Skip to Main Content
Most of the traditional researches of the robotics based on multi-sensor data fusion concentrated on the problems of autonomous navigation and localization, few of them involved the target tracking. The existing fusion method based on the network time-delay applied the traditional synchronized fusion algorithm directly, therefore, it had led many problems, such as the information waiting, resources waste, and the bad real-time performance. This paper unified the predictive estimate and the recursion weighted fusion technology to design a new multi-sensor predictive weighted fusion algorithm, which could adapts to the network time-delay. This algorithm not only can solve the defects of the existing data fusion method based on the network time-delay effectively, but also has good function of real-time prediction. In this paper, the author provided the deduction process of the new algorithm, and demonstrated the practicability and superiority of the new algorithm through the computer simulation example.