未知不对称死区补偿方法在三轴微机电陀螺控制中的应用

An unknown asymmetry dead-zone compensation method for MEMS triaxial gyroscope control design

微机电系统(Micro-Electro-Mechanic System,MEMS)陀螺仪由于微机械加工误差、微米量级的尺度效应和驱动原理上的非线性等因素,驱动梳齿不可避免地受到死区效应影响。针对上述问题,提出基于自适应权值神经网络和比例微分滑模的复合控制方法,分别采用两个RBF神经网络逼近死区的逆和经过驱动梳齿后的实际控制力,以弱耦合的权值自适应律降低控制复杂度,利用滑模控制器的鲁棒性消除两者总的逼近误差,实现微机电陀螺三轴的严格轨迹跟踪。最后采用Lyapunov直接法证明了全局滑模面的可达性和弱耦合自适应律的收敛性。仿真结果表明了该复合控制器的有效性。

Under machining tolerances,due to size effect of micrometer range and nonlinear driven principle,the mechano-electric transducers of Micro-Electro-Mechanic System(MEMS)gyro could be affected by dead-zone effect.This paper presents a composite controll method combined adaptive neural networks and proportional-derivative sliding mode controller to compensate the dead-zone effect.Two adaptive RBF neural networks with a weak-coupling weights adaptive law are utilized to approximate the inverse of dead-zone and the unmeasurable actual control force,and the sliding mode controller is designed to eliminate the total approximation error of two RBFs.Finally,the reachability of sliding mode surface and the convergence of weak-coupling weights adaptive law are analyzed and proved by Lyapunov direct method.Simulation results show the validity and effectiveness of proposed composite controller.