Knowledge-Driven Data Fusion Modeling Using Petri Nets

Aiming at the issue that low data expression ability and low fusion efficiency in data fusion, a data fusion modeling methed based on Petri net by knowledge-driven is proposed. Firstly, the data of multi-sensors was processed. Based on this, the knowledge through the generation of rules to representation knowledge. Then the mapping model and the mapping algorithm are designed by using Petri net theory to establish the fusion model. Finally, the model is validated on the experimental platform of autonomous tracking robot. The experimental results show that three kinds of heterogeneous sensor data fusion can help autonomous tracking robot to better complete the tracking process. The correctness and applicability of this method are verified, which can be widely used in data fusion modeling of multi-source information.