基于优化最小二乘支持向量机的襟翼趋势预测

将最小二乘支持向量机(LS-SVM)应用于飞机襟翼状态趋势研究。首先,通过分析飞机襟翼故障与襟翼动作耗时参数的关系,提出了利用动作耗时趋势来确定该系统未来状态的方法。然后,使用最小二乘支持向量机建立耗时回归预测模型,采用最终预报误差(FPE)准则确定回归模型嵌入维数,提出了自适应网格搜索法,优化最小二乘支持向量机的建模参数,从而实现比现有方法精度高,泛化性能好的预测模型。训练和测试样本取自飞行数据记录系统(QAR)中译码襟翼参数值。与神经网络模型的比较实践表明,该方法具有实用价值。

Trend Prediction of Aircraft Flap System Based on Optimized Least Square Support Vector Machine

Least square support vector machine(LS-SVM) is used to predict the trend of the aircraft flap system.Firstly,by analyzing the relationship between time-consuming of the flap action and the fault of flap system,a model based on the time-consuming trend of the flap action is used to predict the trend of the flap system.Secondly,LS-SVM is used to establish the time-consuming model.The final prediction error(FPE) principle is used to optimize the embedded dimension.Then,an adaptive grid search(AGS) method is proposed to optimize the model parameters of LS-SVM.The training and measuring data set is obtained from quick access recorder(QAR).Finally,the one-step and n-step prediction results of LS-SVM and RBF neural network are compared.The method shows that LSSVM regression model is feasible and accurate for the trend prediction of the flap system.