures has been the focus of researchers. The purpose of this research is to evaluate the effect of network pressure parameters, air temperature and branch age on the number of failures. The data of failures (16631 failures) in one of the areas of Abfai Tehran City during the period (1396-1400) in addition to air temperature data (14600 data) have been collected from Tehran's Mehrabad station. According to the minimum and maximum values of each parameter, equal intervals of 10 years have been considered for age, 0.4 bar for pressure and 5 degrees Celsius for temperature. The number of incidents in each of the intervals is determined and the distribution of branching incidents has been investigated. The evaluation of the data was done with the Normal Quantile Plot method and it shows the normality of the data. For each of the intervals, the failure rate has been calculated according to the accident data and the data of the reference society. The graph of the changes of each rate in different intervals is drawn and the fitting curve is drawn after removing the outliers. R2 parameter for the 3rd degree equation of pressure and temperature incident rate was 0.86 and 0.93, respectively, and for the 2nd degree equation, the incident rate of branching age was 0.95. An increase in pressure leads to an increase in the accident rate, especially after 5 bar pressure. With a 50% increase in pressure, the number of accidents increases by 11.78% and by doubling the pressure, the number of accidents increases by 60.52%. Increasing the age of separation up to 40 years increases the accident rate and after that it has a decreasing trend, which can be technically justified. The effect of temperature on the accident rate has been more noticeable. The calculation equation for the pressure incident rate baxsed on positive field data has a 9% error and for the negative temperature incident rate, it has an error of 2.9%.
Keywords:
#water distribution network #water distribution #failure prediction #evolutionary polynomial regression Keeping place: Central Library of Shahrood University
Visitor: