论文部分内容阅读
结合非线性优化理论和方法提出了易于实现、收敛速度比较快的多层神经网络共轭梯度反传算法。液体火箭发动机参数辨识技术已得到广泛的应用,由于传统的数学方法必须基于发动机已知模型,使得其参数辨识受到极大的限制。文中基于神经网络共轭梯度反传算法进行液体火箭发动机的系统辨识,结合变推力发动机热试车动态数据,得到了满意的仿真结果。
Combined with the theory and method of nonlinear optimization, a multi-layer neural network conjugate gradient backpropagation algorithm is proposed which is easy to implement and has a fast convergence rate. Liquid rocket engine parameter identification technology has been widely used, because the traditional mathematical methods must be based on the known model of the engine, making its parameter identification is greatly limited. In this paper, the system identification of liquid propellant rocket engine based on neural network conjugate gradient backpropagation algorithm is combined with the dynamic test data of variable thrust engine, and the satisfactory simulation result is obtained.