The　model　predictive　control　(MPC)　strategy　with　a　control　Lyapunov　function　(CLF)　as　terminal　cost　is　commonly　used　for　its　guaranteed　stability.　In　most　of　the　cases,　CLF　is　locally　designed,　and　the　region　of　attraction　is　limited,　especially　when　under　control　constraints.　In　this　article,　the　stability　and　the　region　of　attraction　of　constrained　MPC　that　is　applied　to　the　discrete-time　nonlinear　system　are　explicitly　analyzed.　The　region　of　feasibility　is　proposed　to　substitute　the　region　of　attraction,　which　greatly　reduces　the　calculation　burden　of　terminal　constraints　inequalities　and　guarantees　the　stability　of　the　MPC　algorithm.　Also,　the　time-variant　terminal　weighted　factor　is　proposed　to　improve　the　dynamic　performance　of　the　close-loop　system.　Explicit　experiments　verify　the　effectiveness　and　feasibility　of　the　relative　conclusions,　which　provide　practically　feasible　ways　to　stabilize　the　unstable　and/or　fast-dynamic　systems.