The Research on Falling Position Prediction Method of Long Range Guided Rocket Based on Surrogate Model

A novel strategy of predicting falling position based on surrogate model is proposed for meeting the urgent needs of quick and accurate prediction of falling position for long range guided rocket. Through applying surrogate model methodology, this method is established a highly nonlinear functional relationship between falling position, ballistic parameters and launch elements, which relationship is used to predict quickly and accurately for guided rocket’s falling position in any launch conditions. Under the comprehensive consideration of flight status parameters and launch elements that influence accuracy of prediction of falling position, PRS, Kriging and RBF are implemented respectively in the research of falling prediction during climbing section and falling section of the guided rocket. Finally, numerical simulation compares and analyzes the accuracy and rapidity of various surrogate models, the effectiveness and practicality of the method are demonstrated.