中文摘要：

降低大型客机直接维修成本（DMC）是提高经济性的重要途径。笔者综合分析了大型客机直接维修成本影响参数,针对影响参数和大型客机研制阶段数据特点,提出采用基于改进粒子群（IP-SO）-BP神经网络的直接维修成本分析方法,实现定量的直接维修成本分析。建立的大型客机直接维修成本分析模型具有避免陷入局部最小值,加快学习速度的优点,通过实例验证了该方法的有效性。

英文摘要：

Reducing the direct maintenance cost （ DMC） is an important way to improve the economy of a large-scale passenger aircraft. It is necessary to analyze the direct maintenance cost for assisting the economy decision of a large-scale passenger aircraft. In this paper,influencing factors of DMC have been comprehensively analyzed. Considering the characteristics of the influencing factors in the large-scale passenger aircraft development phase,the model of integrating the improved particle swarm optimization （ IPSO） with BP neural network have been brought forward to analyze the DMC quantitatively. The model has the advantage of avoiding to fall into local minimum and speeding up the learning speed. The example demonstrates the effectiveness and validity of the method.