摘要
装备管理中维修人员人力三要素(数量、专业、技术等级)以及它与装备的紧密相关性是运用相似系统法建立模型确定其需求的难点.在力求减少定性的不确定性的同时避免机械地定量化,从维修人员人力特有的角度确定了相似系统法的基准比较系统,结合维修实际确定了相似产品库;从确定相似比的非线性特点出发,运用层次分析法分析了结构相似度,特别是通过创造性地变换样本,离散了普同相似系统法中参数表象的线性关系,构造出合理的人工神经网络(artificial neural network,ANN)模型;设计并优化了网络的结构,从而很好地解决了在确定相似比参数过程中的非线性问题.
In the management of equipment, the three factors of maintenance manpower（quantity, major, technology level）and the close relativity with material are the difficult points of ascertaining by modeling with similar system means. Trying its best to reduce uncertain factors and to avoid mechanically quantifying, the article ascertains the basic comparable system of similar system means from the special point of maintenance manpower. It defines database of similar ‘product＇ on the basis of real maintaining work. And it analyzes structure similarity degree by means of level analyzing means, especially, disperses ostensible linear relation of parameter in general similar system means, and forms a rational ANN model by creative transforming the imputing data according to the non-linear character of defining similar ratio, completes the structure of ANN, and then, preferably settles the non-linearity problem in defining the parameter of similar ratio.
出处
《装备指挥技术学院学报》
2005年第4期27-30,共4页
Journal of the Academy of Equipment Command & Technology
关键词
相似系统
人工神经网络
人员人力
Similar system
artifical neural network（ANN）
personnel and manpower
