For each parameter present in df, adds columns:
*_ok (logical),
*_status one of "ok", "acima_do_maximo", "abaixo_do_minimo",
*__lim_min and *__lim_max (thresholds used),
*__delta (difference to the relevant limit; >0 above max, <0 below min, 0 if ok).
If multiple limit rows exist for the same parameter, *_ok is TRUE if
any row is satisfied; for status/lim_min/lim_max/delta, the first
satisfied row is chosen; if none satisfy, the row with the smallest
absolute violation (min |delta|) is used.