An Automatic Software Behavior Model Generation Method for Industrial Cyber-Physical System
Weiqi Sun, Wenbin Dai · 2020
Industrial Cyber-Physical Systems require more flexibility and resilience to meet the requirement of flexible manufacturing. Model-driven engineering methods are often linked with the development and deployment of distributed automation systems. However, most legacy systems currently do not have a system-level model or even source code, which hinders the maintenance of future-proof systems. With a huge amount of operation data collected by acquisition processes in the existing industrial systems, the system behavior model can be recovered but in an effective way. This paper proposes an automatic software behavior model recovery method based on data mining from industrial controllers. This method can recover and optimize the system models based on the state machines and largely reduce the computing power required for generating a system behavior state machine model. Finally, the proposed method was verified by the FSM model inferring and code generation using a color sorter example.