为了提高飞机调姿的运动安全并减小在误差敏感方向的冲击,采用Newton-Euler方法构建了调姿系统的动力学模型,以实现调姿轨迹的优化。该动力学模型综合考虑了支撑杆变形、驱动丝杆变形及运动误差的影响,可对比不同轨迹时部件运动全过程的运动学特性,实现多目标多约束条件下的轨迹优化。为了提高计算效率,提出了一种类间可分性最优的自适应核主成分算法进行特征提取,并结合模式识别中的自动分类方法,预判可行轨迹的性能,控制搜索范围,减少寻优过程中的计算量。以某型数控定位系统为例,在对150条可行调姿轨迹进行评价和优选后,大部件调姿过程的最大平动速度小于20mm/s,调姿结束时的最大角速度小于0.1rad/s,说明了该方法的可行性和有效性。
To improve the safety and reduce the impact on the sensitive direction in an aircraft component trajectory adjustment system,the trajectory of a large component is optimized by a dynamics model of the adjustment system.This model is derived by the Newton-Euler method,and takes into account the rod deformation,screw movement error and deformation in the NC locator.It can evaluate the kinematic characteristics of the component with different movement trajectories.In order to improve the efficiency and reduce the amount of computation in the optimization process,an adaptive kernel principal component analysis algorithm with optimal separability between classes is proposed in combination with pattern recognition methods to predict the performance of feasible trajectories to control the scope of the search.The effectiveness of the proposed method is verified by an example of an adjustment system.After evaluation of 150 feasible trajectories,the maximum translational velocity of the large component during the adjustment process is less than 20mm/s,and the maximum angular velocity at the end of the adjustment process is less than 0.1rad/s.