You'll need to measure the current, too.
As jm1009 said, you don't need to measure the current. Why would you? When would the optimal voltage position not be the overall optimal position of the panel?
Did you ever finish your project? Any video? Code?
I've seen many projects using a DIY sensor similar to the one in the video that CrossRoads posted (for your convenience, here: http://www.instructables.com/id/Arduino-Solar-Tracker/ )
If I get it right, the sensor takes an average of LDR1 and LDR2 (my image, upper setup), which will be used in the calculations. And an average of 3 and 4. Further it needs an average of LDR1 and LDR3, and an average of LDR2 and LDR4. Having two LDR's to get one value might give you precision, but is it overkilling?
My approach would be like in the lower setup. LDR1 and LDR2 would work as a voltage divider. One analog input (A0) would get the reading. This reading would tell how to adjust the vertical position. 2.5 V should be optimal. A similar setup for LDR3 and LDR4 and another analog input.
One single reading from one analog input would tell how to adjust in one dimension, instead of calculating averages.
From one datasheet of an LDR I read that the resistance of the LDR (in concern) varies from 400 ohm to 1 Mohm. Connecting two LDR's as a voltage divider would make 800 ohm at its least. If that is critical (it would make a 6.25 mA current), one would have to add a 500-1000 ohm resistor in series with each LDR. if 2 Mohm is too much (full darkness) to create a voltage division (which would make the analog input port float), one would have to add a 10-50 Kohm resistor parallel to each LDR.