Abstract: In the simulation of modern distribution network with high-density distributed generators, the traditional numerical integration algorithm is difficult to meet the fast and stable simulation requirements, so an event triggering based improved variational algorithm is proposed. First, the variational iteration method（VIM）is introduced to obtain the approximate solution corresponding to the typical state variables of the bipolar photovoltaic power station in the distribution network. Then, the numerical stability of the algorithm is verified, and its limit step size is deduced. Aiming at the problem that the calculation of partial state variables consumes more simulation resources, the event triggering mechanism is introduced, and the triggering condition of the locked value is deduced by using the norm of error to reduce the calculation frequency and improve the simulation efficiency of the system. Finally, the numerical stability and event triggering effect of the algorithm are simulated and verified. The results show that the event triggering based variational algorithm improves the simulation speed while maintaining high accuracy,and the efficiency advantage is more obvious with the expansion of node scale.