University of Twente Student Theses

Login

Parameter Calibration of Stochastic Degradation Models of Sewer and Water Pipe Networks using Genetic Algorithms

Ottenschot, Bram (2024) Parameter Calibration of Stochastic Degradation Models of Sewer and Water Pipe Networks using Genetic Algorithms.

[img] PDF
625kB
Abstract:Modelling the deterioration of sewer and water pipes is vital for drawing up efficient maintenance strategies such that preventive maintenance can be executed at a certain time in the pipes’ life cycle when it is needed without having to perform regular, expensive CCTV inspections. One such kind of deterioration model is Homogeneous Discrete-Time Markov Chains, which are stochastic models that can predict the severity of damage by working with states symbolising the severity level, and transition for expressing the timeindependent probability of moving from one state to another in discretized time steps. Calibrating these transition probabilities is however not a trivial task. That is why exploring new calibration techniques for deterioration models of sewer systems is useful. This paper focuses on the application of Genetic Algorithms (GA) in calibrating Markov Chains by using evolutionary concepts like selection, crossover and mutation. The result of this research shows promising stable results following the application of the algorithms in the sense that the calibrated models perform well if compared to real-world data, and the fact that the algorithm is stable, e.g. multiple runs converged to similar performing calibrated parameter sets.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:31 mathematics, 54 computer science
Programme:Computer Science BSc (56964)
Link to this item:https://purl.utwente.nl/essays/100946
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page