RBF网络用于边界层转捩中抽吸流优化控制

 在抽吸气流控制边界层转捩问题中 ,将径向基神经网络用于抽吸流速与转捩位置间的函数关系建模 ,构造了网络结构 ,利用一组两通道抽吸流控制转捩的实验数据训练网络 ,获得了优化的网络参数。在此基础上 ,利用训练网络所获得的输入 /输出关系模型求解了最优抽吸流速。结果表明 ,RBF网络可有效地应用于边界层转捩主动控制中的系统函数关系建模。

PLANT IDENTIFICATION IN ACTIVE CONTROL OF LAMINAR BOUNDARY-LAYER TRANSITION BY SUCTION USING RBF NEURAL NETWORK

A Radial Basis Function (RBF) neural network is applied to plant identification for active control of laminar boundary layer transition by suction. A suitable RBF structure is selected and its optimal parameters are obtained by training a network with real experimental data from a two channel suction system. The plant model from the trained network, which represents the plant response, can be used successfully to solve the optimal suction flow rates instead of using an assumed input/output function. Simulation results show that the RBF neural network is an effective tool in plant identification for the nonlinearly constrained optimization problem of laminar flow control.