Answers are grounded in this paper and cite their sources.
A parameter identification algorithm for linear systems is presented. It is based on smoothing test data with successively improved sets of system model parameters. The smoothing pass through the data provides all of the information needed to compute the gradients of the smoothing performance measure with respect to the parameters. The parameters are updated using a quasi-Newton procedure, until convergence is achieved. The advantage of this algorithm over standard maximum likelihood identification algorithms is the computational savings in calculating the gradients. This approach is extended to identify one set of parameters from several test runs. The algorithm is compared to other time-domain algorithms using a simple example with computer simulated data. The performance of this algorithm is demonstrated in identifying the parameters of a linear model describing the rigid body dynamics of the DLR BO-105 research helicopter from flight test data. The identification results are presented and compared to recently published models using maximum likelihood and frequency-domain algorithm. The identified models are in good agreement with each other.
No takes yet. Share an insight, caveat, or question.
Idan et al. (1992) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: