【摘要】 本文针对传统飞行计划算法迭代次数多、时间长等问题,研究了利用BP神经网络缩短飞行计划计算时间的可行性。在传统已知起飞重量飞行计划算法的基础上,通过BP神经网络对主航段下降顶点重量的迭代初值进行预估,减少了迭代次数。使用MATLAB编程构建训练网络,并与传统飞行计划算法程序进行整合,编写了飞行计划优化算法程序。验证计算结果表明,构建的优化算法与传统算法对比具有同等精度,计算时间显著缩短,说明将BP神经网络方法运用到飞行计划计算是可行的。更多还原

【Abstract】 In view of the many iterations and long calculation time of traditional flight planning algorithms, the feasibility of using BP neural network to shorten the flight planning calculation time is studied. On the basis of the traditional flight plan algorithm with known takeoff weight, the initial iterative value of weight at top of descent of the main segment is estimated through the BP neural network, which reduces the number of iterations. A flight planning optimization algorithm program is developed by using MATLAB to build a training network and integrating it with traditional flight planning algorithm program. The verification calculation results show that the optimization algorithm has the same accuracy as the traditional algorithm, and the calculation time is significantly shortened, which indicates that it is feasible to apply the BP neural network method in flight plan calculation.更多还原