Ｍ估计器是一种广泛应用于信号滤波的稳健性估计器。在Ｈｕｂｂｅｒ引入的Ｅｕｌｅｒ距离函数中，参数ｋ的确定十分重要，直接影响滤波性能。通过对信号局部统计特性的分析，提出了一种自适应选择参数ｋ的变参数Ｍ滤波器，为Ｍ估计提供了一般参数选择方法。文中的理论分析和测试结果表明，该算法对复杂噪声有较好的滤波效果。 M estimator is a robust estimator widely used in signal filtering. In the Euler distance function introduced by Hubber, the determination of the parameter k is very important, which directly affects the filtering performance. By analyzing the local statistical properties of the signal, a variable parameter M filter with adaptive selection parameter k is proposed, which provides a general parameter selection method for M estimation. The theoretical analysis and test results show that this algorithm has better filtering effect on complex noises.