首页> 外国专利> METHOD AND A DEVICE FOR RECOGNIZING A PULSE REPETITION INTERVAL MODULATION FORM USING A SUPPORT VECTOR MACHINE, CAPABLE OF MAXIMIZING A MARGIN IN BETWEEN EACH MODULATION FORM USING A DECISION HYPERPLANE

METHOD AND A DEVICE FOR RECOGNIZING A PULSE REPETITION INTERVAL MODULATION FORM USING A SUPPORT VECTOR MACHINE, CAPABLE OF MAXIMIZING A MARGIN IN BETWEEN EACH MODULATION FORM USING A DECISION HYPERPLANE

机译:使用支持向量机识别脉冲重复间隔调制形式的方法和装置,能够使用决策超平面在每个调制形式之间最大化边际

摘要

PURPOSE: A method and a device for recognizing a pulse repetition interval (PRI) modulation form using a support vector machine (SVM) are provided to minimize an error rate in pattern recognition.;CONSTITUTION: A receiving part (10) senses a radar signal to process and deliver to a control part (20). The control part is classified into an SVM training module (22), an input data characteristic vector extracting module (24), a classification result value check module by SVM (26), and a PRI modulation form discrimination module (28) in order to implement a PRI modulation form recognizing method. The SVM training module extracts a feature vector to input to an SVM from training data and includes the extracted feature vector and a PRI modulation form value. The SVM training module classifies the input data to input to SVM and trains the SVM by modulation forms using the input data. The classification result value check module by SVM puts the feature vector into the SVM by modulation forms implemented by the SVM training module. The PRI modulation form discrimination module selects a SVM with the greatest value among discrimination result values, produces a result code according to the selected SVM number, and recognizes the PRI modulation form using the result code.;COPYRIGHT KIPO 2013;[Reference numerals] (10) Receiving unit; (20) Control unit; (22) SVM training module; (24) Input data characteristic vector extraction module; (26) SVM classification result check module; (28) PRI modulation type classifying module
机译:目的:提供一种使用支持​​向量机(SVM)来识别脉冲重复间隔(PRI)调制形式的方法和装置,以最大程度地减少模式识别中的错误率。组成:接收部分(10)感应雷达信号处理并传送到控制部分(20)。为了将控制部分分类,SVM训练模块(22),输入数据特征向量提取模块(24),SVM的分类结果值检查模块(26)和PRI调制形式判别模块(28)实现PRI调制形式识别方法。 SVM训练模块从训练数据提取特征向量以输入到SVM,并且包括提取的特征向量和PRI调制形式值。 SVM训练模块将输入数据分类以输入到SVM,并使用输入数据通过调制形式训练SVM。 SVM的分类结果值检查模块通过SVM训练模块实现的调制形式将特征向量放入SVM。 PRI调制形式判别模块从判别结果值中选择一个最大值的SVM,根据所选的SVM编号生成结果代码,并使用结果代码识别PRI调制形式。; COPYRIGHT KIPO 2013; [参考数字]( 10)接收单元; (20)控制单元; (22)SVM培训模块; (24)输入数据特征向量提取模块; (26)SVM分类结果检查模块; (28)PRI调制类型分类模块

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号