Fig. 3
From: A simple method for identifying parameter correlations in partially observed linear dynamic models

Estimation results of example 3. The identifiable combinations of the parameters are validated by repeatedly fitting the model to one dataset from x 10 = x 20 = x 30 = 0, u(t) = 25. The curves are from the results of 82 runs each of which with a different fixed value of p 21. a: p 21 + p 31 = 3.0; b: The relationship between p 02 + p 12 and p 03 + p 13; c: The relationship between p 02 and p 12; d: The relationship between p 03 and p 13; e: The relationship between p 12 and p 21; f: The relationship between p 13 and p 31; g: p 03 p 31(p 02 + p 12) + p 02 p 21(p 03 + p 13) = 36.8; h: p 31(p 02 + p 12 + p 03) + p 21(p 03 + p 13 + p 02) = 21.4