j+1], three adjacent landmarks in some variable quantity space, an example of QSIM P-transitions and I-transitions starting from the same state is given in table 1.
For representing the behavior of a system, QSIM uses three kinds of qualitative constraints: (1) arithmetic, (2) differential, and (3) functional.
The QSIM model of the bathtub is given in figure 11.
The structural abstraction theorem of QSIM proves that each ODE can be abstracted into a QDE such that any continuously differentiable function that is a solution of the ODE also satisfies the QDE.
Qualitative simulation approaches, exemplified by QSIM, are trapped in the local nature of the algorithm and the specific formulation of a qualitative model in terms of constraints.
First is use more quantitative information, such as the semiquantitative simulation approach, including quantitative extensions of QSIM such as Q2 and Q3 (Berleant and Kuipers 1997; Kuipers 1994) that preserve the underlying qualitative semantics and interval model based simulation (Armengol et al.
SQUID, based on QSIM semiquantitative extensions, only deals with the refinement of a single semiquantitative differential equation that represents the whole model space, whereas both RHEOLO and PRET deal with automated system identification and, in outline, follow the reasoning flow depicted in Figure 13.