| Summary: | Cyber-Physical Systems(CPS) are a group of systems that are involved with both physical processes and computational processes. The interaction of physical components and computational components makes it difficult to analyze, design and verify this type of systems. The problem becomes more complex when certain input or decision of these systems must be initiated by human. Cyber-Physical Systems with human operator in the loop are called Embedded Human Systems(EHS). To ensure the safety of EHS such as traffic control systems, space shuttle control systems, nuclear power plant control systems and so on, it is critically important for human operators to fully understand both physical and computational processes. However, humans are usually easily overwhelmed by concurrent information, the situation becomes worse when it comes to complex EHS with timing constraints.This dissertation proposes a domain specific modeling language that takes advantage of hybrid system abstraction to retain important system behaviors and automatically generates self-configured system verification software. The verification software could effectively reduce the computation time with parallel scheduling algorithm, thus the computation process that violates the design protocol can be halted without wasting computation resources. The modeling environment also allows user to conveniently set design constraints to avoid flaws early in prototype phase and reuse the available model for a family of different platforms. Several verification results of different platforms are shown to demonstrate the efficiency and reusability of the modeling environment.
|
|---|
