神经网络PID的无人动力伞偏航控制研究

Study of Neural Network PID in Heading Control of Unmanned Powered Parachute Aircraft

对于具有非线性、时变和强耦合特性的动力伞系统,采用常规控制方法难以获得满意的控制效果,而采用神经网络与传统PID相结合作为控制器,使得所设计控制器能够针对非线性、时变的气动特性自适应地调整控制参数,实现对PID控制器参数的寻优整定,简化了设计过程。利用Matlab软件对系统进行仿真,给出了相应的仿真结果。

For the unmanned powered parachute system with nonlinear time-varying and strong coupling characters, the conventional PID eontrller is difficult to achieve good control performance. Therefore, a combination of a BP neural network and general PID controller is used in autopilot of unmanned powered parachute system. This algorithm makes the designed autopilot be able to adaptive adjust control parameters with respect to the time varying and nonlinear aerodynamic characteristics, realizes the setting of PID controller parameter and makes the design processing simplely.