A method of restoring models in embedded systems using the event log

Abstract

Background. In connection with the wide dissemination of cyber-physical systems concept, the issues of analyzing the processes of embedded systems are of interest. The purpose of the study is to automate the process of verifying the behavior of embedded systems by restoring the observed behavior model from event logs and its subsequent analysis. Materials and methods. The research was carried out using the methods of process mining and graph theory. Results. The proposed method allows you to extract the model of the observed behavior from the data from the event log. The retrieved model is visualized as a graph, considering the frequency characteristics of the events taking place. To extract the model, the authors proposed a modified inductive algorithm with post-processing of the results using the alignment algorithm. Evaluation of the method’s effectiveness was carried out on the student base of programs for the laboratory stand-designer SDK-1.1M, designed for prototyping embedded systems based on microcontrollers and Internet of things systems. Conclusions. The proposed method made it possible to restore the process model for various programs and then use the resulting models for mutual comparison, identifying compliance and deviations from the desired behavior. Achieved full automation of checking the correctness of the system behavior based on the reproduction of objects (tokenbased replay) to check compliance of process models.

Keywords

• process mining
• process model
• embedded systems
• functional monitoring
• compliance checking