【摘要】 针对单一MEMS惯性陀螺仪精度低、随机噪声大，无法满足高精度应用场景要求的问题，提出了一种基于差分MEMS惯性传感器阵列的虚拟陀螺仪。设计并实现了由2N个消费级传感器构成的虚拟陀螺仪系统，其中正面N个传感器三轴同向，且与反面N个传感器三轴反向。2N个传感器共同组成N个差分传感器对，形成一个差分传感器阵列。同时，基于单一MEMS惯性陀螺仪的输出模型，对差分阵列式虚拟陀螺仪输出进行建模分析，通过将正反面传感器输出进行差分均值融合，实现了一个高精度虚拟陀螺仪输出。最后，利用陀螺仪零偏、零偏稳定性和Allan方差分析虚拟陀螺仪的确定性误差和随机噪声等特性。结果表明，在静止状态下，差分阵列式虚拟陀螺仪能够有效降低陀螺仪的零偏和零偏稳定性。与单一MEMS陀螺仪相比，差分阵列式虚拟陀螺仪的x,y,z轴角度随机游走降低约73%,速度随机游走降低约77%,即差分阵列式虚拟陀螺仪能有效地抑制MEMS惯性陀螺仪的随机噪声，提高传感器输出精度。更多还原

【Abstract】 Aiming at the problem that a single MEMS inertial gyroscope has low accuracy and large random noise, which cannot meet the requirement of high-precision application scenarios, a virtual gyroscope based on differential MEMS inertial sensor array is proposed. A virtual gyroscope system consisting of 2N consumer-grade sensors is designed and implemented, in which the front N sensors are axially-identical but inverse to the other N sensors. The 2N sensors together form N differential sensor pairs to form a differential sensor array. Meanwhile, based on the output model of a single MEMS inertial gyroscope, the output of virtual gyroscope based on differential array is modelled and analyzed, and a high-precision virtual gyroscope output is realized by differential mean fusion of the front and reverse sensor outputs. Finally, the deterministic error and random noise of the virtual gyroscope are analyzed by using the gyroscope bias, bias stability, and Allan variance. The results show that the virtual gyroscope based on differential array can effectively reduce the bias and bias stability of the gyroscope in the static state, and compared with the single MEMS gyroscope, the angular random walk of the x, y and z axes of the differential array virtual gyroscope is reduced by about 73%, and the rate random walk is reduced by about 77%, that is, the differential array virtual gyroscope can effectively suppress the random noise of the MEMS inertial gyroscope and improve the output accuracy of the sensor.更多还原