关键词: 人工神经网络, 机械钻速, 深井钻井, 钻头, 优选, 自适应共振理论

Abstract:

Based on the properties of deep_well drilling in Daqing,such as hard formation,high temperature and big dip,a method for optimizing bit type with artificial neural network is proposed.The factors that affect the bit selection are taken as input neural nods,and the quantitative method is combined with the qualitative method.This new method makes the optimizing results of bit more reliable.Considering the characteristics of high degree of hardness, abrasion and high temperature of Jurassic layer,some special bits were investigated and applied.On the basis of statistical analysis about the bits used in Daqing deep wells and field tests of new type bits,a bit database was developed.The types of bit used in Daqing deep wells were optimized with artificial neural network method.The field tests of two deep wells in Daqing show that the drilling rates using optimized bits are raised more than 20%.

Key words: artificial neural network, drilling rate, bit optimization, deep well drilling, adaptive resonance theory