Selective sensitivity in the frequency domain--II. Applications

✍ Scribed by Uwe Prells; Yakov Ben-Haim

• Publisher: Elsevier Science
• Year: 1993
• Tongue: English
• Weight: 605 KB
• Volume: 7
• Category: Article
• ISSN: 0888-3270
• DOI: 10.1006/mssp.1993.1032

✦ Synopsis

The concept of selective sensitivity is applied in simulation to the identification of a complex truss. It is shown that, of the 36 adjustment parameters in the model, the number of adjustment parameters which must be simultaneously estimated is never more than two. In the presence of additive zero-mean random output errors, the selectivelysensitive adaptive multihypothesis estimation leads to asymptotically unbiased estimates. Furthermore, it is shown that simultaneous estimation of two parameters can be reduced to single-parameter estimations by applying static loads. In the sequence of estimates, the estimates of the current step generally depend on the estimates of the previous step. We demonstrate that, under certain conditions, this sequential dependence can be reduced.