Disclosed is a genetic algorithm-based transformer substation graph modeling automatically checking method. Existing power grids are large in scale and complex in structure, so that manual graph drawing work can be relatively complicated and can result in various kinds of problems existing in drawn SVG (scalable vector graphics) files, mainly including equipment misalignment and equipment overlapping. The genetic algorithm-based transformer substation graph modeling automatically checking method comprises, according to the formats of transformer substation wiring diagram SVG files for describing transformer substation equipment, extracting the position variables of the transformer substation equipment, and according to common graph modeling errors, designing an objective function; taking the position variables of the transformer substation equipment as an initial solution, optimizing the position variables of the equipment through the genetic algorithm, and exiting optimizing calculation when meeting certain conditions to complete transformer substation graph modeling automatic checking.