A Sequence Period Estimation Method Based on Histogram Screening and Least Squares Fitting

Abstract

The invention discloses a sequence period estimation method based on histogram screening and least squares fitting. Firstly, the obtained period sequence is sorted to calculate the difference value, and the difference value sequence is screened by using the histogram method to obtain the filtered difference value. sequence of values. Then calculate the initial period value of the difference sample, use the period value of each sample to estimate the error of all samples in this subclass and obtain the period deviation, and continuously change the period to obtain a new period and the corresponding period deviation. The period corresponding to the minimum deviation value is selected as the period of this sample, and the period values ​​corresponding to all samples are calculated cyclically. For the period value screening of all samples, the least square method is used for exact fitting to obtain the final sample period. The invention combines the histogram with the least squares, which can very well reduce the requirements for the preprocessing process, and obtain high calculation accuracy of the fitting result, overcome the sensitivity to noise interference, and have good generalization .

Description

technical field [0001] The invention relates to the field of signal processing, in particular to a sequence period estimation method based on histogram screening and least square fitting. Background technique [0002] The estimation of signal period is of great significance to remote sensing detection, communication field, power system and data mining. In the field of signal processing, period estimation of time-series signals can analyze signal components, clarify signal types, and perform effective feature extraction, thereby obtaining useful information. [0003] When estimating the period of a sequence, the most important thing is to calculate the period of the sequence samples and improve the accuracy. Compared with the ideal periodic sample, the actually obtained periodic sequence not only has noise interference, but also has sparse data due to lack of data and discontinuity. Samples in the same period will appear to be distributed with different centers. Therefore,...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem that the sequence period cannot be obtained in the case of noise and uneven sample distribution, the present invention proposes a sequence period estimation method based on histogram screening and least squares fitting, using the statistical histogram method to analyze the original The sequence is screened to achieve the purpose of noise reduction and deletion of abnormal data, and the initial cycle is generated based on the screening value to perform global search, to overcome the sparsity of discontinuous samples, to obtain a global solution, which greatly increases the applicability of the algorithm, and finally uses the minimum The square fitting method is used for periodic precise measurement to improve accuracy and is insensitive to noisy background and coefficient samples

Images