| Popis: |
The Adaptive Modeling Language (AML) is an object-oriented, knowledge-based programming language with some unique capabilities which make it a very versatile design environment. Some of these capabilities include dependency tracking, demand driven calculations, run-time model modification and collaborative engineering. Because the model can be modified at run-time, the user of the code is permitted more flexibility in the choosing of the design variables, objectives and constraints than the code developer initially conceived. However, one of the current deficiencies of AML for the design process is optimization. This paper presents an integration of the Adaptive Modeling Language with the Design Optimization Tools (DOTTM), a commercial off-the-shelf optimization algorithm. An example showing the use of DOT in AML is given. User supplied gradient information is obtained from automatic differentiation in AML. An overview of automatic differentiation in AML is given. |
