In　recent　years,　numerous　multisensor　data　fusion　systems　have　been　developed　for　wide　application.　There　are　many　algorithms　in　multisensor　data　fusion,　D-S　evidence　theory　is　a　useful　method　for　dealing　with　uncertainty　problems.　This　paper　describes　the　main　features　of　the　evidential　combination　algorithm　which　is　implemented　in　our　research,　It　also　gives　improvement　to　the　shortage　of　its　combine　rule　and　gives　it＇s　new　rule,　thus　we　can　combine　high　conflicted　evidences　with　different　weights,　which　have　been　calculated　through　different　information　included　in　the　numerous　evidences.　The　examples　have　been　done　to　demonstrate　the　efficiency　of　the　new　combine　rule,　it　can　improve　the　reliability　of　the　combination　results.