Research And Implementation Of Oil-gas Prospecting Costs Forecasting Method

Oil-gas prospecting cost is an important index which reflects the economic benefit of oilfield enterprise. Making an accurate forecast on the prospecting cost can help the investors and enterprise directors to carry out scientific decisions and estimates.As the distinction between oil production and the production of ordinary goods exists, however, the traditional methods turn nonlinear problems into linear or treat them as linear problems so as to forecast oil-gas prospecting cost, which can not achieve an accurate forecast in actual situation, the specific methods must be used to forecast oil-gas prospecting cost.Firstly, the theory concerning neural network is researched in detail. Secondly, starting from analyzing cost behavior and taking the characteristics of oil-gas prospecting cost into account, the influential factors of oil-gas prospecting cost is identified. Based on these, multiple linear regression, LMBP, MOBP, VLBP and BRBP model are used to forecast oil-gas prospecting cost, and each of their forecasting performances is compared. Finally, BP neural network based on principal component analysis is used to forecast oil-gas prospecting cost, and its high prediction accuracy has been proved.