基于支持向量机的舰炮自动弹库方案决策研究

摘要

针对某大口径舰炮自动弹库的多种方案决策问题,提出了基于样本特征加权支持向量回归机的自动弹库方案决策模型.分析了舰炮自动弹库系统方案选择的要求和原则,对系统决策因素集进行模糊量化,构造了舰炮自动弹库决策模型的输入样本.采用舰炮领域专家知识结构信息构造决策因素的权重向量.通过对系统决策因素集和决策方案非线性关系的辨识,建立了基于特征加权核函数的支持向量回归机的舰炮自动弹库方案决策模型.通过应用实例说明了利用支持向量回归机建立决策模型的决策方法有效性和实用性.%Aimed at the decision problem of many schemes on a certain large caliber naval gun automatic magazine, a decision model based on samples feature weighted support vector regression was proposed. The demands and principia for scheme selection were analyzed. In addition, the system decision factor sets were fuzzily quantized, and the input samples of the decision model were structured. Some domain experts' knowledge structure information was used to build the weight vector of decision factor. By means of nonlinear relationship identification for the system decision factor sets and schemes, the decision model founded on the feature weighted kernel function support vector regression was built. The simulation and experiment results show that the decision model based on the support vector regression is effective and practicable.