Within　the　realm　of　smart　manufacturing,　the　growing　complexity　and　scale　of　computational　tasks　expose　the　limitations　of　conventional　computing　architectures　with　rigid　task　allocation　and　fragmented　collaboration,　which　impede　the　efficiency　and　security　of　smart　manufacturing　systems.　To　address　these　challenges,　this　paper　presents　an　innovative　cloud-fog-edge-terminal　collaborative　strategy　that　provides　an　adaptive　and　comprehensive　offloading　solution　for　tasks　with　complex　dependencies　and　large　scale,　thereby　enhancing　the　seamless　collaborative　potential　of　hierarchical　computing　structures.　Additionally,　a　joint　optimization　mathematical　model　for　collaborative　computational　offloading　is　developed,　aiming　to　minimize　task　offloading　time　and　assess　manufacturing　risk.　To　refine　the　solution,　an　advanced　multi-objective　optimization　algorithm　is　formulated　to　identify　the　optimal　solutions.　The　effectiveness　and　practical　applicability　of　the　proposed　method　are　substantiated　through　simulation　experiments　and　empirical　case　studies,　demonstrating　a　performance　enhancement　of　12-29%　over　other　benchmarks.　The　joint　optimization　method　effectively　synchronizes　cloud-fog-edge-terminal　computing　resources,　realizing　efficient　and　secure　task　offloading　and　execution　in　smart　manufacturing　scenarios.