Adaptive intelligent integration detection method of radar range extension target

Abstract

The invention discloses an adaptive intelligent integration detection method of a radar range extension target and belongs to the radar signal processing field. The adaptive intelligent integration detection method of the radar range extension target aims at overcoming the defects that the existing range extension target statistical detection method is poor in adaptive capability to the clutter and target environment and the like under the condition of the unknown clutter background. According to the adaptive intelligent integration detection method of the radar range extension target, normalized processing is performed on a radar echo one-dimensional range profile, integration optimization is performed through data domain description characteristics of a clutter one-dimensional range profile, a detection method of the constant false alarm rate of a one-class support vector machine is established, a kernel function parameter is adjusted to ensure constant false alarm rate characteristics of the detection method, and a clutter database is updated according to a detection result; statistical modeling does not need to be performed on the clutter background; the detection method is suitable for the positive and negative sample quantity seriously asymmetrical radar target detection environment; the false alarm control capability is strong, the computation is less, the engineering implementation is convenient, and the popularization and application value is high.

Description

1. Technical field [0001] The invention belongs to the field of radar signal processing, and in particular relates to an adaptive intelligent fusion detection method for a range-extending target. 2. Background technology [0002] Compared with conventional low-resolution radars, high-range resolution radars have obvious advantages in anti-jamming, anti-reconnaissance, precise detection and imaging, high-precision tracking, target recognition, etc., and have received extensive attention in modern military and civilian fields. And applications have become an important direction of modern radar development. Different from the "point target" of conventional narrow-band radar, the echoes of high-range-resolution radar targets are distributed in different radial range units, presenting a "one-dimensional range image", forming a "range-extended target". The range-extending target signal has strong non-stationary and nonlinear characteristics. Most of the existing range-extending t...

AI Technical Summary

Problems solved by technology

[0007] In the case of unknown radar clutter background, in order to overcome the sharp drop in performance when the target detection method based on statistical testing does not match the actual environment, the traditional support vector machine method has a high misjudgment rate and cannot overcome the serious asymmetry in the number of positive and negative samples, etc. Insufficient, the present invention adopts single-class support vector machine (OCSVM), proposes a kind of simple and effective distance expansion target self-adaptive intelligent fusion detection method

Images