Simulation Research on Multi sensor Robot Odor Source Identification Strategy based on Cross-over Model

Based on the intelligent organism through a variety of sensory information to identify the object's cross-modal perception, this paper proposes a kind of cross-over model based on multi sensor information fusion robot taste
¹
recognition strategy that formally introduces subjective judgment and intuition into statistical inference and decision analysis so as to establish the basis for synthesis information inference and decision analysis process. This strategy can be based on the uncertainty information of multi sensor, the problem of incomplete representation of single sensor information is solved to achieve the purpose of taste recognition. In the simulation experiment, when using the initiative olfactory of the robot to search for and judgment the odor source, the smell and visual information are used to identify the suspected odor source quickly and effectively through the cross-over model taste recognition strategy, which proves the effectiveness of the strategy.