Turbulent target detection method based on bp neural network multi-class classification

A BP neural network and target detection technology, which is applied in the field of turbulent target detection based on multi-class classification of BP neural network, can solve the problems of high computational complexity, inability to guarantee fast operation, poor performance, etc., and achieve effective turbulent target detection Effect

Active Publication Date: 2021-07-13
SHANGHAI JIAO TONG UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although the existing turbulence intensity detection methods have their own advantages, they also have obvious disadvantages, including poor performance in the case of low signal-to-noise ratio, poor spectral estimation performance under the condition of short echo data length, and methods The calculation complexity itself is high, and the speed of operation cannot be guaranteed

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Turbulent target detection method based on bp neural network multi-class classification
  • Turbulent target detection method based on bp neural network multi-class classification
  • Turbulent target detection method based on bp neural network multi-class classification

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0020] This embodiment includes the following steps:

[0021] Step 1. Simulation data preparation:

[0022] The working wavelength λ of the airborne weather radar adopted in this embodiment is 0.03m, and the pulse repetition period T s is 0.001s, parameter k R About 1W·s 2 / m 2 . for different σ V Values, under different SNR conditions, according to the turbulent echo model to generate echo amplitude sequences with lengths of 8, 16, and 32, respectively, and the corresponding intensity levels form the training set and test set of the BP neural network to observe The influence of the values ​​of signal-to-noise ratio and sequence length on the final test results.

[0023] The turbulence echo model refers to: in view of the meteorological target radar echo signal in the turbulent area is a correlated random process, the correlation coefficient between two adjacent pulse echoes of the same meteorological target is: Among them: T s is the radar pulse repetition period; λ ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

A turbulent target detection method based on multi-class classification of BP neural network. According to the turbulent echo model under different signal-to-noise ratio conditions, the echo amplitude sequence and its corresponding intensity level are respectively generated to form the training set and test of BP neural network. Set, so that the trained BP neural network divides the turbulence intensity into several levels to realize turbulence detection. The present invention does not need to rely on empirical formulas and parameterized models, and utilizes the multi-category classification function of the neural network to use the meteorological target radar echo amplitude sequence as the input data of the neural network training set, and the turbulence intensity level as the training set output data. The learning of echo data can effectively establish the relationship between radar echo and turbulence intensity, so that the neural network can be used to classify turbulence intensity to achieve the purpose of turbulence detection.

Description

technical field [0001] The invention relates to a technology in the field of meteorological environment monitoring, in particular to a turbulent target detection method based on BP neural network multi-category classification. Background technique [0002] At present, most meteorological turbulence detection methods use empirical formulas and parametric models. The correctness of empirical formulas and parameter models greatly affects the accuracy of detection results. Among them, the detection of turbulent targets through definite physical principles is called non-parametric method or empirical formula method. Although the existing turbulence intensity detection methods have their own advantages, they also have obvious disadvantages, including poor performance in the case of low signal-to-noise ratio, poor spectral estimation performance under the condition of short echo data length, and methods The calculation complexity itself is high, and the speed of operation cannot b...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & AuthorityPatents(China)
IPC IPC(8): G01S7/41
CPCG01S7/417
Inventor肖刚张强赵俊豪王彦然刘艺博
OwnerSHANGHAI JIAO TONG UNIV