Aiming at the information fusion problem among multi-sensor data of unmanned weapon system, put forward the unmanned weapon system information fusion algorithm based on combination of AHP method and D S evidence theory. According to the influence of environment on sensor accuracy when multi sensor information fusion, confirm every sensor value of multi-sensor basic probability evaluating by analytic hierarchy process (AHP) method, use D S evidence theory for improvement and realize object recognizing. The experiment analysis result shows that improve identification results is well than tradition D S evidence theory, and improve the performance of target identification system to a certain extent.