基于归一化处理的自适应洗出算法

Adaptive Washout Algorithm Based on Normalization

传统的自适应洗出算法随着输入幅值的增大,其输出会变得不稳定,从而造成平台的异常抖动。为了克服这一缺陷,对自适应律提出一种归一化的修正方案。此方法使得自适应洗出算法中的输入幅值的大小不会对自适应增益造成影响,进而确保输出不会因为输入幅值过大而产生发散,提高动感模拟的真实性及稳定性;而且,改进算法没有引入额外的惩罚项,因此并未使算法变得更加复杂。经过特定的运动数据进行验证,结果显示此方法可以降低输出发散,减少平台的异常抖动,提高模拟驾驶的逼真效果。

With the increase of input amplitude,the output of the traditional adaptive washout algorithm appears unstable,resulting in the abnormal jitter of the platform.To overcome this defect,a normalized correction scheme is proposed for the adaptive law.This method makes the adaptive gain independent from the input amplitude,thus ensuring that the output does not diverge with the increase of the input amplitude to improve the authenticity and stability of the motion cueing;moreover,the improved algorithm does not introduce additional penalty terms,so it does not make the algorithm more complex.The improved washout algorithm is tested by using the start-up,uphill-going and braking data of the loader.The results show that this method can reduce the output divergence,reduce the abnormal shaking of the platform,and improve the realistic effect of simulated driving.