基于自适应回归算法的导弹表面温度预测

为预测导弹高速飞行时由于气动加热而升高的表面温度,根据已知的物理模型和试验数据,建立了一种基于自适应回归算法的导弹表面温度预测模型,使用该模型预测出温度数据,然后与实际试验数据进行对比。抽样选取含有55个个体的样本,按个体时间顺序排列样本个体,使用样本中的前50个个体训练预测模型,再使用剩余5个个体对预测模型进行检验。在前50个个体的相应时刻,预测值比实际值平均减小1.24%,标准差为1.27%; 在最后5个个体的相应时刻,预测值比实际值平均减小1.42%,标准差为0.16%。结果表明,采用基于自适应回归算法的导弹表面温度预测模型对导弹的表面温度进行预测具有较高的精度,达到了预测导弹表面温度的目的。

Prediction on Surface Temperature of Missile Based onAdaptive Regression Algorithm

When the missile flies at high speed,the surface temperature rises due to aerodynamic heating.In order to predict the surface temperature of missile,a prediction model of missile surface temperature based on adaptive regression algorithm was established according to the known physical model and test data.The model was used to predict the temperature data,which were compared with the experimental data.A sample of 55 individuals was selected,and the individual samples were arranged according to the time order.The first 50 individuals were used to train the prediction model,and then the remaining 5 individuals were used to test the prediction model.At the corresponding time of the first 50 individuals,the predicted value was reduced by 1.24%,and the standard deviation was 1.27%.At the corresponding time of the last 5 individuals,the predicted value was reduced by an average of 1.42%,and the standard deviation was about 0.16%.The results show that the surface-temperature prediction-model of missile based on adaptive regression algorithm has high precision to predict the surface temperature of missile,and achieves the purpose of predicting the surface temperature of missile.