分数阶非线性系统的自适应预定义时间优化控制OA

For a class of fractional-order nonlinear systems with quantized inputs and unknown external disturbances,a disturbance observer-based adaptive predefined-time command-filter self-triggered optimal control scheme is proposed.First,a predefined-time filter is introduced and compensation signals are designed to overcome the computational complexity of the traditional backstepping method.Neural networks are employed to approximate the unknown nonlinear dynamics of the system,and a fractional-order nonlinear disturbance observer is constructed to accurately estimate the lumped disturbance,including neural network approximation errors and unknown external disturbances.Based on the backstepping framework and reinforcement learning strategy,auxiliary variables are introduced and an actor-critic network is designed to solve the Hamilton-Jacobi-Bellman equation,yielding the optimal virtual controller and achieving optimal control.Meanwhile,in the presence of the quantized input,a self-triggered strategy is further introduced to reduce communication resource consumption.According to fractional-order predefined-time stability theory,the proposed scheme ensures that tracking errors converge to a small neighborhood of the origin within a predefined time,and all signals in the closed-loop system remain bounded.Finally,simulation results verify the proposed method's efficacy and superiority.