Undetermined Problem

[SeptianK]

I’m conducting a R&D project recently. It belongs to an underdetermined problem (no. of equations < no. of variables). Luckily, several tests can be conducted for the same system; nevertheless, each test is always an underdetermined problem. My question is: how to reach the final solution based upon several tests? Many thanks for your valuable suggestions

[waterboatman]

You could take an average. That might give you a pointer.