摘要
装备保障演练是部队装备保障训练的高级形式,是提升装备保障能力的有效途径,装备保障演练综合评价是检验和衡量部队装备保障演练效果和水平的主要手段。装备保障演练评价指标体系多维,导致利用神经网络方法进行评价时效性差、评价精度低,为解决上述问题,提出了主成分分析和BP神经网络相结合的装备保障演练综合评价方法。利用装备保障能力,建立了装备保障演练评价的指标体系。从降低评价指标体系维度的角度出发,建立了装备保障演练综合评价模型,以主成分分析降低指标体系维数,进而以神经网络训练样本数据实现了装备保障演练的综合评价。仿真结果验证了改进方法的正确性、有效性和优越性,为装备保障演练评价提供了一种有效的方法。
Equipment support exercise is the highest phase of equipment support training and an effective ap- proach to improve equipment support capability. Assessment of equipment support exercise is the main measure to test and rate the effect and level of equipment support exercise. In order to overcome the deficiency of low assessment precision using BP neural network because of various indexes, the comprehensive assessment method of equipment support exercise was proposed based on the hybrid algorithm of principle component analysis (PCA) and BP neural network. The index system of equipment support exercise was established from the viewpoint of equipment support ca- pability. The comprehensive assessment model was established from the viewpoint of reducing the dimension of index system. New index system with reduced dimensions was achieved using PCA method. Assessment result which is consistent with the actual assessment level was achieved through training the proposed model using proper data sam- ples. The rightness and effectiveness of the proposed method are verified by simulation result, which provides an ef- fective method for equipment support exercise assessment.
出处
《计算机仿真》
CSCD
北大核心
2013年第10期18-22,共5页
Computer Simulation
关键词
主成分分析
装备保障演练
神经网络
Principle component analysis (PCA)
Equipment support exercise
Neural network