A　design　of　internal　model　control　systems　for　process　with　delay　using　kernel　ridge　regression　(KRR)　is　proposed,　which　fully　uses　the　excellent　nonlinear　estimation　performance　of　KRR　based　on　structural　risk　minimization　principle.　Internal　model　control　system　is　constructed.　Closed-system　stability　and　steady　error　are　analyzed　for　the　existence　of　modeling　errors.　The　simulations　show　that　the　proposed　control　systems　have　the　better　control　performance　than　neural　networks　in　training　samples　with　finite　size　and　noises.