The article presents a new method to accurately estimate parameters in models with errors, even when there are multiple outliers in the data. By considering outliers in both the design matrix and observations, the researchers developed a way to detect and correct for these outliers. Their approach involves creating a location matrix to pinpoint the outliers and using numerical formulas to estimate the model parameters and outliers simultaneously. Through numerical experiments, they demonstrated that their method successfully eliminates the influence of outliers and provides reliable parameter estimates.