About us
Methods of Machine Learning and CAE-based Robust Design Optimization in Virtual Product Development
or
"Searching for the Optimum in the Uncertain in Nature and Engineering".
Engineering in its true sense is an optimization process with the goal of improving engineering systems in terms of efficiency and cost while ensuring the safety and reliability of the systems to meet the inherent limits of nature controllability within societal acceptance. Since the advent of digital computing technology in the early 1960s, this process has increasingly moved to the virtual world to reduce development time and cost. Using mathematical optimization algorithms, machine learning algorithms, fast computing power, and realistic simulations, the development of designs and products is performed within Computer Aided Engineering (CAE) to reduce development times and costs. Developing robust designs is a future challenge in virtual product development. The goal of stochastic optimization research is, for example, to minimize the failure probabilities of structures or the number of necessary prototypes.
The IMH offers mathematical method and software competence in the field of machine learning and CAE-based robust design optimization in virtual product development, which makes it possible to develop safe, reliable and optimal designs and products taking into account the avoidable scattering effects and system properties. In particular, it is possible that within the optimization in the CAE processes the unavoidable uncertainties and scattering of the actions, the system properties and the manufacturing processes can be considered. Beyond the state-of-the-art methods, IMH develops numerical algorithms in the area of partial differential equation solution, optimal control, stochastic analysis and multidisciplinary optimization on request.
