Convergence　of　combination　evaluation　is　studied　through　simulation　experiment　to　resolve　inconsistency　among　the　evaluation　conclusions　drawn　by　different　evaluation　methods.　In　this　experiment,　stochastic　simulated　data　and　practical　examples　data　are　used　respectively　to　simulate　the　combination　evaluation　and　re-combination　evaluation　for　different　kind　methods.　The　results　indicate　that　combination　evaluation　converges　very　well.　After　several　times　of　combination,　combination　evaluation　results　can　reach　consistency　rapidly,　thus　consistent　evaluation　conclusions　are　obtained.