In the presence of main-lobe interference, the main-lobe offset comes out on the adaptive pattern by eigen-projection matrix preprocessing. Aimed at the problem that the fixation of coefficient is difficult to determine in the traditional main-lobe shape-preserving method of diagonal loading technique, an improved ridge regression estimation based on HKB algorithm is proposed. The load can be determined adaptively according to the data received. The new diagonal loading coefficient is obtained by replacing the minimum variance cost function of the residual vector with the error absolute value function when performing the ridge regression estimation. At the same time, the improvement of the proposed method for the adaptive antiinterference total effect pattern is studied. For the obvious deterioration of the sidelobe performance, the improvement of the total effect pattern is studied in combination with the proposed method. The simulation results demonstrate that the proposed method is correct and effective, and not only can effectively solve the mainlobe offset problem of the pre-processed pattern, but also improve the side-lobe performance of the total effect pattern.