Plot clustering method based on fuzzy-C means

An averaging and clustering technology, which can be used in the reflection/re-radiation of radio waves, the use of re-radiation, measurement devices, etc., and can solve problems such as poor tracking effect.

Active Publication Date: 2018-11-16
XIDIAN UNIV +1
View PDF13 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] To sum up, the problem existing in the existing technology is: when the target is relatively dense, the measurement values ​​of multiple targets will be relatively close to each other, and may be divided into one cluster when clustering. When estimating the number of targets and estimating target parameters Large error occurs when different target measurement values ​​may be divided into different clusters, resulting in poor tracking effect; in the case of densely expanded targets, the tracking effect is poor

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
  • Plot clustering method based on fuzzy-C means
  • Plot clustering method based on fuzzy-C means
  • Plot clustering method based on fuzzy-C means

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0083] When the targets are relatively dense, the measured values ​​of multiple targets will be relatively close to each other, and may be divided into one cluster during clustering, which will cause large errors in estimating the number of targets and estimating the target parameters. Coupled with the false alarm measurements generated by nearby electromagnetic interference, it is difficult to meet the requirements. Such as figure 1 As shown, the fuzzy C-means-based clustering method provided by the embodiment of the present invention specifically includes the following steps:

[0084] (1) Measurement value grouping

[0085] Some elliptic gates are defined in the detection area according to the predicted value of the target, and each target corresponds to an elliptic wave gate. The measured value in the elliptic wave gate indicates that the measured value may be generated by the target at that moment. Suppose there are N k tracks exist in the kth frame, resulting in possib...

Embodiment 2

[0120] The method for clustering based on fuzzy C-mean point traces provided by the embodiments of the present invention is the same as in embodiment 1, and the specific steps of the estimated number of targets in the group described in step (3) are as follows:

[0121] Assume that the possible value of the target number c in a group is c 1 , c 2 ,...,c m and c 1 ≤c 2 ≤...≤c m . The number of predicted positions of objects in these elliptic gates is c 0 . The target measurement rates are γ 1 , gamma 2 ,...,γ c0 , the number of measurements produced by the target can be viewed as a Poisson distribution. Therefore, the number of targets is equal to c i The probability estimate for is:

[0122]

[0123] If P(c i ) is not greater than the constant threshold, discard c i . For the efficiency of the algorithm, the selection partition with a small probability will be discarded. Then, calculate the remaining selection partitions step by step.

Embodiment 3

[0125] The fuzzy C-mean point trace clustering method provided by the embodiment of the present invention is the same as embodiment 1-2, and the specific steps of selecting the initial center described in step (4) are as follows:

[0126] Assume that after clutter removal, there are m measured values ​​p k 1 ,p k 2 ,...,p k m exists in the group. c 0 The predicted position of each target is p′ k 1 , p′ k 2 ,...,p′ k c0 , gamma min , γ max is the target measurement rate γ 1 , gamma 2 ,...,γ c0 minimum and maximum values. When the predicted location agrees well with the measured values, the predicted location is used as the initial center. The kernel density estimate for the predicted location is:

[0127]

[0128] Density threshold τ for the i-th target i 'for:

[0129]

[0130] σ' is a normalization constant.

[0131] when greater than τ i ’, the predicted position agrees with the measured value. if c 0 equal to c i , then c 0 The predicted p...

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

The invention belongs to the technical field of a radar tracking system and a similar system, and discloses a plot clustering method based on a fuzzy-C means. The method includes steps: grouping measuring values; removing clutters; estimating the numbers of targets in groups; selecting an initial center; calculating a grade of membership matrix Ut; performing defuzzification; estimating the integrity of a cluster; updating a clustering matrix Ut+1 and a center vi of the cluster; comparing Ut and Ut+1 by employing a matrix norm; if a relation of Ut+1 and Ut shown in the description is less thanor equal to epsilon, stopping the process; otherwise, enabling t=t+1 for a new round of updating; and finally defuzzifying the measuring values according to the clustering matrix. According to the method, the prediction positioning and the measuring rate are considered to find the initial target center; the clustering integrity is considered in an iteration process of an FCM algorithm; and compared with the conventional method, the robustness and the validity are better, the method can be used for correct clustering of multiple maneuvering targets detected by radar, and the targets can be better tracked.

Description

technical field [0001] The invention belongs to the technical field of radar tracking systems and similar systems, and in particular relates to a point trace clustering method based on fuzzy C-means (Fuzzy C-means algorithm for short FCM algorithm), and a radar multi-moving target detection system. Background technique [0002] Currently, the state-of-the-art commonly used in the industry is as follows: radar multi-maneuvering target detection has always been a challenging problem because the number of targets is unknown and time-varying. Due to the low resolution of previous radars, the target only appeared in a single resolution unit; with the improvement of modern radar resolution, the radar beam can collect measurements from multiple reflection points of the aircraft, that is, multiple measurements of one target value. Objects with multiple detections are called "extended objects" or "extended objects", in which case the object is no longer classified as a point object,...

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
IPC IPC(8): G01S13/66G01S7/41
CPCG01S7/41G01S7/414G01S13/66
Inventor 许录平阎博滕欣进丁智青许娜杨升李沐青孙志峰周钇辛吕鹏飞
Owner XIDIAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap