Non-renewable Recourses Life Cycles Modeling Aspects

Abstract

Abstract The task of modeling and forecasting non-renewable resources production using bell-shaped curve-fitting models is considered. Identification of model parameters is based on the genetic algorithm. The purpose of this article is to determine the best genetic algorithm configuration for the task of bell-shaped life cycle models parameters identification. We find the initial population optimal size, and find that heuristic function is the best for crossover and adaptive feasible function is the best for mutation. The experience of modeling 235 time series of oil and gas production is summed up. On the basis of it we made the frequency analysis of models using for every kind of task. It allows to select the preferred models for describing oil, gas and shale gas production at the different levels of aggregation: from a country and a region to a single field.