SRO_ANN: Toolbox for surface response optimization using RBF

SRO_ANN is an integrated MatLab toolbox for multiple surface response optimization using radial basis functions

SRO_ANN, a MatLab toolbox for implementing multiple surface response optimization by artificial neural networks (SRO_ANN) is presented. Radial basis functions, a type of artificial neural networks, are applied through an easily managed graphical user interface. A detailed description of the interface is provided, including a simulated and two literature examples which allow one to show the potentiality of the software. Examples are povided with experimental data corresponding to: (1) the maximization of the research octane number (RON) of fuels, influenced by three factors (reaction temperature, operating pressure and low liquid hourly space velocity), and (2) the optimization of the calcification process for diced tomatoes, evaluated through three different responses (calcium content, firmness and pH), which are affected by three factors (calcium concentration, solution temperature and treatment time). The results show that the application of a nonparametric tool can enhance the performance of optimization modeling tasks.