Date of Award

May 2018

Degree Type

Thesis

Degree Name

Master of Science

Department

Engineering

First Advisor

Lingfeng Wang

Committee Members

Wei Wei, Jun Zhang, Lingfeng Wang

Keywords

Differential Evolution, Particle Swarm Optimization, Plug-in hybrid electric vehicles, Power System, Reliability, Vehicle-to-Grid

Abstract

Plug-in hybrid electric vehicles (PHEVs) feature combined electric and gasoline powertrains with internal combustion engine and electric motors powered by battery packs. The battery packs of PHEVs are mostly charged during off-peaks hours at lower prices and meanwhile absorb the excess power from the grid. Similarly, the power stored in the batteries can also flow back to the electric grid in response to ease the peak load demands.

With the increasing penetration and integration of PHEVs, the reliability of PHEVs is essential to overall power system reliability since the charging mechanisms of PHEVs can influence the reliability of power system. Furthermore, due to the direct integration of PHEVs into the residential distribution network, the PHEVs can work as backup batteries for power systems in case of power outage. Therefore, the reliability analysis of power systems and the ancillary services provided by PHEVs are also proposed in this thesis study.

For the driving pattern of each PHEV, there are three basic elements modeled, which are the departure time, the arrival time and the daily mileage, all represented by probability density functions. Based on these basic concepts, the methodology for modeling the load demand of PHEVs is introduced.

In the proposed system, both the Differential Evolution and the Particle Swarm Optimization are proposed to optimize the control strategies for power systems with integration of PHEVs. Aside from using the minimum cost as a figure of merit when designing and determining the best PHEV charging mechanism, the reliability improvement brought to the power systems by PHEVs and the extra earnings obtained by performing frequency regulation services are also quantified and taken into account. Although the reliability of power systems with PHEV penetrations has been well-studied, the adoption of the Differential Evolution algorithm for minimizing the cost of overall system is not exercised, not to mention a thorough comparative study with other common optimization algorithms. To sum up, the Differential Evolution can not only achieve multiple goals by improving the power quality, reducing the peak load, providing regulation services and minimizing the total virtual cost in this system, it can also offer better results compared with the Particle Swarm Optimization in terms of minimizing the cost.

Share

COinS