When　high　power　density　of　electric　vehicle　(EV)　charging　load　connected　to　the　microgrids　in　expressway　service　areas,　the　stability　of　the　system　is　affected.　Long-distance　transmission　or　transformer　expansion　can　be　costly　and　difficult　to　maintain,　therefore,　this　research　proposed　a　dynamic　capacity　expansion　method.　Combined　with　the　energy　storage　system,　through　the　most　economic　operation,　aiming　for　strong　microgrid　independence　and　minimum　power　fluctuation,　a　multi-objective　model　is　established　to　pursue　the　optimal　dispatch　scheme　by　integration　of　power　assets　using　the　Nondominated　Sorting　Genetic　Algorithm　II　(NSGA-II).　To　address　insufficient　diversity　for　evolutionary　operators　and　slow　speed　of　the　original　NSGA-II,　improvements　are　made　based　on　the　initial　dataset　optimization　and　differential　evolution.　Verified　by　the　simulating　results,　the　proposed　method　dynamically　expands　the　capacity　of　the　microgrid,　thus　mitigates　the　impact　of　heavy　charging　load.　©　2024　IEEE.