针对自抗扰控制~(ADRC) 多参数不易整定 的问题, 提出了一种基于二次函数的非线性~PID (NLPID) 控 制律. 该方法用二次函数模拟~PID 增益参数随误差变化的规律 曲线, 构造了一个非线性~PID 神经网络模型, 利用最速下降法对各模拟曲线的系数进行在线调整, 实现了基于神经网络的自适应 自抗扰控制. 仿真结果表明, 与常规~ADRC 控制方法相比, 文章方法减少了~ADRC 需调整的参数, 并具有较好的控制效果.

Considering the problem of multi-parameter to be tuned in the ADRC, a nonlinear PID control law based on quadratic function is presented in this paper. This method uses quadratic functions to simulate curves of the PID parameters following error changes and constructs a model of nonlinear PID neural network, adjusting coefficients of simulation curves online by steepest descent method and the adaptive ADRC based on neural network is realized. The simulation results show that the approach reduces the parameters of ADRC to be adjusted compared with conventional method and has better control effect.