Mmulti agent system approach to solve distributed energy resource allocation problem to increase the stability and reliability of the distribution system

Uditha S Navaratne, Purdue University

Abstract

The modern power distribution system differs drastically from the conventional power distribution system. Smart grid and micro-grid technologies make the modern distribution system more complicated. In modern power distribution network two way power flow is possible due to the installation of distributed energy resources compared to the one way power flow from generation to consumer in conventional power distribution system. Utility companies had full control over the conventional power distribution system. With the more and more DERs in placed by privet sectors and some times by individual home owners managing the power distribution system becomes more and more complex. Maintaining the reliability and the stability of the system is becoming a challenge in modern power distribution system. Addition of DERs has significant effect on the reliability of the network. Planning is critical for reliable and efficient development of the modern distribution system and associated micro-grids. Micro-grid and distribution system planning is computationally complex, time consuming and requires expertise in different disciplines. Multi Agent Systems are widely used to simulate the behavior of different agents collaborating to achieve single or multiple objective functions. Expertise in different areas can be represented by agents with expert knowledge and decision making capabilities. This thesis investigates the use of MAS to automate the distribution energy resource planning process while increasing the stability and reliability of modern distribution systems and associated micro-grids. Planning process use MAS to find the best location/s, best DER type/s and best capacity/s for a given portion of a distribution system.

Degree

Ph.D.

Advisors

Kulatunga, Purdue University.

Subject Area

Electrical engineering

COinS