为实现民机液压系统的健康预测,将贝叶斯框架应用于LS-SVM参数的选优.选用径向基核函数,选择了预测回归模型的算法和区间预测公式.用训练样本建立了液压系统的健康预测模型,用测试样本验证了公式的有效性;对液压数据参数进行预测,将预测值带入健康评估模型中得到预测结果.结果表明,基于贝叶斯框架下的LS-SVM回归模型可以很好地用于民机液压系统的健康预测.

In order to predict the health of civil aircraft hydraulic system, based on the LS-SVM regression under Bayesian evidence framework model, the Radial Basis Function (RBF) kernel is used. The LS-SVM regression parameters are selected and tuned, the model algorithm and interval prediction formula are selected, and then the prediction model of hydraulic system is built by the train sample, the validation of formula is verified by the test sample. The parameters of hydraulic system are predicted, and then the predicted values are put into the health assessment model. The results show that the LS-SVM regression under Bayesian evidence framework model is able to be well applied to the health prediction of civil aircraft hydraulic system.