CAJ | 학술논문

　　针对BP神经网络易陷入局部极小问题以及收敛速度慢的问题，引入量子粒子群优化算法和BP神经网络相结合的方法，共享BP神经网络强大的灵活性和量子粒子群全局搜索能力强的优势，通过改进QPSO的平均最优位置的计算方法，实现基于 BP 神经网络和量子粒子群的油田节能指标预测。以大庆某采油厂注水泵机组单耗数据为训练数据，预测结果表明该方法能达到良好的预测效果，具有可行性。
　　침대BP신경망락역함입국부겁소문제이급수렴속도만적문제，인입양자입자군우화산법화BP신경망락상결합적방법，공향BP신경망락강대적령활성화양자입자군전국수색능력강적우세，통과개진QPSO적평균최우위치적계산방법，실현기우 BP 신경망락화양자입자군적유전절능지표예측。이대경모채유엄주수빙궤조단모수거위훈련수거，예측결과표명해방법능체도량호적예측효과，구유가행성。
According to the fact that BP neural network is easy to fall into local minimum and the slow convergence problems, the paper introduces QPSO and BP neural network combination method, which shares the advantage of BP neural network robust flexibility and the powerful global searching ability of QPSO, through improved the calculation method of average optimal position of QPSO to make the BP neural network and QPSO oilfield energy conservation index prediction success. Using the injection pump unit consumption data of Daqing Oilfield Company as training data, by training the new mehtod with the data of samples, the forecast results show that the proposed method can achieve good forecast effect and have feasibility.