One-dimensional linear array direction finding method under two-dimensional angle dependence error based on deep learning

Abstract

The invention discloses a one-dimensional linear array direction finding method under a two-dimensional angle dependence error based on deep learning. According to the method, based on the characteristic that deep learning is good at approximating a complex nonlinear function, the problem of two-dimensional angle dependent array error calibration is solved through machine learning. In order to process azimuth angle dependence and pitch angle dependence of array errors at the same time, two-dimensional data acquisition is carried out, namely, different azimuth array steering vectors are acquired at different pitch angles. The measurement data are expanded by adopting local array flow pattern interpolation so as to reduce the over-fitting risk of the deep learning model; and deep learning iscarried out on the data with the lowest signal-to-noise ratio to enable the data to adapt to noisy signals. The method is used for improving the precision of one-dimensional linear array direction finding of the two-dimensional angle dependent array error, reducing the residual array error, correcting the dependence of the array error on the azimuth angle and the pitch angle, and enabling the direction finding method to still have good performance at different pitch angles.

Description

technical field [0001] The invention belongs to the field of array direction finding, in particular to the direction finding of radar, communication, sonar, microphone and other receiver sensor arrays under the existence of array errors, and specifically relates to a two-dimensional angle-dependent azimuth and pitch sensor based on deep learning. A 1D linear array direction finding method for array errors. Background technique [0002] Sensor arrays are widely used in radar, communication, sonar, microphones. The premise of direction finding with sensor array is that the response of the array, that is, the array steering vector is known precisely. In the ideal case of no array error, the responses of each sensor are the same and independent, the position of the sensor is precisely known, and the array steering vector has an exact analytical expression. But this is not the case in practical applications: there are generally three types of array errors in sensor arrays, name...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the deficiencies of the prior art, the present invention proposes a one-dimensional linear array direction finding method based on two-dimensional angle-dependent errors based on deep learning, so as to solve the problem of large residual array errors or high computational complexity and large amount of stored data in the calibration method in the prior art. big problem

Images