论文部分内容阅读
对于包含人为因素飞行员模型的人机系统,传统针对确定模型的控制器方法无法完全满足飞行轨迹精度。针对飞行员模型在环系统,提出了采用衰减记忆方法、线性代价递增切换算法的去伪PI控制器设计方法。去伪控制理论是一种仅利用模型输入输出数据的无模型数据驱动控制理论。在去伪控制理论中,监视器用于选择最优控制器,控制器切换算法负责在需要时进行控制器切换。衰减记忆法的引入可减弱历史数据对切换判断的影响。对特定飞行场景进行了蒙特卡罗仿真,仿真结果表明,设计的控制器可减小飞行任务中的高度轨迹偏差,提高飞行品质。
For the man-machine system including the human-factor pilot model, the traditional controller method for determining the model can not fully satisfy the flight path accuracy. Aiming at the pilot system in the loop, a de-pseudo PI controller design method using attenuation memory method and linear cost incremental switching algorithm is proposed. Pseudo-control theory is a model-free data driven control theory that uses only the model input and output data. In de-dummy control theory, the monitor is used to select the optimal controller, and the controller switching algorithm is responsible for switching the controller when needed. The introduction of attenuated memory reduces the impact of historical data on the switching decision. The Monte Carlo simulation of a specific flight scenario is carried out. The simulation results show that the designed controller can reduce the altitude trajectory deviation and improve the flight quality.