Look up "Solar Insolence" for your location.
There are parameters available to calculate how much energy can be produced by the "Standard" of 1000W/M^2 averaged over a year for flat-on-ground plate, fixed-tilt plates, and tracking plates.
For Austin, Tx the average daily insolence for a fixed-tilt plate was about 4.5 kwh/m^2/day, averaged over a whole year. This figure accounts for night, seasons, dust in the air, and weather. It does not account for tree shading or dirt on the panels. Say I installed a 3KW solar system. I don't know exactly what the surface area is, but I don't need to know. The rating means that 1KW/M^2=3KW output for the array, and 0.5KW/M^2 late in the day means 1.5KW out.
I get 4.5 "standard" hours worth of sun on a fixed-tilt panel tilted at an ideal angle. The 3KW rating comes from a "standard" 1KW/M^2 irradiance. So it's 13.5KWH/day output, averaged over a year. Or 4,927.5 KWH/yr total.
However, the panel output goes down substantially with elevated temp, and the standard cell rating method is for 25C, which isn't applicable to most installations. Almost no installations have any measures to cool the cells inside sealed panels. We're a lot hotter during the most productive hours. This can easily reduce the total average output by 20% or more.
The extra energy gained by using a "tracker" is not all that great considering the cost of "moving mounting systems". It could easily cost 2x or more what a fixed system costs, but won't put out 2x more power. This would generally mean it's more cost-effective to add 2x as many panels as opposed to tracking.
There is some seasonal benefit to pointing the panels lower towards the horizon in winter, as the sun is lower then. It's lower cost anyways than a mechnical east-west tracker, and no significant breakdown issues, but still you're going up there on the roof, maybe spending an hour or two, rebolting the mounts' tilt angle for a few $ worth of power over the course of the season.
SO... you may want to create a test jig capable of measuring power output under some sort of "standard" 1KW/M^2 light source to characterize them, if you don't have mfg specs. But the daily output figure, over a year, is much easier to get, and much more accurate, from external data sources. Your data will be full of data confusion from panel dust, temp, seasons (if you don't keep it up for a year), etc.