A new method to identify maximum power point for photovoltaic generation

A New Method to Identify Maximum Power Point for

Photovoltaic Generation

Le Tien Phong

Electrical Faculty

Thai Nguyen University of Technology

Thai Nguyen, Viet Nam

[email protected]

Ngo Duc Minh

Electrical Faculty

Thai Nguyen University of Technology

Thai Nguyen, Viet Nam

[email protected]

Nguyen Van Lien

Electrical Institute

Ha Noi University of Technology and Science

Ha Noi, Viet Nam

[email protected]

Abstract—In power systems having photovoltaic generations,

controlling the optimal energy is executed by maximum power

point algorithms. Most of these algorithms require the

information of equivalent photovoltaic parameters such as series

and parallel resistor, photo-generated current, reserve saturation

current, thermal voltage. Unfortunately, this information is

unknown and complicated methods are needed to approximate

the value of these parameters. Considering that fact, this paper

uses Newton-Raphson method to calculate fast unknown

parameters and proposes a new algorithm based on half-fold

diagram method to identify the value of current, voltage at the

maximum power point of a photovoltaic generation when solar

irradiance and temperature change. Results receiving from

proposed algorithm are exact, reliable, and it can help reducing

the calculation effort for the processor and enhancing efficiency

for whole of this system.

Keywords—Half-fold diagram; maximum power point tracking;

maximum power point; Newton-Raphson algorithm; offline

technique; online technique; photovoltaic generation; photovoltaic

parameters.

I. INTRODUCTION

Controlling operation modes for a photovoltaic generation

(PVg) at maximum power point (MPP) has an important role to

enhance efficiency for this type of generation. This is also the

problem solved by many researchers in theory and experiments

recently because of the complexity of mathematics model and

the operation character of PVg. So, maximum power point

tracking (MPPT) can be classified by two main types called

online and offline [1], [2].

Offline techniques such as Open Voltage (OV), Short

Circuit (SC), Artificial Neural Network (ANN), Fuzzy Logic

(FL),... base on the fundamental mathematics model and

known parameters of PVg to have values of parameters at MPP

[1], [3]-[6]. On the other hand, online technics such as Perturb

and Observe (P&O), Extremum Seeking Control (ESC),

Incremental Conductance (INC),... use the information about

current, voltage in real time to find MPP by creating

perturbation to have values of parameters at MPP [1], [2], [5]-

[7].

Each technique has own advantages, disadvantages

individually [1], [2], [6]. It depends on the way of using input

parameters and the complexity of that technique. For example,

OV or SC method has to open or short-circuit the two terminals

of PVg to predict MPP while ANN and FL techniques have to

collect a real operation data to recalculate parameters of

transfer function at each time. It makes the program quite

complex and spends much time to execute. Moreover, online

techniques have an advantage that is easy to execute but they

make some perturbation for whole of system because of always

changing pulse control although it doesn’t have any change of

PV input parameters, so all energy of PVg can’t be exploited at

anytime.

The current tendency to find MPP is that combines online

and offline techniques [1], [7]. It means that, at anytime, by the

information about irradiance S and temperature T, control

program has to manage completely about the information about

MPP using offline techniques. Online techniques only use to

reverify and help enhancing the degree of accuracy in the

process of finding MPP.

One problem made by offline techniques is that has to

define values of parameters such as series and parallel resistor,

photo-generated current, reserve saturation current, thermal

voltage in the mathematics model of PVg which aren’t

promulgated widly. Reference [8] calculated these parameters

using Gauss-Seidel algorithm but it was made easier by

omitting non-linear elements in I-V relationship at each time

which are responsible to each pair of value (S, T).

The purpose of this paper is that presents the way to

calculate parameters of PVg fast and exactly using Newton￾987-1-4673-6547-5/15/$31.00 ©2015 IEEE

987-1-4673-6547-5/15/$31.00 ©2015 IEEE

987-1-4673-6547-5/15/$31.00 ©2015 IEEE