Add the GAA (General Aviation Aircraft) design problem

[gialmisi]

Background

The General Aviation Aircraft (GAA) problem is a well-known multiobjective benchmark from aircraft conceptual design: design variables (cruise speed, aspect ratio, sweep, propeller diameter, wing loading, etc.) trade off up to 10 objectives (e.g. takeoff noise, empty weight, direct operating cost, ride roughness, fuel, cruise speed, …). The objectives come from response-surface (polynomial) models, so it can be implemented analytically. DESDEO has no implementation currently.

Note the existing implementation (see Reference below)!

Template to copy: desdeo/problem/testproblems/re_problem.py (engineering problem with polynomial objective expressions).

What to do

• Add gaa_problem() -> Problem in desdeo/problem/testproblems/gaa_problem.py with the design variables, the (up to) 10 response-surface objectives, and bounds/constraints from the reference.
• Allow selecting a subset of objectives if helpful; write the formulas/metamodels in the docstring. Do not write the full formulas, as they are very long, keep the LaTeX informative.
• Register in __init__.py; add a test asserting values at a known point.

Reference

• Woodruff, M. J., Reed, P. M., & Simpson, T. W. (2013). Many objective visual analytics: rethinking the design of complex engineered systems. Structural and Multidisciplinary Optimization, 48, 201–219.
• Previous implementation in DESDEO found here: https://github.com/industrial-optimization-group/desdeo-problem/blob/master/desdeo_problem/testproblems/GAA.py

Done when

• gaa_problem is importable, evaluates correctly for the chosen objective set, and has a passing test.