Progress beyond the state of the art:
As several varying parameters are concerned (4 in our case), a Monte-Carlo based statistical solution is the most appropriate. This solution is not intrusive, robust and less time consuming than other methods such as Stochastic Collocation that are efficient only for low numbers of varying parameters. This MonteCarlo method accommodated well enhancement such as surrogate models based on Kriging and optimization algorithms based on COBYLA.
Using deterministic calculation modules as black boxes is of particular interest for generating statistical data bases and probability assessment. This approach is entirely in line with current trends based on deep learning and Artificial Intelligence. To this extent, such approaches provide a new justification to keep working on the EM-physics reliability of those deterministic computer methods (prediction of the amplitude of the resonances of transmission-lines responses, frequency-varying losses and the loss models to apply on wires and shields…) and improve their commuter efficiency (GPU-CPU, parallelization, use of built-in libraries such as OpenTurns…).
Specific modules have been developed to optimize the calculation procedure and its relevance (frequency management, assembling of bundle cross-sections, end-load junction calculation, frequency-dependent losses…. Particular efforts have been made on the calculation of Transmission-Line p.u.l. parameters (CLIG and Smolyak-based interpolation technique).
Impact:
The impact of the project is also in terms of industrial process since this application addresses installation of cable harnesses in aircraft for which safety must be maintained while optimizing room inside the aircraft. From a societal point of view, the ANALYST application seeks for more confidence in safety assessment. The way to define high level rules to segregate cable bundles is a fully new approach compared to current installation rules based on empirical rules. Especially, the way the optimum distance as well as derogation distances are evaluated are all expressed in terms of a probability level, which is related to security confidence (not to trigger EM susceptibility).
The capability of a better defining the cable-bundle constitution, their routes, their protection has a real impact to protect environment and save energy, especially from the saving weight perspective in order to save electrical energy.
The ANALYST activity with it specific orientation to minimum distance optimization addresses both research and testing (optimization techniques, simulation and experimental tests). From this perspective, the potential of the outputs of ANALYST in terms of education are very high.
