RQ1:Preciseness

Table 4 presents the experimental results of evaluating the preciseness of the abstract model. The average preciseness over all three benchmark systems is 92.22%. And for ACC, AFC, and CSTR, the average preciseness is 86.04%, 99.99%, and 99.91%, respectively. It can be observed that ACC has a lower preciseness compared to the other two systems, which both have a value above 99.9%. The reason for this could be that, the system state space of ACC has a higher dimension than AFC and CSTR, which requires more tuning on the abstraction parameters, e.g. c and 𝑘.