大射电望远镜舱索系统的控制与实验

Control and Experiment for Large Radio Telescope

大射电望远镜(LT)馈源舱的大范围空间扫描运动是通过6根大跨度的绳索收放来完成的。针对系统中的索结构之间强耦合以及模型变结构的特点,推导了输入索和输出索之间的关系矩阵的递推公式,并证明了递推公式的稳定性和收敛性,在此基础上提出了一种具有神经网络非参数模型自适应控制的方法,最后通过LT50m室外模型实验证明了该算法的有效性。

In large radio telescope （LT） project, six large span cables were employed to tune the cable--cabin--supporting system to implement the cabin＇s trajectory tracking. Considering the special characteristics of the strong coupling among cables, a recursively formula between input length a/ld out length of cables was deduced based on none parameter adaptive theory. The stability and convergence of the formula were established respectively. Then an adaptive control algorithm with neural network was presented. Practical LT50m model experimental results were executed to demonstrate the method validity.