In order to improve the three-axis accelerometer′s measurement accuracy, an error model is established based on its error source and producing mechanism. Nonlinear observation equation is de-duced by the error model and constant amplitude of gravity field. Through introduced two unknown pa-rameter vectors, nonlinear observation equation is mathematically converted into a linear form relative to new unknown parameter. Then batch least-square method is used to estimate the new unknown parame-ters. Biases, scale factors and non orthogonal error are solved by inverse analytic arithmetic. The experi-mental results show that the maximum error of accelerometer can be reduced about 100 times and the ac-celerometer accuracy can reach 10-3 m/s2 . Experiments have good repeatability. The method is a high ef-ficiency batch least-square method and the compensation effect is remarkable.%为提高三轴加速度计测量精度,分析其误差来源和产生机理,建立了误差模型；根据误差模型和重力场幅值不变特性,推导出非线性观测方程。引入两个未知的参数向量,通过数学变换把非线性观测方程变形为关于新未知参数的线性方程；然后采用批量最小二乘法求出新的未知参数,并根据误差模型参数和新未知参数之间的关系,利用逆解析运算求出对应的零点偏差、灵敏度误差和三轴不正交误差。实验结果表明,加速度计正负误差峰值间距可以减小100倍左右,精度可达10-3 m/s2,并具有很好的可重复性。该方法算法效率较高,补偿效果显著。
展开▼
