1: \begin{abstract}

2:   This paper considers the dynamic economic dispatch problem for a

3:   group of distributed energy resources (DERs) with storage that communicate over a

4:   weight-balanced strongly connected digraph.  The objective is to

5:   collectively meet a certain load profile over a finite time horizon

6:   while minimizing the aggregate cost.  At each time slot, each

7:   DER decides on the amount of generated power, the amount sent

8:   to/drawn from the storage unit, and the amount injected into the

9:   grid to satisfy the load.  Additional constraints include bounds on

10:   the amount of generated power, ramp constraints on the difference in

11:   generation across successive time slots, and bounds on the amount of

12:   power in storage.  We synthesize a provably-correct distributed

13:   algorithm that solves the resulting finite-horizon optimization

14:   problem starting from any initial condition. Our design consists of

15:   two interconnected systems, one estimating the mismatch

16:   between the injection and the total load at each time slot, and

17:   another using this estimate to reduce the mismatch and optimize the

18:   total cost of generation while meeting the

19:   constraints. 

20: \end{abstract}

21: