Abstract
Reliable frequency estimation is considered crucial for several industrial applications, such as power control, power system protection and monitoring. In certain applications is required an accurate and fast frequency estimations of the sinusoidal signals. Several algorithms are found in literature able to solve this task. However, their responses under low Signal to Noise Ratio (SNR) are not reliable. In this paper is proposed an optimization by least squares of the well-known zero-crossing technique in order to enhance performance under such condition. To validate the algorithm, it was performed some comparisons with other methods using different values of frequency. The presented technique has shown better results regarding the estimation error.
