一种神经形态马赫数PID自适应控制器研究

本文从自适应控制和优化学习算法的角度出发，将传统的PID控制和神经网络BP算法相结合，构造了具有自适应、自学习功能的神经形态PID自适应马赫数控制器(NNCPID)。通过对网络控制模型的训练、学习和系统仿真，验证了该NNCPID控制器在改善马赫数控制性能、提高系统实时性、鲁棒性等方面，具有很大的优越性。

Research of Self-Adaptive PID Mach Controller Based on Neural Network

From the point of self-adaptive control and optimizing study algorithm, this thesis combines traditional PID control and BP neural network algorithm together to construct an self-adaptive neural network Mach PID controller which is called NNCPID. The topology structure of NN is constructed, the NN model is trained and a series of simulation experiment are done. It is proved that NNCPID is capable of improving the dynamic, real-time and robust character of Mach controller.