In　the　current　cloud　computing　environment,　task　scheduling　and　resource　allocation　are　the　key　and　difficult　points　in　the　performance　improvement.　However,　there　are　numerous　problems　of　workflow,　such　as　Montage,　Inspiral,　Cybershake　etc.　They　have　similar　workflow　structures,　which　affect　the　efficiency　of　task　scheduling　and　resource　distribution.　In　addition,　the　result　obtained　by　the　traditional　evolutionary　algorithm　is　the　allocation　sequence　of　the　virtual　machine　in　the　cloud　computing　environment　only　for　single　task,　which　is　a　great　waste　of　resources.　Aiming　at　these　problems,　the　multiple　workflow　tasks are　processed　in　this　paper　by　using　implicit　information　transfer　at　the　same　time,　that　is,　to　reasonably　use　the　allocation　sequence　of　each　task　to　exchange　information　so　as　to　share　a　better　virtual　machine　allocation.　Meanwhile,　using　the　potential　relationship　and　differences　between　different　tasks　are　better　able　to　make　population　has　better　convergence　and　diversity.　We　proposed　a　multifactorial　evolutionary　algorithm　based　on　combinatorial　population　(CP-MFEA)　for　multitasking　workflows.　This　paper　constructs　nine　sets　of　multi-task　combination　problems,　and　compares　the　method　with　the　traditional　single-task　evolutionary　algorithm,　the　purpose　is　to　describe　the　superiority　of　this　method　clearly.　Through　the　experimental　results,　we　can　notice　that　CP-MFEA’s　ability　is　much　more　obvious　than　single-task　evolutionary　algorithms.　©　2022,　Springer　Nature　Singapore　Pte　Ltd.