The　performance　of　state　of　charge　(SOC)　estimation　can　be　improved　by　using　multi-dimensional　signals.　However,　conventional　battery　models　developed　for　SOC　estimation　have　difficulties　in　simulating　the　coupling　relationship　between　the　mechanical　and　electrical　characteristics.　In　this　paper,　an　electromechanical　coupling　model　(EmCM)　of　a　lithium-ion　pouch　battery　module　is　established　for　SOC　estimation　in　real-time.　To　achieve　the　closed-loop　SOC　estimation　based　on　force　signal　feedback,　the　stack　pressure　is　chosen　as　the　model　output.　The　current　and　SOC　are　set　as　the　model　input　and　a　state　variable,　respectively.　On　this　basis,　a　novel　SOC　estimation　method　through　current　and　stack　pressure　is　proposed.　The　model　parameters　are　identified　by　a　genetic　algorithm,　and　SOC　estimation　is　performed　using　the　Extended　Kalman　filter　algorithm.　The　experiment　results　indicate　that　the　proposed　EmCM　can　depict　the　stack　pressure　variations　with　high　accuracy.　The　SOC　estimation　error　can　be　controlled　within　±2.8%　for　both　Li　[NiCoMn]O2　cells　and　LiFePO4　cells.　©　2022　Elsevier　Ltd