Multi-objective mapping optimization for distributed Integrated Modular Avionics
Integrated Modular Avionics (IMA) is a promising resource sharing concept in the avionics domain. It reduces weight and cost of current aircraft. Saving potentials rise with the evolution of IMA, e.g. distributed IMA (DIMA). However, the inherently rising number of functions and the technological degrees of freedom make the selection of the optimal DIMA architecture a complex task. Complexity can be reduced by supporting the IMA designer with solving subtasks of the design process as mathematical optimization problems. Important subtasks are the software and hardware mapping, i.e. for a given aircraft structure finding the optimal distribution of DIMA devices, and for a given hardware topology finding the optimal distribution of aircraft system functions. It is shown how to solve these problems as multi-objective mathematical programming problems for mass and cost objectives, whereby optimizer input is automatically derived from a DIMA architecture domain model. The proposed methods are demonstrated on a reference DIMA architecture consisting of several aircraft systems and a state-of-the-art DIMA platform, which has been constructed and optimized manually. Optimization results reveal major optimization potentials, all possible trade-offs, as well as correlations between objectives. In summary, a valuable support of the DIMA design process by multi-criteria optimization with a practical example is presented.