A framework to model big data driven complex cyber physical control systems

Abstract

First, big data driven cyber physical systems (CPS) have big data characteristics. Big Data is usually described by the following aspects - Volume, Velocity, Variety, Veracity, Validity, Value, and Volatility. Second, big data driven CPS have special characteristics and requirements that must be met during system development. These big data characteristics and special characteristics requires methods and techniques for the data specification, modeling, capture, transfer, management and algorithms for their collection, transfer, analysis, storage and processing. The design methods for big data driven cyber physical systems not only meet multiple V's characteristics, but also must meet special characteristics of CPS such as spatio-temporal requirements and real time communication requirements. In this paper, a model based approach is proposed for modeling big data driven cyber physical systems based on integration of AADL, Modelicaml and clock theory, the proposed approach allows the design of object-oriented and component model for big data driven CPS, the main advantages of the proposed approach are its capacity to take into account big data properties and cyber physical system properties through specialized concepts in rigorous, easy and expressive manner. The proposed approach is illustrated by a case study of conceptual design of aviation cyber physical systems.