Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Solar panel

  1. Jan 27, 2010 #1
    our proposed thesis will be about solar panel..we want to design a circuit that will maximize the usage of the energy of the sun, whenever it is available in greater amount using microcontroller. we would like to ask some ideas to realize this project in order for us to complete our requirements and be able to graduate this coming march. hoping to receive your reply as soon as possible.thanks..GODBLESS.
     
  2. jcsd
  3. Jan 28, 2010 #2

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

  4. Jan 28, 2010 #3
    I did this same project when I was in college - so it's no longer available as a thesis :)

    Easy, cheesy. Nah, I was stepping the output voltage up to the value of the battery bank. I used a boost regulator topology and controlled the PWM. It's a classic linear programming problem.

    1. Start by assuming given PWM value will be optimal (x.nom)
    2. Output a slightly higher PWM value to the boost circuit (PWM = x.nom + h)
    2. Record the output current (y.high)
    3. Decrement the PWM value (PWM = x.nom - h)
    4. Record the output current (y.low)
    5. Ascertain whether the initial guess, x.nom, was high or low using (y.high - y.low)
    6. Increment decrement x.nom accordingly.
    7. Goto step 2

    Now, you may be thinking that I solved your problem for you - I haven't. I gave you the starting point that most people already know. The problem with this method is transients. What happens when a cloud comes by? Typically, the output shoots up as the cloud comes near and drops harshly when the cloud is overhead. How do you deal with this effect on your algorithm?

    Then, there's the battery voltage. You may be thinking it's steady, but again, it's not. your job is to deliver the maximum current to the battery, but optimal PWM will change as the battery is loaded by pumps, refrigerators, whatever.

    Go forth, mull it over. By now, you have the education to do an excellent job, so put your basics in place, make some theories, and test them. Leave extra inputs to your micro and make mistakes fast.

    Best Wishes,

    Mike Fortner
     
  5. Jan 28, 2010 #4

    berkeman

    User Avatar

    Staff: Mentor

    For school projects, we generally allow them to be posted in the technical forums like EE. That's one exception in the Homework Help rules. Especially for more complex, upper-division projects.
     
  6. Jan 29, 2010 #5
    I thought the fact that you only measure the current was a consequence of assuming that the battery voltage was steady during the time it takes to perturb and observe? If you maximize the output power (which would be equivalent to maximizing the current for a load with constant voltage) from the panels, wouldn't this be sufficient if your goal was to maximize charge current to the battery?

    When you say the battery voltage is not steady, is that because you assume it is loaded with equipment that draws some kind of high frequency pulsed current or something of that sort?
     
  7. Jan 29, 2010 #6
    It's not my place to do this work, and your answers don't have to be perfect. Perfection is the enemy of good. But, it's good to see what you can do to address these very real situations.

    You have a method to ascertain improvements in x.nom. Must you react at every measurement, or can you filter / average the new value for x.nom? Is there an optimal sample rate to ignore an inverter (100 or 120Hz ripple)? Does your tracker react quickly or slowly? Does it decide when changes are excessive (Cloud)? Does it use a feed-forward measurement to compensate / ignore anything?

    Again, it doesn't have to be perfect - especially if a feature will make the project late. Just consider what gives you a good, stable tracker. Make mistakes quickly.

    - Mike
     
  8. Jan 29, 2010 #7
    Admittedly I'm drunk, but still -- I can't really relate your reply as to why the battery voltage matters when you're trying to track the maximum power point of the solar panel. Why is it not enough to ignore the battery voltage and simply focus on maximizing the output current of the solar panels? Unless you perturb and observe at a turtle's pace, I can't see how the battery voltage enters the equation.
     
  9. Jan 30, 2010 #8
    thanks for the replies..i actually getting something in your ideas..im looking forward for more ideas..as of now we are planning to program a module for our microcontroller that maximize the usage of energy from the sun for our thesis project..GODBLESS
     
  10. Jan 30, 2010 #9

    uart

    User Avatar
    Science Advisor

    Yes Gnurf I agree. For the purpose of battery charging it is sufficient to just maximize the current.

    BTW. Mike's post following your post does not appear to be a reply to your post. This is just a hunch, but it looks to me like a repsonse to further questions asked of Mike by the OP through the forums PM (private messaging) feature.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook