|
|||||
|
|
| 基于知识发现的火箭武器研制费用预测 | |
| 引用本文： | 王锋,吴晓云,马大为. 基于知识发现的火箭武器研制费用预测[J]. 系统工程与电子技术, 2007, 29(1): 69-72 |
| 作者姓名： | 王锋 吴晓云 马大为 |
| 作者单位： | 1. 南京理工大学机械工程学院,江苏,南京,210094;防空兵指挥学院自动化指挥控制教研室,河南,郑州,450052 2. 炮兵学院火箭炮教研室,安徽,合肥,230031 3. 南京理工大学机械工程学院,江苏,南京,210094 |
| 摘 要： | 小样本的火箭武器研制费用预测通常难于应用线性回归方法,而灰色理论方法在实际中仍不能较好地解决费用与武器特征参数间存在的非线性问题。提出了融合粗集理论和神经网络预测火箭武器研制费用的新方法,利用粗集知识约简后的特征要素作为神经网络的输入,实现火箭武器研制费用的预测,并用实例证明了基于粗集-神经网络的费用预测精度高于灰色模型预测精度。 |
| 关 键 词： | 人工智能 粗集 神经网络 武器研制 费用预测 |
| 文章编号： | 1001-506X(2007)01-0069-04 |
| 修稿时间： | 2006-01-07 |
Cost forecasting of rockers weapons development based on knowledge mining |
|
| WANG Feng,WU Xiao-yun,MA Da-wei. Cost forecasting of rockers weapons development based on knowledge mining[J]. System Engineering and Electronics, 2007, 29(1): 69-72 | |
| Authors: | WANG Feng WU Xiao-yun MA Da-wei |
| Abstract: | It is difficult to predict the cost of rockets weapon research under the small sample situation for the linearity regression method,while the nonlinear problem between the cost and characteristic parameters can't be solved by the gray theory.So a new method combined knowledge mining by rough set and neural network is proposed,which can use the relative reduce theory to mine the knowledge behind these elements,then certain elements after being reduced is selected as the input of neural network.Based on rough set and neural net work,the development cost is forecasted.An example is calculated to prove that the precision of the new method is higher than that of the gray model. |
| Keywords: | artificial intelligent rough set neural network weapon development cost forecast |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |