基于支持向量分类的装备完好性预测模型

对抗的战场环境和任务的变化，越来越需要装备战备完好性来保障作战行动。保持或提高装备战备完好性是装备保障的核心和中心工作。利用基于结构风险最小化的支持向量分类（Support Vector Classification，SVC）方法对装备的战备完好性进行了预测，提高了机器学习方法的预测能力。并以车辆装备发动机的技术状况数据为实例，建立了预测模型，通过参数选择，提高了模型预测的正确率、命中率等指标。结论表明：支持向量分类方法是预测装备战备完好性的有效方法。

Forecasting Model of Equipment Readiness Based on SVC

Battle field environment of oppositions and changed missions, increasingly depend on equipment readiness for military operations. Keeping or improving the readiness is the key and core task in equipment ILS （Integrated Logistic Support）. In this paper, SVC （Support Vector Classification） with SRM （Structure Risk Minimization） is used to forecast readiness and sustainable capability, which can be improved by machine learning. The status parameters of armored vehicle engine are used as an example to establish a forecast model to improve the indexes such as correctness and target-hitting rates through preferences. The conclusion shows that SVC is an effective method to forecast the equipment readiness.