21
Agent Based Load Management for Microgrid MOHAMED ABBAS ELTAHIR AMJAD MOHEY EDDIN AWAD [email protected] [email protected] SHARIEF FADUL BABIKIR [email protected] IEEE INTERNATIONAL CONFERENCE ON COMMUNICATION, CONTROL, COMPUTING, AND ELECTRONIC ENGINEERING (ICCCCEE 2017) ICCCCEE2017(C) 1 16/1/2016 Department of Electrical and Electronic Engineering - University of Khartoum - Khartoum, Sudan

Agent based Load Management for Microgrid

Embed Size (px)

Citation preview

Page 1: Agent based Load Management for Microgrid

Agent Based Load Management for Microgrid

MOHAMED ABBAS ELTAHIR A MJAD MOHEY EDDIN AWAD M O H A M E D . A . E LTA H I R @ H O T M A I L . C O M A M J A D 5 7 3 5 7 @ H O T M A I L . C O M

SHA RIEF FADUL BABIKIRS . B A B I K I R @ V I R G I N . C O M

IEEE INTERNATIONAL CONFERENCE ON COMMUNICATION, CONTROL, COMPUTING, AND ELECTRONIC ENGINEERING (ICCCCEE

2017)

ICCCCEE2017(C) 1

16/1/2016Department of Electrical and Electronic Engineering - University of Khartoum - Khartoum, Sudan

Page 2: Agent based Load Management for Microgrid

OutlineIntroductionMethodologyTesting and ResultsDiscussionFuture Work

ICCCCEE2017(C) 2

Page 3: Agent based Load Management for Microgrid

INTRODUCTION • The increasing demand on the power grid.

• Infeasibility of large central generation units due to the high cost and the losses in the transmission.

• The high penetration of distributed generation.

• The rapid expansion, the complexity of control…

Smart Grid• Microgrid: is a localized grouping of distributed energy resources and controllable interconnected load all connected to the grid through single point of common coupling.

• Decentralized Control.

• Multi Agent Systems: is a representation of an intelligent society of agents with goals and behaviors to interact with the environment and the other agents.

BackgroundAim

ICCCCEE2017(C) 3

Page 4: Agent based Load Management for Microgrid

INTRODUCTION • An agent is a software entity with certain goals, certain

capabilities, certain knowledge, and autonomous in respect of its actions.

• JADE Platform (Java Agent DEvelopment Framework).

BackgroundAim

ICCCCEE2017(C) 4

Page 5: Agent based Load Management for Microgrid

INTRODUCTION • To apply the structure of Microgrid on particular zone in

Sudan to initiate the investigation of the Smart Grid structure on the Sudan National Power Grid.

• To design a load operator agent.

• To provide suitable integration between the MATLAB environment and JADE platform.

BackgroundAim

ICCCCEE2017(C) 5

Page 6: Agent based Load Management for Microgrid

METHODOLOGY

Shambat Al-Kobaneya Distribution Network.

Load flow study using NEPLAN.

Photovoltaic generation seizing and design.

Components Specification and filter design.

Simulation using MATLAB/SIMULINK.Microgrid DesignAgent Design Integration

ICCCCEE2017(C) 6

Page 7: Agent based Load Management for Microgrid

ICCCCEE2017(C) 7

Al-kobaneya Distribution Network in NEPLAN

Page 8: Agent based Load Management for Microgrid

ICCCCEE2017(C) 8

Microgrid Model in MATLAB/SIMULINK

Page 9: Agent based Load Management for Microgrid

ICCCCEE2017(C) 9

Photovoltaic Subsystem

Page 10: Agent based Load Management for Microgrid

ICCCCEE2017(C) 10

Residential load structure

Page 11: Agent based Load Management for Microgrid

METHODOLOGY

Design Agent Algorithm to do the two functions: Load Shedding (on priority base) and control the PV units.

Design Agent Operator class.

Running the Agent via JADE platform in the container.Microgrid DesignAgent Design Integration

ICCCCEE2017(C) 11

Page 12: Agent based Load Management for Microgrid

ICCCCEE2017(C) 12

Agent Algorithm

Page 13: Agent based Load Management for Microgrid

ICCCCEE2017(C) 13

JADE GUI

Page 14: Agent based Load Management for Microgrid

METHODOLOGY

TCP/IP Protocol.

Server: JADE implements Java Server Socket.

Client: SIMULINK implements dynamic tcpip object through the S-Function block.

Microgrid DesignAgent Design Integration

ICCCCEE2017(C) 14

Page 15: Agent based Load Management for Microgrid

TESTING AND RESULTS

• This scenario mimics the normal daily load profile in the selected area.

• First the demand starts at 1.9 MW.

• Then it increases to reach the peak at 2.6 MW.

• After that it goes down again to 1.9 MW.

• The agent is given 2.18 MW as a limitation power.

Normal and Shedding scenario

PV units control

ICCCCEE2017(C) 15

Page 16: Agent based Load Management for Microgrid

ICCCCEE2017(C) 16

Page 17: Agent based Load Management for Microgrid

TESTING AND RESULTS

• This scenario tests the agent ability to control the two controllable PV units when the load is very low.

• First the load start at 1.7 MW.

• Then it drops to 1.3 MW.

• After that the load goes down to 1 MW.

• And Finally it experience a further decrease to 309 KW.

Normal and Shedding scenario

PV units control

ICCCCEE2017(C) 17

Page 18: Agent based Load Management for Microgrid

ICCCCEE2017(C) 18

Page 19: Agent based Load Management for Microgrid

DISCUSSION The designed system runs smoothly in the normal operation condition satisfying the power quality standards (voltage level, frequency, and harmonic distortion).

Agent was able to perform its function and provide real time response signals through the TCP/IP connection.

The Agent operation on the load and the PV units doesn’t affect the system stability, indicated from the stable voltage level and frequency.

Priority algorithm can be unfair for the load with the lowest priority.

Controlling the PV units was based on the needed demand and doesn’t consider the generation optimization or the economical dispatch.

ICCCCEE2017(C) 19

Page 20: Agent based Load Management for Microgrid

FUTURE WORK Improving the Microgrid modeling by include more design details like the lines impedance from the feeder and to simulate the daily irradiance and temperature variance in the selected zone.

Exploiting the potential of the Multi Agent System by designing agents with dedicated goal to contribute in the decision making, as to introduce agents for the generation optimization.

Implementing the proposed structure for small house scale control where the priority algorithm can be more suitable in controlling the house appliances, which will enable demand side response.

ICCCCEE2017(C) 20

Page 21: Agent based Load Management for Microgrid

Click icon to add picture

ICCCCEE2017(C) 21

Thank you