微陀螺滑模控制器中的积分算法设计（英）

为了去除不定积分环节产生的直流偏置项，提出了一种基于一次移动平均法的改进积分算法。利用一个整数周期的积分信号值构成一个集合，计算出集合的算术平均值作为瞬时直流偏置，将原积分信号减去该直流偏置，最终可得到无直流偏置项的积分信号。为实时更新直流偏置，在每个时钟上升沿将一个积分信号的新采样值加入集合尾部，并去除集合首部的数值点，从而确保集合长度维持固定值并能实时反映出最新的直流偏置。以FPGA为信号处理平台，采用Modelsim仿真和实验测试相结合的方法验证了该算法的可行性。该方法已成功用于电磁式微陀螺仪的滑模控制器之中。同已有方法相比，该积分算法复杂度低、实时性好、易于实现。

Integral algorithm design in sliding mode controller of MEMS gyroscope

In order to remove the undesired DC offset of integral signal, a modified integral algorithm based on the Single Moving Average Method is proposed. In this algorithm, the N samples during one cycle are collected as a set, and the average value of them is calculated as the instantaneous DC offset. Therefore, a pure AC signal without any offset is obtained by subtracting the DC offset from the integral signal. On every rising edge of the clock, a new sampling point is added into the set and the oldest one should be removed. In this way, the length of the set remains constant and the DC offset can be updated in real time. Both the simulation and test results demonstrate the effectiveness of this integral algorithm. This algorithm has been successfully used in the sliding mode controller of a MEMS gyroscope. Compared with the existing integral algorithm, the proposed method is less complex, easier to be realized and can achieve better real time response.