Fast Multidisciplinary Design Optimization in the Development of Mechatronic Systems

Multiobjective multidisciplinary optimization supports the development of mechatronic systems. A suitable approach is required for a short calculation time and sufficient results. All aspects of the system model (mechatronic system, cost model and time optimal control problem) are incorporated into one nonlinear optimization model, following the all-at-once approach. The dynamic simulation is discretized in time and optimization variables are introduced for the state at each time step. Formulating the problem in an algebraic modeling language and solving it by the interior point method allows very fast solution times. This enables a fast turnaround time during the preliminary design phase in the product development. Sensitivities of the objective with respect to model parameters and design constraints are generated by the solver and used to guide the modeling and development process. Using these sensitivities, the model can be improved where necessary while keeping the model complexity low by simplifying less important parts. As example an electromechanical actuating system is considered, in which the rotary motion of a motor is converted to a translational movement with a gate-tape gear.