Preliminary airframe design using co-evolutionary multiobjective genetic algorithms

A novel multiobjective optimisation approach utilising a genetic algorithm (GA) for the preliminary design of airframes is introduced. Concurrent GA processes each optimise one objective related to the problem. The fitness measure for individuals within each GA is adjusted by comparing the values of the variable parameters of identified solutions relating to a single objective with those of the solutions of the other GA's. A penalty relating to the degree of diversity of their variable values as compared to those of the other GA processes is then imposed taking into consideration a generational parameter constraint map. Initial convergence upon individual objectives leads to overall convergence of all processes upon a single feasible design region. A sensitivity analysis to ensure that relative importance of a parameter is taken into account is also introduced. Design paths from the run are stored and can be used by the designer to explore not only the optimum solution provided by the method but also solutions which are biased towards each of the design objectives without further function calls to the design model.