Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.

  

A solver-free, privacy preserving method for the problem of optimal Electric Vehicle charging coordination

Author(s):

R. Vujanic
Conference/Journal:

University of Sydney, Australia
Abstract:

In this talk we discuss the task of coordinating the charging pattern of a large fleet of electric vehicles. We identify several constraints limiting controls both on a local level (the batteries) as well as on a global scope (the distribution network). Due to the presence of these constraints, we choose to introduce an optimal controller to tackle the coordination task. The optimization problem we formulate amounts then to a large scale mixed integer program.

We propose to utilize Lagrangian duality to obtain decomposition of the problem. In contrast to the convex counterpart, duality in mixed integer optimization is usually unable to produce optimal solutions directly. However, for the particular optimization problem structure at hand, the duality gap is known to vanish (in relative terms) as the size of the problem increases. Roughly speaking this indicates that, as they grow, these programs resemble more and more convex programs.

We explore the geometric properties that lead to this effect, and exploit them to design a distributed solution method that is simple to implement and is privacy preserving. For the particular case of EV charging coordination, using our method solutions can be obtained by performing simple local computations, avoiding the necessity of deploying optimization problem solvers altogether. We also provide feasibility and performance guarantees for the solutions recovered; these bounds indicate that the quality of the solutions produced by the method improves as the size of the problems considered increases, making it particularly interesting for very large scale instances for which general purpose solvers are generally inadequate.

Year:

2014
Type of Publication:

(06)Talk
Supervisor:



File Download:

Request a copy of this publication.
(Uses JavaScript)
% No recipe for automatically generating a BibTex entry for (06)Talk
Permanent link