基于BP神经网络的坦克炮火力威力评价模型  

Model of Appraising Firepower-might of Tank-gun Based on Back-Propagation Network

作　　者：王永彬[1] 徐大杰[1] 侯胜高[1] 

机构地区：[1]蚌埠坦克学院,安徽蚌埠233013

出　　处：《指挥控制与仿真》2009年第3期68-71,共4页Command Control & Simulation

摘　　要：从分析不同型号坦克炮的主要性能指标入手,应用BP神经网络理论和方法探索其火力威力评价的问题。建立了坦克炮火力威力评价的BP神经网络模型,对原始评价指标进行了归一化处理,并构造了3组训练样本,利用MATLAB6.5中神经网络工具箱的图形用户界面GUI对评价指标和样本进行训练、仿真,并对结果进行比较。结果表明,BP神经网络模型能很好地解决评价指标与结果之间的非线性关系,评价坦克炮火力威力简便可靠,客观性强;研究结果可为坦克炮的进一步开发和研制提供理论参考。This paper proposes a new way to analyze the main function index of different Tank-guns, and applies the theories and methods of BP Neural Network to solve the problem of appraising Firepower-might of Tank-gun. The model of appraising Firepower-might of Tank-gun is established and based on Back-Propagation Network, the original data is dealt with Return a turn, and the model is exercised using the three date. By the GUI in Nerve Network Tool boxes in MATLAB6.5, the indexes and samples of valuation are trained and simulated, and the results are compared. Results indicate that the model of Back-Propagation Network can solve the no -linearity relations between the indexes and the results. It can be simple and credible to evaluate the Firepower-might of Tank-gun and it is impersonal. The research outcome can provide the theory preference for developing and researching Tank-guns of new level.

关 键 词：坦克炮 火力威力 评价 BP神经网络 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]