Real time economic dispatch for power networks: A distributed economic model predictive control approach

Fast power fluctuations pose increasing challenges on the existing control structure for power networks. One challenge is how to incorporate economic performance and constraint satisfaction in the operation. Current state of the art controllers are based on online steady-state optimization algorithms, which guarantee optimal steady-state performance. A natural extension of this trend is to consider economic model predictive control (EMPC), a dynamic optimization method, which can give guarantees on transient economic performance and constraint satisfaction. We show that the real time economic dispatch problem can be posed as an EMPC problem and provide corresponding transient guarantees for feasibility, stability and economic performance. Next, we show how the corresponding optimization problem can be solved online with dual distributed optimization and study stopping conditions due to real time requirements. This leads to an inexact solution of the optimization problem and we provide guarantees for this inexact distributed EMPC. Finally, we present simulation results showing constraint satisfaction and superior economic performance of the EMPC approach compared to state of the art solutions.