Research On Context Modeling And Reasoning For Smart Vehicle Space

With the rapid development of pervasive computing environment and smart spaces, context-aware computing, as an important sub-areas, needs to be able to adapt to highly dynamic heterogeneous computing environment, and requires all entities in the smart space (eg, equipment, services and agents) must be able to aware the context they are in, and make user-centric timely and effective changes to the context. However, owning to the characteristic of single, low, unstable, imprecise and so on of the original context that can be aware in the current smart space, how to identify a valid context and how to infer the user's intent and state using the context are the key issues of our research. Moreover, with the growing proliferation of sharing of context information, how to achieve effective and quantify the uncertain knowledge information in it become another hot spot of research.This article focuses on these issues for research and practical work, we introduce ontology into the smart vehicle space context modeling, and make the truly knowledge reuse and sharing of the system possible. Also we propose an ontology mapping framework that drived by uncertainty context reasoning mechanism to deal with the requirement in the field of smart space. In the uncertainty context reasoning mechanism, we choose the Dynamic Bayesian Network and Hidden Markov Model to solve the reasoning calculates of the context, so that we can ensure the reliability, accuracy and efficiency of the system and meet the user's demand for services. The main work of this article is as follows:First, we start to do the context modeling of smart vehicle space based on ontology, and describe it by OWL. At the same time, to meet the context reasoning of a uniform standardized reasoning ontology, we adopt ontology mapping methods to adapt the reasoning of the context uncertainty information and satisfy the mathematical structure of reasoning.Then, we propose dynamic Bayesian network and Hidden Markov Model-based reasoning mechanism to solve the problem of uncertainty and integrity, and build the reasoning methods by combining the mathematical probability and the state of driver's body to make the context reasoning in smart vehicle space.Finally, in the view of application, we integrate the ontology mapping method and uncertainty context reasoning mechanism applying to the recognization of driver's body state in the smart vehicle space, and analyze the experimental data and the result of application to verify the change that can be aware by the system and make accurate state recognition, meanwhile, remind drivers to ensure the safe driving behavior.