Increasingly, electricity network managers, through their grid codes, require renewable energy production systems to participate in system services, which includes requirements such as the stability of these production systems, the quality of the energy injected into the networks, the ability to withstand voltage dips, etc. To meet these requirements, the use of appropriate commands for the control of the production systems is necessary. Various control methods have been proposed, among which direct torque control (DTC) stands out. However, several studies have highlighted the impact of parametric variations on this control method. The contribution of the work presented in this article is the improvement of DTC when combined with a fuzzy estimate applied to a wind production system based on an asynchronous machine. Robustness tests were simulated to highlight the sensitivity of this control to variations in the stator resistance of asynchronous machines. To make this command robust and stable, a fuzzy estimator was used with this command. The simulation results demonstrated that this combination (DTC with a fuzzy estimator) makes the wind system more stable. To assess the effectiveness of the proposed solution, the root mean square error index was used.