This work proposes a robust and non-gaussian version of the shrinkage-based EnKF implementation, the EnKF-KA. The proposed method is based in the robust H filter and in its ensemble time-local version the EnTLHF, using an adaptive inflation factor depending on the shrinkage covariance estimated matrix. This implies a theoretical and solid background to construct robust filters from the well-known covariance inflation technique. The method is tested using the Lorenz-96 model to evaluate the robustness and performance under different scenarios as ensemble size, observation error, errors in the model specifications, and ensemble gaussianity. The results suggest good robustness of the proposed method in all the evaluated cases compared with the standard EnKF, the shrinkage-based EnKF-KA, and the robust EnTLHF.