In recent years, Economic Model Predictive Control (EMPC) has received considerable attention of many research groups. The present tutorial survey summarizes state-of-the-art approaches in EMPC. In this context EMPC is to be understood as receding-horizon optimal control with a stage cost that does not simply penalize the distance to a desired equilibrium but encodes more sophisticated economic objectives. This survey provides a comprehensive overview of EMPC stability results: with and without terminal constraints, with and without dissipativity assumptions, with averaged constraints, formulations with multiple objectives and generalized terminal constraints as well as Lyapunov-based approaches.

In this work, we propose a hierarchical distributed model predictive control strategy to operate interconnected microgrids (MGs) with the goal of increasing the overall infeed of renewable energy sources. In particular, we investigate how renewable infeed of MGs can be increased by using a trans- mission network allowing the exchange of energy. To obtain an model predictive control scheme which is scalable with respect to the number of MGs and preserves their independent structure, we make use of the alternating direction method of multipliers leading to local controllers communicating through a central entity. This entity is in charge of the power lines and ensures that the constraints on the transmission capacities are met. The results are illustrated in a numerical case study.

We propose a deep neural network architecture and associated loss functions for a training algorithm for computing approximate Lyapunov functions of systems of nonlinear ordinary differential equations. Under the assumption that the system admits a compositional Lyapunov function, we prove that the number of neurons needed for an approximation of a Lyapunov function with fixed accuracy grows only polynomially in the state dimension, i.e., the proposed approach is able to overcome the curse of dimensionality. We show that nonlinear systems satisfying a small-gain condition admit compositional Lyapunov functions. Numerical examples in up to ten space dimensions illustrate the performance of the training scheme.