This work suggests a reliable continuous-time model predictive direct power control for doubly fed induction generators (DFIG). In a finite time horizon, the Taylor series expansion can be used to predict the stator current in the synchronous reference frame. In order to reduce the difference between actual stator currents and their reference values over the anticipated time period, the required rotor voltage is calculated using stator current predictions. Due to the proposed technique’s high sensitivity to parameter fluctuations and outside disturbances, disturbance observers are incorporated into the control loop to eliminate the stator current’s steady-state error. Both steady-state and transient performance can be determined using some fundamental design parameters. In this article, stator current is calculated directly from the intended stator’s active and reactive powers without taking into account the machine’s specifics. As a result, the DFIG’s control structure does not need an additional power control loop to function properly. The grid-connected DFIG’s experimental results demonstrate the effectiveness of the suggested approach, and satisfactory outcomes are obtained.