This paper presents the RBF4AERO benchmark technology platform, developed in the framework of the EU-funded RBF4AERO project. The platform enables the so-called Benchmark Management System (BMS) used for benchmark submission and results reporting. The BMS is deployed using three modules, namely the Graphical User Interface (GUI), the Work-flow Manager (WM) and the Benchmarking Database System (BDS) which cooperate during the whole optimization benchmark life-cycle. The GUI is the only component which interacts with the end-user. It enables the optimization benchmark submission, along with the progress, results and computational platform resources monitoring. The configuration of the Optimization (OT) and the Morpher Tool (MT) is a prerequisite for the optimization benchmark submission. In an optimization scenario the WM, which is practically the controller of the system, queries the OT in order to get a table of samples and gives back the results of the simulator (for instance a CFD tool). The evaluated individuals serve as training patterns of a Response Surface Model (RSM) which is, then, used for an Evolutionary Algorithms based optimization. The resulting 'optimal' solution(s) are delivered back to the WM for re-evaluation on the CFD tool. For each evaluation on the CFD tool, when a new geometrical shape is required, the computational grid is morphed using the MT based on radial basis functions.
This paper presents the RBF4AERO benchmark technology platform, developed in the framework of the EU-funded RBF4AERO project. The platform enables the so-called Benchmark Management System (BMS) used for benchmark submission and results reporting. The BMS is deployed using three modules, namely the Graphical User Interface (GUI), the Workflow Manager (WM) and the Benchmarking Database System (BDS) which cooperate during the whole optimization benchmark life-cycle. The GUI is the only component which interacts with the end-user. It enables the optimization benchmark submission, along with the progress, results and computational platform resources monitoring. The configuration of the Optimization (OT) and the Morpher Tool (MT) is a pre-requisite for the optimization benchmark submission. In an optimization scenario the WM, which is practically the controller of the system, queries the OT in order to get a table of samples and gives back the results of the simulator (for instance a CFD tool). The evaluated individuals serve as training patterns of a Response Surface Model (RSM) which is, then, used for an Evolutionary Algorithms based optimization. The resulting 'optimal' solution(s) are delivered back to the WM for re-evaluation on the CFD tool. For each evaluation on the CFD tool, when a new geometrical shape is required, the computational grid is morphed using the MT based on radial basis functions.
Bernaschi, M., Sabellico, A., Urso, G., Costa, E., Porziani, S., Lagasco, F., et al. (2016). The RBF4AERO benchmark technology platform. In Proceedings of the VII European Congress on Computational Methods in Applied Sciences and Engineering (pp.4156-4163). National Technical University of Athens [10.7712/100016.2100.11229].
The RBF4AERO benchmark technology platform
BERNASCHI, MASSIMO;COSTA, EMILIANO;PORZIANI, STEFANO;GROTH, CORRADO;CELLA, UBALDO;BIANCOLINI, MARCO EVANGELOS;
2016-01-01
Abstract
This paper presents the RBF4AERO benchmark technology platform, developed in the framework of the EU-funded RBF4AERO project. The platform enables the so-called Benchmark Management System (BMS) used for benchmark submission and results reporting. The BMS is deployed using three modules, namely the Graphical User Interface (GUI), the Workflow Manager (WM) and the Benchmarking Database System (BDS) which cooperate during the whole optimization benchmark life-cycle. The GUI is the only component which interacts with the end-user. It enables the optimization benchmark submission, along with the progress, results and computational platform resources monitoring. The configuration of the Optimization (OT) and the Morpher Tool (MT) is a pre-requisite for the optimization benchmark submission. In an optimization scenario the WM, which is practically the controller of the system, queries the OT in order to get a table of samples and gives back the results of the simulator (for instance a CFD tool). The evaluated individuals serve as training patterns of a Response Surface Model (RSM) which is, then, used for an Evolutionary Algorithms based optimization. The resulting 'optimal' solution(s) are delivered back to the WM for re-evaluation on the CFD tool. For each evaluation on the CFD tool, when a new geometrical shape is required, the computational grid is morphed using the MT based on radial basis functions.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
