Grid-connected pre-synchronization control method and system based on neural network adaptive control

The invention discloses a grid-connected pre-synchronization control method based on neural network adaptive control. The grid-connected pre-synchronization control method comprises the following steps: converting three-phase output voltage Ugabc of a power grid into Ugd and Ugq according to Park transformation; adjusting the Ugq to zero according to a CMAC and PID composite adaptive control algorithm so as to obtain a power grid voltage angular frequency omega g and a voltage phase angle theta g during stable operation of the power grid; adding theta g to Park transformation of the three-phase output voltage Uabc, wherein Ud and Uq are obtained; adjusting Uq to zero according to the composite adaptive control algorithm to obtain a phase-locked tracking compensation angular frequency omegasc; adding omega sc to a droop control link and superposing the omega sc to the angular frequency omega 0 of the power grid voltage, wherein the reference angular frequency omega of the output voltage of the inverter is obta