Abstract: To solve the voltage deviation problem caused by the high penetration of distributed generation (DG) in the distribution networks, a strategy is proposed for optimal allocation of static var generator (SVG) with consideration of voltage mitigation of DG grid-connected inverter (GCIN) and voltage detection based active power filter (VDAPF). Firstly, a partitioning method based on community theory is proposed. And then, the dominant nodes in each region are selected as the candidate nodes of SVG to be installed. A multi-objective SVG optimal allocation model is established with minimum total cost of SVG and optimal mitigation effect of system voltage deviation, and the improved genetic algorithm is used to solve the allocation model. Considering the uncertainties of the remaining capacity of GCIN and VDAPF, the multi-scenario analysis technique is adopted to construct a series of voltage mitigation operation scenarios. A case study of the IEEE 33-bus distribution network is conducted to verify the effectiveness and rationality of the proposed strategy.