A Relational Abstraction of Structure and Behavior for Cyber-Physical System Design

Model-based approaches are essential for designing cyber-physical systems, which adopt the formal models to simultaneously form the specifications and enable the verification at an early stage. Aimed to model the complex structure and continuous-discrete hybrid behavior of cyber-physical systems, this paper mathematically defines a dynamic relational system so that the cyber-physical system can be regarded as dynamic relational systems in a hierarchical structure and each dynamical relational system is a triple of dynamic attributes, subsystems, and hybrid relations between attributes and subsystems. Every hybrid relation contains a tuple and a predicate to govern the system behaviors. By utilizing the dynamic relational system, a parametric abstraction is then performed to specify the design requirements and schemes. It can represent the structure and behaviors of multiple cyber-physical system design schemes in an integrated manner. With a mathematical foundation, the constructed relational models are beneficial for structural analysis and behavior verification. An implementation case of a friction-driven plate conveyor is presented to illustrate the design specification with relational models, and the connectivity analysis and behavior verifications are carried out to show the effectiveness and engineering practicability of the achieved models.