基于BP神经网络的三轴转台偏航控制

摘要

The three-axis flight motion simulator table is a nonlinear and strongly-coupled system. With its research focus on the servo control system of the three-axis flight motion simulator table and the mathematical model of the brushless DC motor which is used as the drive mechanisms of the rotary table,make an comparison between the traditional PID control arithmetic and NN PID control way arithmetic based on model forecast. MATLAB simulation is conducted to test stepping and slop input signals. Simulation result indicates that this method can get satisfied control result, and it has the capability of better self-adaptability and robustness.%三轴飞行仿真转台系统作为一种非线性、强耦合系统,难以用精确的数学模型进行描述.文中以三轴飞行转台系统中的回转转台伺服控制系统为研究对象,在分析了回转转台的驱动机构一直流无刷电机的数学模型的基础上,对常规PID控制算法和BP神经网络PID控制算法进行了比较分析.对BP神经网络算法和计算流程进行了简介.同时对阶跃和正弦信号进行MATLAB仿真.仿真结果表明,该方法可以实现有效的控制,并且与传统PID算法相比,具有更好的自适应性和鲁棒性.