RULE-BASED OBJECT-ORIENTED APPROACH FOR MODELING REAL-TIME SYSTEMS

The paper describes a system in which the rules, entities, and events of a real-time environment are represented as objects using the object-oriented paradigm. All rules are stored in a static rule base, and only rules that are relevant to the active events are maintained in the production rule set. A selection criterion with real-time factors is presented to enhance the performance of the matching process between the rules in the production rule set and working memory. Also a scheduling algorithm is proposed that incorporates time criticalness in scheduling transactions. Both selection and scheduling strategies are based on dynamic priority criteria.