The traditional Kalman filter algorithm is not robust with uncertain parameters. In view of this and based on the ensemble robust filters, the optimal method of data assimilation constructed from observations, which was referred to as the ensemble timelocal robust filter of inflating the observational covariance matrices, was presented, and the rule of the algorithm and the inference of the formula of the approach presented were given. The approach presented was compared with the ensemble Kalman filter method on the robustness and the assimilation accuracy using the strongly nonlinear Lorenz-96 model and on the basis of the changeable condition of the performance level parameters, the force parameters, and the size of observations and ensemble. The results suggest that: the root mean square errors of the ensemble Kalman filter method are much larger than those of the time-local robust filter; the ensemble Kalman filter produces filter divergence with a relatively small observation or ensemble size, while the root mean square error of the robust filter has slightly change; compared with the traditional ensemble Kalman filter algorithm, the time-local H_∞filter approaches using the observation inflation is more robust on the changes of system parameters, and improve the accuracy of filtering.