基于奇异谱分解的微机械加速度计振动噪声抑制方法

微机械（MEMS）惯性传感器成本低的同时噪声较大，易受振动信号的干扰。为了利用微机械惯性传感器构成低成本姿态估计系统，提出了一种基于奇异谱分解（SSA）的振动噪声预处理方法。SSA方法的实质是利用延迟扩维矩阵进行主成分分析，其延迟相关的算法能够有效地分离出加速度计测量值中的趋势项与周期项，趋势项中包含有需要的姿态变化信号，周期项即为低频振动噪声，根据过零点检测方法提取出趋势项，将该趋势项作为加速度计的测量值，即可实现对振动噪声信号的抑制，有效地提高姿态估计精度。实际的跑车实验验证了本方法的可行性和有效性。

De-noising method for MEMS accelerometers based on singular spectrum analysis

Micro-Electro Mechanical Systems (MEMS) based inertial sensors are low-cost but their performances are also degraded because of the large uncertainties in their output and the effects caused by vibrations. To estimate the attitude using the MEMS inertial sensors, a pretreatment method to mitigate the noise of the inertial sensors is proposed based on the singular spectrum analysis (SSA). SSA belongs to the general category of PCA methods. With the so-called lagged covariance matrix of this approach, the trend and periodic components are separated by SSA. As a result, the true attitude signal is contained in the trend component, while the vibrations are contained in the periodic components. Then, the trend component is extracted by the use of the number of zero-crossing. Finally, the true attitude measurements pretreated by the SSA are utilized as the measurement input of the fusion filter for accurate attitude estimation. The car tests verified the feasibility and the capacity of the method to improve the accuracy of the attitude estimation effectively.