Securing critical loads in a PV-based microgrid with a multi-agent system. Taesic Kim 03/04/2013

Transcription

1 Securing critical loads in a PV-based microgrid with a multi-agent system Taesic Kim 03/04/2013 1

2 Reference M. Pipa0anasomporn, H. Feroze, S. Rahman, Securing cri<cal loads in a PV- based microgrid with a mul<- agent system," Renew Energy, 39 (1) (2012), p Appendix H. Feroze, Mul<- agent systems in microgrids: Design and implementa<on, thesis, Virginia Polytechnic Ins<tute and State University, Arlington, Virginia, August 7th,

3 Outline Introduc<on Design of mul<- agent system Cyber- Physical system and their interac<ons Simula<on results and discussion Conclusion Praises Cri<ques Proposed MAS for DC microgrid 3

4 Smart Grid Uses informa<on technologies to improve how electricity travels from power plants to consumers Allows consumers to interact with the grid Integrates new and improved technologies into the opera<on of the grid Issues: Distributed Genera<on, Demand Response and Load Control, Energy Storage, An<cipated Massive Amount of Energy Transac<on, Security 4

5 Micro Grid A subset of an electric power system Single customer; Par<al feeder; full feeder Islanded grid 5

6 Control Methods and Topologies Tradi<onal power system (SCADA) problems Centralized No local supervisory control unit No fault isola<on Relied en<rely on electricity from the main grid Limits communica<on with different DER devices *SCADA: Supervisory Control and Data Acquisi<on 6

7 IDAPS: Intelligent Distributed Autonomous Power Systems Distributed Loosely connected APSs Autonomous Can perform automa<c control without human interven<on, such as fault isola<on Intelligent Demand- side management Securing cri<cal loads 7

8 Opera<ng strategies Normal mode: Micro Grid a part of the local u<lity Op<mal opera<on Outage mode Isola<on Operate autonomously to secure cri<cal loads Intelligent, distributed and autonomous 8

9 Design of the MAS Four agents: control; DER; User; database Goal: island the microgrid and secure cri<cal loads during external fault condi<ons 9

10 Agent architecture and specifica<ons Control agent Monitoring the health of the u<lity network Isola<on and restora<on DER agent Monitoring and controlling DER power and its connec<on DER info: ID, type, power ra<ng (kw), local fuel availability, cost func<on (price) at which users agree to sell, DER availability (maintenance) 10

11 Agent architecture and specifica<ons User agent Customer gateway Monitor voltage, current, ac<ve and reac<ve power consump<on at each cri<cal and non- cri<cal loads Database agent Storing system informa<on, recording the messages and data shared among agents Serves as a data access point for other agents, and keeps track of all available agents and their capabili<es 11

12 Ini<aliza<on: Agent collabora<ve diagram Name server: register and maintaining directory of the agents and their loca<ons Facilitator: contain abili<es of all agents in the MAS 12

13 Agent collabora<ve diagram 13

14 Physical system Distributed energy resources (DER) Solar panels with ba0ery storage During grid- connected mode (normal) VS islanded mode: voltage magnitude, frequency and phase control Assume DER has sufficient energy to secure all cri<cal loads during outage 14

15 Physical system Load profile and their priority A group of three homes Cri<cal load VS non- cri<cal load Cri<cal loads: refrigera<on, freezer, cooking, and ligh<ng Non- cri<cal loads: space cooling, water hea<ng, dryer and others In 2007, Pacific Northwest Na<onal Lab Customers change their usage behaviors in response to variable electricity rate 15

16 Simula<on and discussion Island the microgrid once a fault is detected Secure cri<cal loads during emergency Resynchronize the microgrid to the main grid afer external fault is cleared 16

17 IDAPS mul<- agent system 17

18 Demand power of loads 18

19 Cri<cal load control 19

20 Islanded mode: isola<on and stabiliza<on Autonomous decisions such as seamless transi<on from grid- connected to island mode 20

21 Resynchroniza<on of the microgrid 21

22 Conclusion MAS was designed and implemented in a simulated environment to control and manage a PV- based This paper demonstrates the agent s abili<es to isolate the PV- based microgrid and to secure cri<cal loads during fault condi<ons, as well as resynchronizing the microgrid afer the external fault is cleared 22

23 Praises Detailed descrip<on of the development of MAS for the control of a PV- based microgrid A guide for the prac<cal implementa<on of an agent- based approach for resilient opera<on of a microgrid that has a solar photovoltaic (PV) system coupled with ba0ery storage 23

24 Cri<ques Can t say that database agent is an agent week descrip<on of how to decide the requested power level to DER DER unit has a limited capacity Need to include Trading feature in MAS to be microgrid Advanced components and subsystem having different ownerships will be DER agents or USER agents Ex) PHEV/EV: user agent, Smart Ba0ery System: DER agent 24

25 Appendix: Mul<ple owners Five user agents sharing DER unit ( one DER agent) 25

26 Appendix: Exchange of message 26

27 Appendix: Transi<on process 27

28 Appendix B: Envisioned IDAPS system In real world, there will be many IDAPS systems 28

29 Different owners DERs, Storage, loads will have different owners Each owner would like to make their own decision locally 29

30 Advanced Components and Subsystems Plug- in Hybrid Electric Vehicle (PHEV) Grid- to- Vehicle(G2V) and Vehicle- to- Grid(V2G) Peak load leveling 30

31 Grid- to- Vehicle (G2V) 31

32 DC Microgrid using a Mul<- Agent System- based controller Extending and coordina<ng IDAPS to DC microgrid Several DER agents having different ownerships can make a decision by nego<a<ng and coopera<ng with other agents Deferrable load (PHEVs) will be considered in MAS 32

33 Thank you. Questions, Comments,? 33