This article aims to improve the output voltage efficiency of a Permanent Magnet Synchronous Generator (PMSG) with the benefit combination of Neural Networks and Fuzzy Logic. The previous both give a Neuro-fuzzy system tool. The PMSG is associated with a vertical-axis tidal turbine to produce electrical energy. PMSG and the vertical tidal turbine make up the studied system. Regarding the regulation and control of the Pulse Wave Modulation (PWM) rectifier at the output of the system, two controllers are explained and their performance is tested and compared: Maximum Power Point Tracking (MPPT) at Tip Speed Ratio (TSR) and MPPT Neurofuzzy. The two control algorithms were simulated for a turbine operating at maximum mean efficiency with a flow of 1.5 m/s. It is shown that when an unregulated generator feeds a resistive and inductive load, the amplitude of the output voltages is more undulating compared with a tidal turbine equipped with MPPTTSR and MPPT Neuro-fuzzy control systems. It shows that the optimal MPPT-TSR improves and stabilizes voltages by seeking the reference rotational speed as a variable to minimize the error between the measured and reference rotational speeds. The MPPT Neuro-fuzzy controller, on the other hand, offers a clear improvement and better stabilization of voltages by providing the reference rotational speed as instantaneous to minimize the error between the measured and reference rotational speeds. As for the MPPT Neuro-fuzzy controller, it offers a clear improvement and better stabilization of voltages by providing the reference speed of rotation as instantaneous to minimize the error between the recorded rotational speed and the specified reference speed.