Optimization

Simulation-based or black-box optimization is a well-known discipline in the field of applied mathematics and engineering. It is mainly used in the industrial and design fields.

The aim is to improve certain performances of a physical system that are typically only calculable by running complex and expensive numerical simulation programs. These programs, having set some decision parameters (the optimization variables), simulate the behavior of the system or process under study and extract the performance to be improved. Each simulation may in turn require the execution of several tasks in succession, and thus require many computational resources and a lot of time. In this context, standard optimization techniques are often inefficient. It is therefore necessary to use or develop ad hoc algorithms for the specific problem.

Deix is able to provide tools and design methods for the efficient solution of this particular class of problems. In this area, the expertise of Deix is the state-of-the-art in applied research. Through its founding partners, Deix has participated and contributed to the development of one of the most renowned open-source software libraries for simulation-based optimization: click here to access the library. The library includes, among others, algorithms for the solution of global optimization problems, multi-objective optimization problems, integer mixed variable optimization problems.