小样本条件下航空装备费用预测

钟诗胜¹, 付旭云¹, 胡淑荣²

1.哈尔滨工业大学机电工程学院;2.国家林业局管理干部学院



摘要:

为了提高小样本条件下的航空装备费用预测的精度,将信息扩散方法和支持向量机相结合,提出了信息扩散支持向量机预测模型,对模型的拓扑结构、建模步骤进行了描述.为了采用粒子群优化算法为模型选择合适的参数,考虑到模型参数既有连续变量,又有离散变量,提出对粒子位置的各个分量采用不同的更新策略.将信息扩散支持向量机应用于军用飞机机载电子设备的生产费用预测,预测结果的平均相对误差绝对值为3.3%,表明该方法可以满足工程的实际需要.

关键词:  小样本  费用预测  信息扩散  支持向量机  粒子群优化算法

DOI：10.11918/j.issn.0367-6234.2011.05.010

分类号:V267

基金项目:国家自然科学基金重点资助项目(60939003)



Aviation equipment cost prediction under small sample size

ZHONG Shi-sheng¹, FU Xu-yun¹, HU Shu-rong²

1.School of Mechatronics Engineering,Harbin Institute of Technology,150001 Harbin,China;2.State Academy of Forestry Administration,102600 Beijing,China

Abstract:

To improve prediction accuracy of aviation equipment cost under small sample size,a prediction model named information diffusion support vector machine is proposed after combining the information diffusion method and support vector machine.The topology and modeling process of the model are also described.The model parameters include both continuous parameters and discrete parameters,so that a different update strategy is adopted to each component of the particle position when solving the problem of model parameter selection using the particle swarm optimization.Finally,the information diffusion support vector machine is applied to the production cost prediction of military aircraft avionics equipment.The average absolute relative error of the result is 3.3%,which can satisfy actual requirement.

Key words:  small sample  cost prediction  information diffusion  support vector machine  particle swarm optimization