Fuzzy random programming models of oilfield for increasing output of oilfields

Abstract

Measures programming can help oilfields to extend production life, reduce mining difficulty and improve final recovery ratio. It has the typical uncertainty character, and the stochatic and fuzzy programming models have been constructed. In this paper, we regard the parameters including unit well times production and unit well times cost of each measure as fuzzy random variables, and construct fuzzy random programming models of oilfield measures, including the expected production model and chance-constrained production model. Then we propose a hybrid intelligent algorithm ingrating fuzzy random simulation, neural network and genetic algorithm to solve the uncertain programming models. Finally, we provide two real examples to testify the validity of the fuzzy randon models and the hybrid intelligent algorithm.