This paper addresses the problem of estimating the supersonic air velocity from an incomplete measurement of an acoustic sensor array, that is, when some of the data are missing. Most sparsity-inducing estimators for air velocity require the discretization of the estimation grid. When the true air velocity is not on the discretized grid, this leads to grid mismatch, which results in a degradation of the performance. To deal with such an off-grid air velocity estimation problem, an off-grid, iteratively stochastic maximum-likelihood algorithm of estimating air velocity (abbreviated as Sparse Asymptotic Minimum Variance-Stochastic Maximum Likelihood-Air Velocity Estimation) is proposed, based on sparse asymptotic minimum-variance criterion. The proposed algorithm can achieve the off-grid air velocity estimation with high accuracy, eliminating the need of additional grid-refinement techniques. Compared with other off-the-shelf sparse estimation methods, the proposed algorithm is computationally more efficient. Moreover, the proposed algorithm is suitable for estimating the air velocity in case of missing data. Simulations are implemented to illustrate the effectiveness of the proposed algorithm.
Off-Grid Sparse Estimator for Air Velocity in Missing-Data Case
2016
Article (Journal)
English
Off-Grid Sparse Estimator for Air Velocity in Missing-Data Case
Online Contents | 2016
|Missing path recovery method based on sparse license plate recognition data
European Patent Office | 2023
|Adaptive Angular Velocity Estimator (AAS 09-314)
British Library Conference Proceedings | 2010
|