# Re solar cell parameters by ill20582

VIEWS: 9 PAGES: 2

• pg 1
```									                                          Re: solar cell parameters

Re: solar cell parameters

Source: http://sci.tech−archive.net/Archive/sci.energy/2007−10/msg00039.html

• From: "R.H. Allen" <kkarie@xxxxxxxxxxx>
• Date: Thu, 04 Oct 2007 09:56:40 −0400

HI all,

I am working on the topic maximum power point tracking from
solar array under partial shading conditions
Can any one of u tell how to model a solar array under partial shading
conditions in MATLAB simulink (preferably mathematical modelling).

The first thing to understand is that the I−V curve of a solar array is modeled in exactly the same way an
individual solar cell or module is −− both use exactly the same equation, the only difference is the parameter
values you use. As a first attempt, just to keep things simple, I would suggest modeling the array in two parts:
the shaded part and the unshaded part. For the shaded part, just set the light−generated current portion of the
I−V curve equation to zero or something small. If the two portions of the array are in series, each will have to
have the same current flowing through it, and if they are in parallel, each will have to have the same voltage
across it.

A fair amount has been published on using Matlab and Simulink to model PV cells and arrays. I find about 10
papers in each of the last three European Photovoltaic Solar Energy Conferences that mention Simulink
(though I haven't actually looked at them to see whether it is just a mention or something more substantive).
Papers on the topic have been published in the other major PV conferences as well. I suggest starting there for
some ideas on how to proceed.

One conception of a full−blown Simulink model is to have a separate block for each module that models the
I−V curve for that individual module, with the modules connected by resistors that represent the series
resistance in the interconnects. That may also be overkill, depending on your goals and your assumptions.

HOw to find the paramerters of a solar array (i.e series,shunt
resistances)

Ultimately you will need to obtain these through some sort of experimental data. A lot of the parameters you
will need are functions of design and/or manufacturing, so there's no fundamental theoretical way to estimate
them. Depending on the purpose of your project you might be able to use "typical" values that you gather
through a survey of literature. If all you're studying is the effect of shading, you may find that the precise
values of the parameters may have little influence on your conclusions anyway (which is something you
should certainly test).

Re: solar cell parameters                                                                                        1
Re: solar cell parameters

When i searched in the web I found some methods to find
them with different solar intensities.

Are you referring to King's method for obtaining temperature− and intensity−dependent values of the
parameters for an individual PV module? The appropriate coefficients for a fairly large number of modules
are published via some of the same documents that describe the method (which you can find through Google
−− I believe they're at sandia.gov or nrel.gov), so you can use those coefficients to find values for the
parameters you need. These parameters will be for individual modules, of course, not entire PV arrays.

Will it be possible to find the
parameters with one light intensity (i.e with one value of solar
insolation)

Sure. The simplest way is to fit measured I−V data to the equation for the I−V curve, but the accuracy of this
method is questionable. A rather large volume of literature has been published on this topic −− search on
"estimate light i−v parameters" and similar. Some people feel that estimating the parameters from the *dark*
I−V curve is more accurate, so you might look into that too.

One thing that most people agree on is that these parameters can vary with insolation and temperature, but
whether you wish to account for that depends on how detailed and complex you want your model to be. I
suggest keeping it simple and using fixed parameters to start. Do some sensitivity analysis to see how your
conclusions change with small variations in the parameter values. If it makes a difference, you might consider
trying something along the lines of King's method and let the parameter values vary with your assumptions