The　thesis,　based　on　complexity　adaptive　system　theory,　simulates　the　evolvement　process　of　industrial　clusters　innovation　network　by　making　use　of　simulation　means.　It　is　shown　that　innovation　agents　adjust　constantly　action　rules　themselves　in　order　to　adapt　to　the　circumstance,　which　results　in　the　emergence　of　innovation　network.　At　the　same　time,　some　complications,　knowledge　difference　and　the　extent　of　specialization　of　innovation　agents　and　the　means　knowledge　is　transferred,　are　influencing　weightily　the　evolvement.　©　2008　IEEE.