时变大纯滞后系统的单神经元预测控制

为对时变大纯滞后系统实现快速有效的实时控制 ,将单神经元与 L evinson预测器相结合采用比例积分微分调节器 (PID)控制方式 ,设计了单神经元预测控制器 ,其权值和可整定参数能够在线自适应调整 ,克服了大滞后对象控制结果不能及时反馈的不足。应用该控制策略对大滞后一阶和二阶对象的仿真研究表明 ,对大滞后时变系统具有很强的适应性和鲁棒性 ,各种控制性能明显优于常规 PID等其它控制

Single-neuron predictive control for time-variable large delay systems

A real time control method was developed for time variable large delay systems. A predictive controller based on a single neuron was designed combining a single neuron with the Levinson predictive mechanism to build a proportion integral differential (PID) controller. This controller's weights and variable parameters were adaptively adjusted on line, to minimize the control signal delay. Experiments with single order and two order delay systems, verified that the controller processes excellent self turn and robustness and that it's overall control performance is beffer than the PID controller.