基于改进遗传算法的四旋翼无人机参数辨识

摘要

为克服传统遗传算法在参数辨识过程中收敛速度慢的问题，提出了一种基于改进遗传算法的四悬翼无人机参数辨识方法。该方法引入梯度算子为遗传进化提供指示性的方向，利用遗传算法的全局搜索性保证算法的全局收敛，并根据简化四旋翼无人机数学模型设计了优化函数。利用四悬翼无人机实飞数据进行了实验测试，实验结果验证了本文方法的有效性和快速收敛性。%In order to work out the low convergence speed problem in the process of parameter identification in which the genetic algorithm is adopted as the main section, a new method of parameter identification process is developed for the quadrotor UAV model, which is constructed based on an improved method of the GA process. This method introduced a gradient operator into the GA process to offer a sort of direction for algorithm evolution. The global convergence of the algorithm is guaranteed by the global search property of the GA and an optimal function is designed with the simplified quadrotor model. A test with the data gained from the actual flight experiment of the quadrotor is made and results proved the effectiveness and rapid convergence property of this method.