摘要
航空兵部队成建制空运转场飞行架次需求预测,对机关、航空兵部队拟定空运计划、进行空运准备等都具有重要的意义。运用序列后向选择方法(SBS)对影响飞行架次的特征因素进行逐层淘汰,利用支持向量机(SVM)理论建立单因素非线性回归模型,进而对飞行架次进行预测。预测结果表明:同多因素SVM预测模型相比,单因素SVM预测模型虽在预测精度上没有显著提高,但其减少了预测的前期工作量,方便了机关和部队的使用,实现了飞行架次预测的实时性要求。
Sortie requirement prediction of military transporter for air arm aircraft ferry is beneficial to constitute an aircraft ferry plan and prepare for the aircraft ferry.Features which affect the sortie were eliminated through Sequential backward selection（SBS）,the single feature nonlinear regression model based on Support Vector Machine（SVM） was established,then the sortie requirement was predicted.The result of prediction shows that although the method of singe feature SVM doesn？？t have better result compared with Multiple feature SVM,it reduces the work before prediction,promotes the army＇s convenience and raises the speed of prediction.
出处
《飞机设计》
2010年第6期62-65,共4页
Aircraft Design
关键词
空运
序列后向选择法
支持向量机
架次预测
Air transportation
Sequential backward selection
Support vector machine
Sortie requirement prediction
