Title :
Fast static analysis of real-time rule-based systems to verify their fixed point convergence
Author :
Cheng, Albert Mo Kim ; Wang, Chih-Kan
Author_Institution :
Dept. of Comput. Sci., Texas Univ., Austin, TX, USA
Abstract :
A class of real-time rule-based decision systems in which decisions are computed by an equational rule-based (EQL) program is described. The timing analysis of interest is to verify that a real-time EQL program converges to stable values in bounded time at each invocation. Techniques for determining whether an EQL rule-based program is guaranteed to converge to stable values in bounded time are presented. An approach that is based on static analysis of the rules of the EQL program and that does not require generating a reachability graph is discussed. The approach is utilized to perform a pre-run-time analysis on two real expert system programs, the Integrated Status Assessment Expert System (ISA) and the Fuel Cell Monitoring Expert System (FCE), to verify whether the variables in these programs will always converge to stable values in bounded time at each invocation.<>
Keywords :
convergence; decision support systems; expert systems; program verification; real-time systems; EQL rule-based program; FCE; Fuel Cell Monitoring Expert System; ISA; Integrated Status Assessment Expert System; bounded time; equational rule-based; fixed point convergence; pre-run-time analysis; reachability graph; real expert system programs; real-time EQL program; real-time rule-based decision systems; stable values; static analysis; timing analysis; Contracts; Convergence; Design automation; Hydrogen; Knowledge based systems; Performance analysis; Real time systems; Space vehicles; State-space methods; Timing;
Conference_Titel :
Computer Assurance, 1990. COMPASS '90, Systems Integrity, Software Safety and Process Security., Proceedings of the Fifth Annual Conference on
Conference_Location :
Gaithersburg, MD, USA
DOI :
10.1109/CMPASS.1990.175401