Behavior Verification of Hybrid Real-time Requirements by Qualitative Formalism (An Extended Abstract)

Although modern control theories have been successfully applied in solving a variety of problems, they are often mathematically and physically too specific to describe and analyze qualitative properties of hybrid real-time systems. In this paper, we propose to use qualitative formal methods, Compositional Modeling Language (CML) and Causal Functional Representation Language (CFRL) in particular, to specify continuous plant dynamics and the required system behavior and to simulate system behavior using a qualitative simulator known as the Device Modeling Environment (DME). Using Electrical Power System (EPS) as an example, we demonstrate effectiveness of our approach by illustrating how a simple SCR-style specification can be transformed and analyzed. Key Words Real-time, Hybrid system, Qualitative simulation, Formal method, Behavior verification, Requirements engineering 1. Introduction It is well-known in software engineering community that significant portion of software failures found in operational software can be traced to errors made during requirement engineering phase and that it is expensive to correct such errors. Hybrid real-time and embedded software development is an area where software requirements must be subject to highly rigorous and systematic analysis. Examples include software to control nuclear power plants, commercial and military jets, satellites, or manufacturing plants. A typical closed-loop process-control system consist of the following components (Figure 1): plant, controller, actuators, and sensors. …±°¶ ́±®®§ ́| bb•™ bbbŠ™ ’®£°¶|