Description
Gas Electron Multiplier (GEM) detectors are central to modern particle physics instrumentation, yet their performance is often limited by a persistent discrepancy between simulated and experimentally measured gain. This book examines one of the main causes of that discrepancy: the charging-up effect, in which electrons and ions accumulate on the insulating Kapton surfaces of GEM holes and modify the local electric field over time. Numerical Modeling of Charging-Up in Gas Electron Multipliers: A Genetic Programming Approach presents a detailed study of time-dependent gain variation in single- and triple-GEM detectors. It introduces the physics of charge accumulation, dielectric polarization, field modification, and gain stabilization, then develops a data-driven modeling framework based on Genetic Programming (GP). Unlike conventional fitting methods or black-box machine learning models, GP performs symbolic regression and produces explicit mathematical expressions that relate detector gain to time, radiation rate, and applied voltage. Through systematic comparison with traditional exponential fitting functions, the book demonstrates that GP-derived models can achieve high predictive accuracy, with correlation coefficients exceeding 0.97 in most cases and substantially reduced mean squared error. The approach is shown to be especially effective in generating unified models valid across multiple operating conditions, offering a practical path toward correcting gain instabilities and improving agreement between simulation and experiment. Written for advanced students, researchers, and practitioners in experimental particle physics, detector development, and computational modeling, this volume combines theoretical background, methodological explanation, model results, and academic commentary. It highlights the value of evolutionary computation as an interpretable and powerful tool for solving complex modeling problems in modern radiation detector systems.