关键词: 非线性系统, 神经网络, 自适应控制, 离散Nussbaum增益, PID控制

Abstract: For a class of single-input-single-output nonlinear discrete-time systems with unknown control direction, an adaptive neural network control was developed by incorporating a conventional incremental digital PID (proportional-integral-derivative) controller in an adaptive neural network term to guarantee the stability of the closed-loop systems. The conventional PID controller was utilized to stabilize the approximate linear system, while the adaptive neural network was introduced to deal with the influence of nonlinear terms on closed-loop systems. A discrete Nussbaum gain was introduced into the adaptation law of the weights in neural network to resolve the unknown control direction problem. It was proved that all signals of the closed-loop system are bounded with the tracking error converges on a compact set. Simulation results verified the effectiveness of the proposed control method.