System Optimization is the adaptation of an existing system with a fitting parametrization according to a previously determined optimal system structure. The optimal spatio-temporal decomposition into subsystems, their patterns of interaction and the propagation of uncertainties are central questions. Like in Systems- and Requirements-Engineering, optimization has to be holistic.
System Optimization can add profound value to a variety of applications like interdisciplinary engineering design, the high connectivity of large scale economic models or the multi-scale complexity of intelligent systems.
At HRI-EU, we approach System Optimization with stochastic optimization methods inspired by biological evolution. Biological evolution is “situated design” or “optimization in-vivo”. This means that the design and operation process are occurring concurrently in the same environment. In our research, we focus on Many Objective and Many Disciplinary Optimization tasks. For system characteristics, we explore Robustness as inherent property and Evolvability, enabling systems to continuously evolve.
Cooperative Charging of Electric Vehicles
With a strong increase in the number of battery electric and plug-in hybrid vehicles, smart charging approaches are needed to reduce the negative impact of electric mobility on the stable operation of the power grid to reduce social costs as well as environmental impacts.
Our cooperative approach enables the driver of an EV (Electric Vehicle) to choose their preferred charging conditions. The decision incentives provided in form of dynamic charging price offers, reward choices that are beneficial to the electric grid and the charging operator. For instance, a longer charging duration results in a lower electricity price for the driver and more flexibility in the charging process for the operator.
This dynamical determination of optimal price offer and charging profiles combines mixed integer linear programming with evolutionary multi-objective optimization. Additional constraints like price fair-ness can be added into the price finding optimization, creating a smart energy service system in which driver, energy provider and mobility provider are engaged.
This approach enables the utilisation of machine learning and optimization to develop and test robust energy management systems.
For more information
S. Limmer and T. Rodemann, “Peak Load Reduction through Dynamic Pricing for Electric Vehicle Charging”, International Journal of Electrical Power & Energy Systems, 2019