将二次型性能指标引入单神经元并利用自学习功能构成了自适应 Ｐ Ｉ Ｄ 控制器；分析了神经元自适应控制系统的稳定性和学习算法的收敛性．仿真结果表明，这种控制方案具有良好的自适应性． The quadratic performance index is introduced into the single neuron and self-learning function is used to construct the adaptive PID controller. The stability of the neuron adaptive control system and the convergence of the learning algorithm are analyzed. Simulation results show that this control scheme has good adaptability.