Browse Publications Technical Papers 2019-01-1354

Regression Techniques for Parameter Estimation of a Synchronous Machine from Sudden Short-Circuit Testing 2019-01-1354

The Air Force Research Laboratory (AFRL) Intelligent Power program has procured several 40 kVA, 400 Hz, brushless synchronous generators at various stages of their life cycle. The history of these machines is unknown with differences in manufacturing, materials, ageing, heating, and non-ideal maintenance procedures contribute to the variability of the measured machine model parameters. The work presented in this paper summarizes efforts by AFRL to improve the fidelity of hardware parameterization of synchronous machine models by leveraging the sudden short-circuit (SSC) laboratory tests. This procedure is used to observed the dynamic response of the synchronous machines and determine the fitted d-axis operational impedances and time constants. Relevant background theory and captured data will be provided to demonstrate the process of approximating machine parameters; however, the primary contribution of this paper is the numerical regression used to curve fit the experimental data. The decomposition of the dynamic response of an electric machine in an SSC event is widely published. The typical approach of the curve fitment is often loosely explained, neglected, or leverages genetic algorithms. Furthermore, most publications discuss the problem from a perspective of larger utility synchronous machines; which have longer decays and the separation between the subtransient and transient time constants is more apparent compared to aerospace machines. In this paper, the authors present an integration regression technique to fit a double exponential decay (with unknown decay time constants) to an unknown steady-state value in the presence of noise. This enables the automated processing of hundreds of laboratory test runs with consistent results. This approach lacks the propagation error associated with isolating the (sub)transient timescales and convergence errors found in genetic algorithms if the solver is not properly initialized/configured.


Subscribers can view annotate, and download all of SAE's content. Learn More »


Attention: This item is not yet published. Pre-Order to be notified, via email, when it becomes available.
Members save up to 40% off list price.
Login to see discount.
Special Offer: With TechSelect, you decide what SAE Technical Papers you need, when you need them, and how much you want to pay.