This paper describes an integrated optimization process of dynamic systems including design parameters and control algorithms. In contrast to known approaches the developed approach is based on an optimization loop including the evaluation of the dynamical behavior of technical systems with respect to the behavior and related properties in time and frequency domain. This includes as well the behavior of the system, the objective function as the formulation of the restrictions to be considered for the dynamical behavior (stationary and instationary). The proposed approach is declared in detail and will be illustrated using two typical technical applications as examples. The first application example is the optimization of the control system of an active magnetic bearing (AMB) rotor system. Hereby the modeling of the AMB rotor system is briefly introduced. An H∞ controller is designed for the control of the system. The performance both in time and frequency domain is optimized in parallel. The algorithm will be explained by simulation examples. The second example is the optimization of the pow-ermanagement system of a fuel cell/supercap-based hybrid electric powertrain. Hereby the modeling of the electric power flow within the powertrain system is demonstrated and its influence on certain system properties like availability, efficiency, and typical aging mechanisms is discussed. The proposed method leads to near-optimal results in a few steps for both of the systems introduced.
Application of the Dynamic Iterative Learning Control to the Heteroplanar Active Magnetic Bearing
