【摘要】 四轴飞行器是一个非线性、时变的控制系统,由于典型PID算法缺乏自适应能力,典型的PID控制算法对四轴飞行器无法达到较好的控制效果,于是提出了基于BP神经网络的智能PID控制算法。本文首先通过选取适当的坐标系统对旋翼受力进行了近似和简化处理建立其数学模型,根据得到各通道的传递函数,然后使用BP神经网络PID控制方法对四轴飞行器进行控制,实现PID参数自动调整,结果表明BP神经网络PID控制对非线性系统及其参数具有良好的控制效果,而且具有更好的适应性,同时也具有很好的逼近和容错能力。更多还原

【Abstract】 Quadrocopter is a nonlinear, time-varying control system. The typical PID control algorithm can’t achieve a good control effect due to the lack of adaptability. Then this work proposed the intelligent PID control algorithm based on BP neural network. Firstly, the force of the rotor was simplified to establish the mathematical model by choosing an appropriate coordinate system and obtain transfer functions for each channel based on the formula. Then by using the BP neural network PID control algorithm to control the quadrocopter, it can achieve automatic adjustment in PID parameters. The results show that BP neural network PID control algorithm has good control effect on the nonlinear system and parameters, not only adaptability and robustness, but also approximation and fault tolerance.更多还原