针对二自由度机械手的控制问题，本文提出了一种基于函数联接神经网络的反馈学习控制器，提高了控制系统的自适应速度，改善了系统的起始特性．为了在改善系统自适应速度的同时增强系统的稳态特性，本文又提出了基于函数联接神经网络的串级学习控制算法．仿真和实验都证明了控制方法的优越性． In order to solve the control problem of manipulator with two degrees of freedom, this paper presents a feedback learning controller based on functional neural network, which improves the adaptive speed of the control system and improves the initial characteristic of the system. In order to improve the system’s steady-state characteristics while improving the adaptive speed of the system, this paper also proposed a cascade learning control algorithm based on function-linked neural networks. Both simulation and experiment prove the superiority of the control method.