Modeling and Decision-Making Methods for a Class of Cyber-Physical Systems Based on Modified Hybrid Stochastic Timed Petri Net

Cyber-physical system (CPS) is a multidimensional complex system that integrates computation, communication, and physical environment with essential and broad application prospects. However, there is a kind of CPS containing discrete events, continuous processes, stochastic phenomena, time-delay, and decision. It suffers from several sophisticated modeling and decision-making challenges, including deep integration of cyber and physical world, intensive temporal properties, model uncertainty, concurrency, and behavioral control. In this article, we deal with the above problems from the perspective of hybrid systems. We propose the modeling methods of CPS based on the modified hybrid stochastic timed Petri net (M-HSTPN) with three-tier architecture and introduce the decision place to strengthen the decision-making ability of CPS. Its advantages lie in describing multiple problems in CPS at the same time by adopting a unified open architecture. By integrating the various modeling and decision-making challenges under a unified framework, it is easy to analyze the problem systematically. Taking the humanoid soccer robot CPSs (HSR-CPS) as an example, by analyzing factors that affect the action cycle and interception success rate of the HSRs, we show the powerful modeling and decision-making capabilities of the M-HSTPN for CPS.