A graph-based approach for timing analysis and refinement of OPS5 knowledge-based systems

We examine the problem of predicting the timing behavior of knowledge-based systems for real-time applications. In particular, we describe a suite of tools which analyze OPS5 programs to understand their timing properties. First, a graphical representation of an OPS5 program is defined and evaluated. This graph represents the logical control flows of an OPS5 program. Most of our analysis is based on this data structure. Second, we describe a novel tool which verifies that an OPS5 program can terminate in finite time. If the termination of the OPS5 program is not expected, the "culprit" conditions are detected. These conditions are then used to correct the problem by adding extra rules to the original program. Third, another tool is introduced to aid timing analysis of OPS5 programs. This tool generates a set of test data which maximize the program execution time. Other functions are also provided to facilitate the timing analysis.