European Commission logo
English English
CORDIS - EU research results
CORDIS
Content archived on 2024-05-30

Hybrid Adaptive Robust Optimization Strategy for EWIS High Dimensional systems

Article Category

Article available in the following languages:

Optimised design of aircraft electrical wiring interconnection systems

As electric systems replace conventional hydraulic and pneumatic systems in aircraft, the complexity of the electrical wiring interconnection system (EWIS) has increased significantly. New software will optimise the designs quickly and efficiently.

Transport and Mobility icon Transport and Mobility
Industrial Technologies icon Industrial Technologies

The EU is moving towards all-electric aircraft. Optimising the high-dimensional wiring system, the operation of which depends on a high number of variables and constraints, is a task that is well suited for modelling and optimisation algorithms. Scientists working on the EU-funded project HAROS-HD (Hybrid adaptive robust optimization strategy for EWIS high dimensional systems) developed the prototype software solution to do just that. The software combines surrogate modelling technologies in creating new surrogate-based optimisation techniques and algorithms. Surrogate models approximate the multivariate behaviour of systems using a limited number of computationally expensive simulations. Surrogate-based optimisation makes use of such models, searching for the optimal design that greatly improves design performance. Scientists developed advanced graph decomposition techniques that help structure information regarding the EWIS problem and extract dependencies. The team used active learning algorithms that intelligently sample the design space to identify feasible regions and produce the most accurate surrogate models of the EWIS discrete system possible. Algorithms for optimisation-based self-organising maps (SOMBAS) were created to conduct quick space filling of the feasible design regions and identification of the boundaries. Additional techniques speed up the execution of the active learning and SOMBAS algorithms. In parallel, scientists improved cross-entropy algorithms and developed a variant pattern search algorithm of Hooke-Jeeves to handle complex and large-scale optimisation problems. The cross-entropy method uses a probabilistic approach to convert discrete problems into continuous ones, while the pattern search algorithm explores the design space more efficiently. The new software solution enables efficient and effective optimisation of high-dimensional systems such as EWIS. Researchers successfully implemented the prototype HAROS-HD system on a cable harness containing 48 wires for an aircraft EWIS, with over 400 design variables and 10 000 constraints. Commercialisation will save important time and money associated with the critical design phase of EWIS optimisation while reducing the weight associated with the wiring through enhanced design. This will increase the competitiveness of the EU aerospace industry while decreasing the environmental impact of flight.

Keywords

Aircraft, electrical wiring interconnection system, high-dimensional, HAROS-HD, design space

Discover other articles in the same domain of application