Inverse Relational Reasoning Based on 3D-CRN Direction Relation Model

In recent years, although a number of spatial directional relation models have been proposed, the reasoning ability of spatial directional relations is limited due to the inherent defects of some models and the weak research on the corresponding reasoning techniques, which reduces the usefulness of the models to a certain extent. Based on the 3D-CRN directional relation expression and inference model, combining three-dimensional algebra and qualitative coordinates, establishing equivalent links, an automatic qualitative inference algorithm for the inverse relation between principal directional relations is proposed, which gets rid of the dependence on manual inference tables, greatly improves the operational efficiency, and increases the practicality of the model. The results of theoretical analysis and example verification show that the algorithm effectively improves the accuracy of reasoning and reduces the time complexity of reasoning.