基于BP神经网络固冲发动机燃气流量调节控制

GA-BP Neural Network Control for Ducted Rocket Gas Regulating System

在针阀式固冲发动机燃气流量调节优化控制问题的研究中,为优化燃气流量调节,建立了针阀式燃气流量调节燃气发生器工作过程中的动态模型,分析了动态模型的频域特性,并通过仿真进行了验证。针对燃气流量调节系统是一非最小相位系统,具有负调特性,设计了一种遗传算法优化BP神经网络PID控制器。仿真结果表明所提出的控制方法可以有效减小固体火箭冲压发动机燃气流量调节的负调峰值和超调量,缩短系统调节时间,有效地提高了固冲发动机的工作性能。

The control method of the Needle-Type gas regulating system for a ducted rocket was studied in this article. Firstly, the dynamic model of the Needle-Type gas regulating system for the ducted rocket was founded and the frequency-domain characteristics was analyzed. Then a neural network PID controller with back propagation was pro- posed, the weights of neural network were optimized with Genetic Algorithm to achieve the optimal combination property. The simulation results show that the discrete model can be directly used in computer simulation, the GA-BP Neural Network controller can eliminate the undershoot and reduce the overshoot and the adjusting time.