考虑作动器输出饱合的光电平台终端滑模神经网络控制

Sliding mode neural network control of photoelectric platform terminal considering actuator output saturation

光电惯性稳定平台被广泛的应用于无人系统移动载体上,针对两轴光电平台系统参数不确定与外界扰动问题,提出自适应终端滑模神经网络控制算法对光电平台伺服系统进行稳定跟踪控制。利用径向基神经网络估计平台动力学系统中的未知非线性函数,同时,考虑无刷直流力矩电机在实际应用中的输出饱和特性,引入辅助函数用以补偿理想控制力矩和实际输出力矩之间的误差,提高了光电负载图像的稳定性与动态目标跟踪的快速性。最后,通过李亚普诺夫原理验证了闭环系统的稳定性与渐进收敛性。外界随机振动试验结果表明,提出的终端滑模神经网络控制算法减振效果陀螺均方根值为4.7 mrad/s,比传统的滑模控制的减振效果提升了8%,相比于传统的PID控制提高了13.3%。所提出的控制方法能够较好抑制移动载体对光电设备的扰动,抗干扰能力强。

Photoelectric inertial stabilization platform is widely used in mobile carrier of unmanned system.Here,aiming at the problem of parameter uncertainty and external disturbance of dual-axis photoelectric platform system,an adaptive terminal sliding mode neural network control algorithm was proposed to do stabilization and track control for servo system of photoelectric platform.Radial basis function neural network was used to estimate unknown nonlinear function in platform dynamic system.Meanwhile,considering output saturation characteristics of brushless DC torque motor in practical application,auxiliary function was introduced to compensate error between ideal control torque and actual output torque,improve stability of photoelectric load image and rapidity of dynamic target tracking.Finally,stability and asymptotic convergence of closed-loop system were verified using Lyapunov principle.The external random vibration experiment results showed that the proposed terminal sliding mode neural network control algorithm has a vibration reduction effect to make gyro have RMS value of 4.7 mrad/s,it is 8% higher than the damping effect of the traditional sliding mode control and 13.3% higher than that of the traditional PID control;the proposed control method can better suppress disturbance of mobile carrier to photoelectric equipment,and have a strong anti-interference ability.