Some thoughts that may help you design your tracking algorithms.
Many years ago I read a detailed analysis in a technical journal (IBM Systems Journal) describing analysis of some factors affecting insolation (power delivered) from solar panels. I don't have details now, but key points I remember:
- Basic calculation is that collection efficiency varies with cosine of angle between perpendicular to panel face and the sun. Now remember that the peak of a cosine curve is quite flat, so moderate angular errors have little effect.
- Atmospheric pollution and haze affected that dramatically - reducing the sensitivity to the angle - making accuracy of "point to the sun" even less important.
- A reflector, perhaps wider than the panel mounted horizontally in front of the panel increased collection markedly at very low cost.
- If tilt angle is fixed, consider tilting to optimise winter collection - summer collection will be OK.
Hope these ideas help your design. ciao.