CAJ | 학술논문

装甲装备器材保障具有规模大、时间紧、消耗大、不确定因素多、决策难度大等特点。准确的需求预测是实施主动的、精细化的器材保障的重要前提条件。利用BP神经网络较强自学习能力和自适应能力对器材需求规律进行学习，并借助遗传算法提高BP神经网络的收敛速度，设计了一种基于遗传算法改进的BP神经网络模型预测方法，对装甲装备器材进行需求预测。通过实例计算表明，该方法比单纯BP神经网络方法具有预测精度高、收敛速度快的优点。
장갑장비기재보장구유규모대、시간긴、소모대、불학정인소다、결책난도대등특점。준학적수구예측시실시주동적、정세화적기재보장적중요전제조건。이용BP신경망락교강자학습능력화자괄응능력대기재수구규률진행학습，병차조유전산법제고BP신경망락적수렴속도，설계료일충기우유전산법개진적BP신경망락모형예측방법，대장갑장비기재진행수구예측。통과실례계산표명，해방법비단순BP신경망락방법구유예측정도고、수렴속도쾌적우점。
Armored equipment material support has some particular features,including large scale, time urgency,large consumption,a lot of uncertain factors and difficult decision. Accurate demand forecasting is an important prerequisite to implement an initiative and refinement equipment protection. In this paper,BP neural network learrning and self-adaptive ability is used to learn the law of equipment demand,genetic algorithm is used to improve BP neural network convergence speed. A genetic algorithm improved BP neural network algorithm is proposed for forecasting equipment demond. The experiments show that the proposed method offers the advantages of high precision and fast convergence in contrast with BP neural network.