Model Equivalence Verification via Semantic Unit Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In safety-critical systems, existing verification methods fail to comprehensively verify model transformations due to increasing complexity, as they rely on testing-based evaluations and simple syntax checkers that cannot capture the semantics of reference models, leading to incomplete structural coverage and dependence on design processes.
Innovation Solution
A system for verifying model equivalence is developed, which includes a processing unit that generates and compares reference and comparison model representations, ensuring compatibility through data flow diagram and semantic unit verification, independent of the design process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If testing-based evaluations and simple syntax checkers are used to verify model transformations, then the verification process is simple and fast, but the verification is incomplete and cannot capture the semantics of reference models
Solution Approach 1:
The verification process is divided into multiple independent verification stages: syntax verification, structural coverage verification, and semantic equivalence verification. Each stage focuses on a specific aspect of model transformation, allowing comprehensive verification without overwhelming complexity in a single system.
Solution Approach 2:
The patent introduces an intermediate representation layer that translates reference models and comparison models into a common verification format. This intermediary mechanism enables semantic comparison without requiring direct complex interaction between the verification system and the full model specifications.
2Reliability
If comprehensive semantic verification is implemented, then model equivalence can be accurately verified, but the verification process becomes dependent on the design process
Solution Approach 1:
The verification framework performs preliminary actions by pre-defining the verification criteria and independent verification paths before the actual verification runs. This allows comprehensive semantic verification to be conducted independently of the specific design process, as the verification logic is prepared in advance and can be applied universally.
3Reliability
If industry standards like RTCA DO-178B are strictly enforced, then verification rigor is improved, but the verification process becomes more complex and time-consuming
Solution Approach 1:
The verification system operates continuously by automatically executing verification checks as part of the model transformation process. Rather than performing verification only at the end, the system continuously verifies syntax, structure, and semantics throughout the transformation pipeline, reducing total verification time while maintaining RTCA DO-178B compliance.
Data Source
AI summary
Systems and methods for verifying model equivalence are provided. In one implementation, a system includes: a memory device that stores a reference model (RM) and comparison model (CM), wherein the CM and the RM are constrained by a set of rules; and a processing unit that generates a reference model representation (RMR) from the RM and stores the RMR on the memory device; the processing unit further generates a comparison model representation (CMR) from the comparison model (CM) and stores the CMR on the memory device, wherein the processing unit further to: verifies that the CMR compatibly implements the RMR; verifies that a CM data flow diagram derived from the CMR compatibly implements a RM data flow diagram derived from the RMR; and verifies that every CM semantic unit implements a behavior that corresponds to a RM semantic unit and every RM semantic unit is accounted for in the CM.