机构地区: 国防科学技术大学
出 处: 《汽车技术》 2004年第3期22-24,27,共4页
摘 要: 常用的ABS轮速信号的滤波技术是滤波增益为常数的指数平滑数据滤波技术,由于没有考虑到测量信号的方差变化特性,滤波效果不能满足ABS系统的要求。文中根据实测轮速信号的方差数据,利用二阶多项式数据拟合得到了系统的测量方差。在假定系统方程和过程方差的情况下,轮速信号的处理采用了卡尔曼滤波器的结构。实际试验结果表明,可变滤波增益的滤波器滤波精度高、滞后小,可以满足计算ABS其它控制量的要求。 The typical filter technique in ABS wheel speed signals is the exponential smoothing data filter technique whose filter gain is constant.Because the variation characteristics of measured signal variance are not to be considered,its filter effect can not meet the requirements of ABS.In this paper,measurement variance of the system is obtained with second order polynomial data fitling based on actual variance data of wheel speed signals.Assuming the system equation and process variance,KALMAN filter structures are used in wheel speed signal processing.It is shown by tests that the gain-variable filter is precise and less lag,and meets the requirements to calculate other control variables of ABS.