OptiCat: a versatile open-source optimisation platform for experimental design

F. Clerc , D. Farrusseng (1) , C. Mirodatos (1)

1 IRCELYON - Institut de recherches sur la catalyse et l'environnement de Lyon

A new open-source software for experimental design is presented. It consists at a platform which enables to design specific optimization algorithms by simple drag and drop from a toolbox. Complex workflows can be implemented and tested within a few minutes by experimentalists. OptiCat can integrate MS Excel©, MatLab© and Statistica© functionalities for batch calculations. Optimization results on a custom benchmark are presented as examples. Four different algorithms are studied: genetic algorithm, evolutionary strategy, Tabu Search and simulated annealing. Advantages to use OptiCat in the frame of parallel experimentation in Material Science and Catalysis are addressed.

Mots clés: Design of experiments; Genetic algorithm; Evolutionary strategy; Open-source software; Simulated annealing; Benchmark; Optimization

Domaines: Environnement et Société; Catalyse

F. Clerc, D. Farrusseng, C. Mirodatos. OptiCat: a versatile open-source optimisation platform for experimental design. Chemometrics and Intelligent Laboratory Systems, 2008, 93 (2), pp.167-171. ⟨10.1016/j.chemolab.2008.05.006⟩. ⟨hal-00308465⟩