Changes　in　pedestrian　dynamics　caused　by　social　distancing　policies　place　new　demands　on　pedestrian　motion　modeling　during　the　pandemic.　This　study　summarizes　pedestrian　movement　characteristics　during　the　pandemic,　based　on　which,　the　traditional　floor-field　cellular　automata　model　was　improved　by　introducing　two　floor　fields　related　to　pedestrian　density　to　simulate　social　distancing　in　crowded　places.　Especially,　the　cumulative　density　field　guides　pedestrians　in　route　selection,　thereby　compensating　for　the　limitation　of　the　previous　models　in　which　only　local　repulsion　was　considered.　By　selecting　an　appropriate　combination　of　parameters,　the　desired　social　distancing　behavior　can　be　observed.　Then,　the　rationality　of　our　model　is　verified　by　the　fundamental　diagram.　Moreover,　to　assess　the　influences　of　social　distancing　on　the　risk　of　disease　transmission,　we　considered　both　person-person　transmission　and　environment-person　transmission.　The　simulation　results　show　that　although　social　distancing　is　effective　in　preventing　interpersonal　transmission,　an　increase　in　environmental　transmission　may　somewhat　offset　this　effect.　We　also　examined　the　influence　of　individual　motion　heterogeneity　on　infection　spread　and　found　that　the　containment　was　the　best　when　only　patients　complied　with　the　social　distancing　restriction.　The　trade-off　between　safety　and　efficiency　associated　with　social　distancing　was　also　initially　explored　in　this　study.