8
Control Characteristics of a Fibonacci- Search-based Maximum Power Point Tracker When a Photovoltaic Array Is Partially Shaded Masafumi Miyatake, Takeshi Inada, Isao Hiratsuka, Hongyan Zhao, Hisayo Otsuka, Motomu Nakano 1

Control Characteristics of a Fibonacci- Search-based

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Control Characteristics of a Fibonacci- Search-based

Control Characteristics of a Fibonacci-Search-based Maximum Power Point Tracker When a Photovoltaic Array Is

Partially Shaded

Masafumi Miyatake, Takeshi Inada, Isao Hiratsuka, Hongyan Zhao,

Hisayo Otsuka, Motomu Nakano

1

Page 2: Control Characteristics of a Fibonacci- Search-based

Department of Electrical and Electronics Engineering, Sophia UniversityKioi-cho 7-1, Chiyoda-ku, Tokyo 102-8554, Japan

E-mail : [email protected] : http://power.ee.sophia.ac.jp/

宮 武 研 究 室宮 武 研 究 室MIYATAKE LABORATORY

Power Electronics

Provides Energy

with Preserving Environment.

powerful !!

2

Page 3: Control Characteristics of a Fibonacci- Search-based

Introduction

V

I

maximum area

maximum powervariation of insolation

variation of optimal voltage

General Background

Objectives・Simple and fast algorithm

・small amount of calculation

・intuitive parameter tuning

・Reliable algorithm

・able to find the global maximum power point

・Applicable to any type of power conditioners

・independent of the circuit configuration

→ Various applications are expected.

・Needs for photovoltaic (PV) power

・reduction of CO2 emission

・distributed power generation

・Features of PV power generation

・dependent on weather

・change of the operating point

generating maximum power

→ Maximum Power Point Tracker (MPPT)

Partial Shadow Effect・Partially shaded PV array

・with series connection of PV modules

・two or more maximum power points

・difficulty in finding the global maximum

power point

3

Page 4: Control Characteristics of a Fibonacci- Search-based

Fundamentals of the MPPT

cn+1 = cn + cn−1

c1 = c2 =1

c1,c2,L cN

simple and fast !

Using Fibonacci Sequence・Using simple optimization methods

・no need for differential operation

・no need for assumption of function itself

→ Difficult to use some methods applying

Maximum Principle

・Two types of optimization techniques

・one dimensional line search algorithms

・very simple

・Fibonacci search algorithm

・(golden section search algorithm)

・introducing meta-heuristics

・easy to find global maximum point

・Particle Swarm Optimization (PSO)

Approach

only comparing and restricting

・The ratio of divisions satisfies the Sequence.

P

x

optimal

widening

shifting direction

If the optimal point is out of the search range...

4

Page 5: Control Characteristics of a Fibonacci- Search-based

Improvement of the Proposed MPPT

direct insolation

shadowedinsolation

V1

V2

I

I

V V

I

V1V2 V1+V2

Effect of partial shade

Partial shade and series modules often generate two or more maximum points.

P

x

optimal

widening the range as wide as possible

sudden change of PInitialization of search range

in order to find global maximum power point easier.

detected

example of P-V characteristics

need for wide range search

5

Page 6: Control Characteristics of a Fibonacci- Search-based

Experimental System

white sheet

To make artificial shadowCircuit Configuration

An Intelligent Power Module (IPM) is used as a boost converter.

It can control three PV arrays simultaneously.

realizing sudden change of insolation

6

Page 7: Control Characteristics of a Fibonacci- Search-based

Results of the Experiments

raised by 44%

shaded shaded

・The improved MPPT does NOT always finding

the global maximum theoretically.・It can raise output power with partially shaded

array in most cases. ・balancing simplicity and performance

・reasonable MPPT

・Fast power fluctuation can be found.measured P-V characteristics

local

global

resistance loadbattery load

local maximum

initialized

7

Page 8: Control Characteristics of a Fibonacci- Search-based

Further Works

BPFPV

power

+

-

PI

WTG power

+

+

WTGMPPT

converter circuit

LPF

wind speed and WTG rotor speed

0

+

+

command of power variation

PI controller

Low Pass Filter

Band Pass Filter

power fluctuation canceller

maximum power point tracker

agent

maximum point

swarm

agent

agent

agent

Application of PSO method

Hybridization of generators

simulation result

PV-WTG hybrid system

・Controlling the movement of agents

・easy to find the global maximum power point

・applicable to multi-dimensional control

independent operating voltage

uvw

IPM+

-

+

-+

-

+

-

more difficult conditions with fewer sensors

Fast power fluctuation of PV is cancelled with a WTG

8