Key points are not available for this paper at this time.
This paper proposes a new multivariate chart corresponding to a T2- chart robust to outliers using three methods, namely the Rousseuw and Leroy algorithm, Maronna and Zamar, and the family of "concentration algorithms" by Olive and Hawkins. Then the comparison between the proposed and classical method of the researcher Shewhart depending on the total variance (trace variance matrix) and the general variance (determinant of the variance matrix) to obtain the most efficient paintings against outliers through simulation and real data and using a program in MATLAB language designed for this purpose. The study concluded that the proposed charts dealt with the problem of the influence of outliers and were more efficient than the classical method.
Sedeeq et al. (Wed,) studied this question.