An Agent-Based Approach to Reconciling Data Heterogeneity in Cyber-Physical Systems

摘要

Computing and communication devices in any cyber-physical system (CPS) of non-trivial scale exhibit significant heterogeneity. Critical infrastructure systems, which are prime examples of CPSs, are no exception. The extent of networking capability, decentralized control, and more generally, integration between the cyber and physical infrastructures can vary greatly within a large-scale CPS. Other manifestations of heterogeneity in CPSs are in the resolution, syntax, and semantics of data collected by sensors from the physical infrastructure. Similar challenges complicate the use of databases that maintain past sensor data, device settings, or information about the physical infrastructure. The work presented in this paper aims to address these challenges by using the summary schemas model (SSM), which enables heterogeneous data sources to be queried with an unrestricted view and/or terminology. This support for imprecise queries significantly broadens the scope of data that can be used for intelligent decision support and carries the promise of increased reliability and performance for the CPS. We seek to ensure that ambiguity and imprecision do not accompany this expanded scope. The ultimate goal of a CPS is to fortify and streamline the operation of its physical infrastructure. The success of this task is contingent upon correct and efficient interpretation of data describing the state of the physical components, and the constraints to which it is subject. To this end, we propose agent-based semantic interpretation services that extract meaningful and useful information from raw data from heterogeneous sources, aided by the SSM. The proposed approach is described in the context of intelligent water distribution networks, which are cyber-physical critical infrastructure systems responsible for reliable delivery of potable water. The methodology is general, and can be extended to a broad range of CPSs, including smart power grids and intelligent transportation systems.