Real-time optimization method for thrust performance of ramjet

The invention discloses a real-time optimization method for the thrust performance of a ramjet. The method comprises the following steps: constructing a multi-dimensional optimization space according to the position of each air inlet of the ramjet engine and the incident angle of the air inlet, constructing a sample through a Latin hypercube design method, calculating the thrust through a CFD method, constructing an RBF neural network approximate prediction model, using a leapfrog algorithm for optimization, and calculating an individual fitness value during global search by using an RBF neural network. The position and the incident angle of the air inlet of the engine can be adjusted in real time under the actual complex working condition, and the thrust of the ramjet can be optimized in real time in the actual operation process of the engine. 本发明公开了一种冲压发动机推力性能的实时优化方法，包括以冲压发动机各进气口的进气口位置以及进气口入射角度构建多维度优化空间，采用拉丁超立方设计方法构建样本，以CFD方法计算推力并构造RBF神经网络近似预测模型，采用蛙跳算法优化，在全局搜索时以RBF神经网络计算个体适应度值。本发明使发动机在实际复杂工况下实时调整进气口的位置与入射角度