基于差分进化支持向量机的作战效能评估方法
Evaluation Method for Operational Effectiveness Based on Support Vector Machine with Differential Evolution
杨健为 1徐坚 1吴小役 1鲁玉祥 1魏继卿1
作者信息
- 1. 西北机电工程研究所，陕西咸阳 712099
- 折叠
摘要
武器系统作战效能的评估具有重要意义。针对作战效能评估过程中影响因素复杂、小样本、非线性等问题，引入基于最小二乘法的支持向量机回归算法，用于作战效能的学习与预测。为了提高预测精度，引入差分进化算法进行支持向量机的参数优化选取。以地地导弹武器系统效能为例，分别采用 BP神经网络算法、经典支持向量机算法与本文算法进行仿真计算，结果表明差分进化支持向量机算法可很好地实现武器系统作战效能评估，具有较好的计算精度。
Abstract
It is of great significance to evaluate the operational effectiveness of weapon system. With regard to the problems such as complicated impact factors,small sample and nonlinearity,the least square support vector machine was used to study and predict the evaluation of operational effectiveness. When establishing the parameters of SVM,Differential evolution was introduced to enhance the accura-cy of the prediction. With the evaluation of surface to surface missile as an example,this new method, BP neural network and traditional SVM were used for simulation calculation. It is shown that the meth-od of LS-SVM with Differential Evolution works well for the evaluation of operational effectiveness with a better computing accuracy.
关键词
作战效能/支持向量机/差分进化算法:BP神经网络Key words
operational effectiveness/support vector machine/differential evolution/BP neural net-work引用本文复制引用
出版年
2016
NETL
NSTL
维普
万方数据