摘要
随着机载云台技术的快速发展,搭载稳定云台的无人机具备快速目标搜索和跟踪能力。由于云台常在复杂环境下执行任务,不可避免地会受到风阻力、机械振动、负载扰动等多种因素影响。传统的云台控制系统采用经典的PID控制,但在复杂的环境中,由于外扰突变,传统的PID不能跟踪突变量的变化,造成控制系统控制精度下降和响应速率减慢。设计基于ARM的硬件平台,运用神经网络云台控制技术,在方向恒定、风速5.95 m/s情况下,云台的姿态角度误差在±5°以内,能够实现云台的稳定控制。
With the rapid development of airborne pan-tilt-zoom (PTZ) technology, the steady PTZ can make UAV have fast target search and tracking capabilities. As the PTZ often performs tasks in complex en- vironments, it will inevitably be affected by wind resistance, mechanical vibration, load disturbance and other factors. The traditional PTZ control system adopts the classical PID control. However, in complex environ- ments, the traditional PID cannot track the change of the mutation quantity due to the external disturbance, which results in decreased control accuracy and slow response speed. The design in this paper is based on ARM hardware platform and uses the neural network PTZ control technology. In the case of constant wind direction and wind speed 5.95 m/s,the stable control of the PTZ can be achieved in the design with the PTZ attitude angle error within plus or minus five degrees.
出处
《沈阳航空航天大学学报》
2017年第3期70-75,共6页
Journal of Shenyang Aerospace University
基金
辽宁省高等学校优秀人才支持计划第二层次(项目编号:LJQ2014017)