Paper describes the use of algorithmic knowledge models for overnight testing automation. The arhitecture of intellectual testing system is described. Algorithmic models formalization is proposed. Knowledge model interaction and execution is presented for a real world test automation system.
The article deals with software architecture for testing and verification methods for solving systems of equations in partial derivatives describing physical processes. Developed automated generation of processor architecture problems, their solution and comparative analysis with other methods of analysis tasks mathematical physics. Analyzed with software on shell COMSOL 4.2.