Pipe failure rate prediction in water distribution networks using multivariate adaptive regression splines and random forest techniques

2019 ◽  
Vol 16 (9) ◽  
pp. 653-661 ◽  
Author(s):  
Akbar Shirzad ◽  
Mir Jafar Sadegh Safari
Keyword(s):  
Random Forest ◽  
Failure Rate ◽  
Water Distribution ◽  
Distribution Networks ◽  
Multivariate Adaptive Regression Splines ◽  
Water Distribution Networks ◽  
Regression Splines ◽  
Pipe Failure ◽  
Adaptive Regression ◽  
Adaptive Regression Splines

scholarly journals Novel Approaches to Modelling Flammability Characteristics of Polymethyl Methacrylate (PMMA) via Multivariate Adaptive Regression Splines and Random Forest Methods

2020 ◽  
pp. 1-14
Author(s):  
Mohammed Okoe Alhassan ◽  
Michael Boakye Osei
Keyword(s):  
Random Forest ◽  
Polymethyl Methacrylate ◽  
Polymeric Materials ◽  
Multivariate Adaptive Regression Splines ◽  
Coefficient Of Determination ◽  
Regression Splines ◽  
Soft Computing Techniques ◽  
Flammability Characteristics ◽  
Adaptive Regression ◽  
Adaptive Regression Splines

Soft-computing techniques for fire safety parameter predictions in flammability studies are essential for describing a material fire behaviour. This study proposed, two novel Artificial Intelligence developed models, Multivariate Adaptive Regression Splines (MARS) and Random Forest (RF) methods, to model and predict peak heat release rate (pHRR) of Polymethyl methacrylate (PMMA) from Microscale Combustion Calorimetry (MCC) experiment. From the statistical analysis, MARS presented the highest coefficient of determination (R2) values of (0.9998) and (0.9996) for training and testing respectively, with low MAD, MAPE and RMSE values. Comparatively, MARS outperformed RF in the predictions of pHRR, through its model algorithms that generated optimized equations for pHRR predictions, covering all non-linearity points of the experimental data. Amongst the input variables (sample mass, THR, HRC, pTemp and pTime), heating rate (β), highly influenced pHRR outcome predictions from MARS and RF models. However, to validate the performance and applicability of the proposed models. Results of MARS and RF were benchmarked with that from Artificial Neural Network (ANN) methods. The MARS and RF models observed the least error deviation when compared with pHRR results for PMMA from the ANN models. This study therefore, recommends the adoption of MARS and RF in the predictions of flammability characteristics of polymeric materials.

scholarly journals Vulnerability of water distribution networks with real-life pipe failure statistics

2021 ◽  
Author(s):  
Richárd Wéber ◽  
Tamás Huzsvár ◽  
Csaba Hős
Keyword(s):  
Failure Rate ◽  
Water Distribution ◽  
Ph Value ◽  
Real Life ◽  
Distribution Networks ◽  
Water Distribution Networks ◽  
Pipe Material ◽  
Power Law Distribution ◽  
Pipe Failure ◽  
Failure Statistics

Abstract Reasons for occasional, random pipe bursts in water distribution networks (WDNs) might come from numerous factors (e.g. pH value of the soil, the pipeline material). Still, the isolation of the damaged section is inevitable. While the corresponding area is segregated by closing the isolation valves, there is a shortfall in drinking water service. This paper analyses the vulnerability of segments of WDNs from the viewpoint of the consumers that is the product of the failure rate and the relative demand loss. Real pipe failure database, pipe material and pipe age data are used to increase the accuracy of the failure rate estimation for 27 real-life WDNs from Hungary. The vulnerability analysis revealed the highly exposed nature of the local vulnerabilities; the distribution of local vulnerability values follows a power-law distribution. This phenomenon is also found by investigating the artificial WDNs from the literature using N rule in terms of isolation valve layout, namely the ky networks, with similar results.

Sign in / Sign up

Export Citation Format

Share Document