Coordinating regulation and demand response in electric power grids using multirate model predictive control

In this paper we explore the question of whether it is possible to reduce demand-supply imbalances in the grid, by jointly controlling both the supply-side electric power regulation together with the demand-side energy consumption by residential and commercial consumers. The later is known as demand response and has become a key idea in the Smart Grid vision. Specifically we focus on the potential performance improvements that arise from the complimentary nature of the dynamics of the two: regulation allows for frequent control updates but suffers from slower dynamics; demand response has faster dynamics but does not allow as frequent control updates. We propose a multirate model predictive control approach. This captures the varying dynamics and update rates, as well as the nonlinearities due to saturation and ramp rate limits, and we use a total variation constraint to limit the switching of the demand response signal. The multirate MPC approach results in a quadratic program that must be solved at each time step. We also present a much simpler heuristic controller which delivers reasonably good performance. In addition, we show that our approach has the flexibility to be implemented in the two most likely deployment scenarios: where a direct demand-supply imbalance reference tracking signal is available; or where an indirect market price based imbalance signal is available. Numerical examples are presented to show the efficacy of this joint control approach. Specifically, it is shown that fast demand response can be successfully used in combination with traditional supply-side regulation, to achieve better overall performance.