MEMS陀螺阵列的RCC-OBE估计融合方法

为提高微机电系统（MEMS）陀螺的精度，提出一种基于松弛Chebyshev中心（RCC）的最优定界椭球（OBE）算法，并用于陀螺阵列信号的融合。以单个陀螺误差输出模型为基础，建立了阵列系统的机动融合模型；由于噪声统计特性的不确定会导致传统融合方法精度下降，引入仅要求噪声未知但有界的集员估计理论，运用OBE算法实现角速率信号的稳健估计；在OBE算法中，往往采用椭球几何中心作为真实值的点估计，但该中心并没有理论上的最优特性，而可行集的Chebyshev中心具有很多优良特性，同时，考虑到准确的Chebyshev中心求解十分困难，转而求解可行集的RCC，作为速率信号的点估计，设计了以RCC作为输出的OBE更新过程和新的参数优化准则。采用6个陀螺构成的阵列进行了验证试验，结果表明基于该算法的阵列估计融合方法在获得角速率保证边界的基础上，可以进一步提高MEMS陀螺精度。

RCC-OBE estimation fusion approach for MEMS gyro array

In order to improve the accuracy of micro-electro-mechanical system (MEMS) gyro, an optimal bounding ellipsoid (OBE) algorithm based on relaxed Chebyshev center (RCC) is proposed and used to fuse gyro array signals. On the basis of the error model of single gyro, the maneuvering fusion model of the array system is established. Because of the uncertainty of the noise statistics, the accuracy of the traditional fusion method is reduced. The set-membership estimation theory with unknown but bounded disturbances is introduced and the OBE algorithm is used to achieve the robust estimation of the angular rate. In the OBE algorithm, the ellipsoid geometry center is often used as the point estimate of the true value, but it is not optimal theoretically. The Chebyshev center of the feasible set has many excellent features. Meanwhile, considering that it is very difficult to solve the exact Chebyshev center, the relaxed Chebyshev center is used as a substitute for the point estimate of the true angular rate. Then an OBE update process with RCC as output is designed and a novel parameter optimization criterion is proposed. The verification experiment is performed by using a gyro array composed by six gyroscopes. The experimental results show that the estimation fusion method based on the proposed algorithm can obtain the angle rate guaranteed boundary and further improve the MEMS gyroscope accuracy.