论文部分内容阅读
针对提高逆系统建模中神经网络的逼近效果和动态性能问题,根据PID神经元网络工作原理,提出一种具有动态激励函数的新型PID神经元模型—输出反馈型PID神经元(OFPID),输出激励采用连续的Sigmoidal函数,使神经元具有等效的IIR突触,采用梯度下降法实现OFPID神经元网络的权值调整,将其应用于非线性系统的神经网络逆控制系统,从而提高非线性系统的解耦效果和控制性能。仿真实验证明,提出的新型神经元网络是一种良好的非线性系统建模和控制工具。
In order to improve the approximation effect and dynamic performance of neural network in inverse system modeling, a novel PID neuron with output function (OFPID) is presented according to the working principle of PID neural network. The output The continuous sigmoidal function was used to make the neurons have the equivalent IIR synapse, and the gradient descent method was used to adjust the weight of the OFPID neural network. This method was applied to the neural network inverse control system of nonlinear system to improve the nonlinearity System decoupling effects and control performance. Simulation results show that the proposed new neural network is a good nonlinear system modeling and control tool.