CAJ | 학술논문

针对火箭主动段飞行过程中，插值外推等算法时间短，目标丢失后测控设备无法快速重捕目标的问题，采用BP神经网络，以实际测量的相关联数据作为训练样本，通过指导式学习算法对火箭主动段的弹道进行预测。仿真实验表明，相比较以往的插值外推算法，基于BP神经网络对火箭主动段外弹道的预测时间长、精度高，能有效地提高引导的精度和可靠性。
침대화전주동단비행과정중，삽치외추등산법시간단，목표주실후측공설비무법쾌속중포목표적문제，채용BP신경망락，이실제측량적상관련수거작위훈련양본，통과지도식학습산법대화전주동단적탄도진행예측。방진실험표명，상비교이왕적삽치외추산법，기우BP신경망락대화전주동단외탄도적예측시간장、정도고，능유효지제고인도적정도화가고성。
In the Launch Vehicle Powered Flight, it is the problem that time is short in extrapolation of interpolation algorithm, and can not re-capture fast when the target loss. Use the BP neural network and take reality and relevant data as the training sample. Forecast the trajectory of launch vehicle by algorithm of learning. Simulation results show that, compared with the interpolation and extrapolation algorithms in the past, forecast based on BP neural network has characteristics of long time and high precision, and it can improve the guidance accuracy and reliability effectively.