| 基于支持向量机和粒子群算法的电子鼻伤口感染检测 |
| |
| 引用本文： | 闫嘉,田逢春,何庆华,冯敬伟,贾鹏飞,孙诚,樊澍.基于支持向量机和粒子群算法的电子鼻伤口感染检测[J].世界科技研究与发展,2012(2):261-264. |
| |
| 作者姓名： | 闫嘉 田逢春 何庆华 冯敬伟 贾鹏飞 孙诚 樊澍 |
| |
| 作者单位： | 1. 重庆大学通信工程学院,重庆400030
2. 第三军医大学大坪医院外科研究所,重庆400042 |
| |
| 基金项目： | 重庆市自然科学基金计划重点项目“基于电子鼻技术的人体创伤反应气味模式识别算法研究”（CSTC,2009BA2021）,重庆大学研究生科技创新基金“基于电子鼻的伤口感染检测研究”（20091181A0100326）,重庆大学研究生创新团队建设项目（200909C1016）资助 |
| |
| 摘 要： | 针对传统的伤口感染诊断方法耗时长,操作复杂等问题,提出了一种基于电子鼻和支持向量机(SVM)的方法进行伤口感染检测,分别检测非感染和三种常见病原菌感染的大白鼠伤口顶空气体,然后利用 SVM对实验数据进行识别.同时,鉴于传感器阵列的优化以及 SVM参数选择对其分类准确率有重大的影响,提出一种基于粒子群算法(PSO)的传感器阵列和 SVM参数同步优化方法.实验结果表明,SVM结合 PSO与传统的神经网络以及遗传算法相比,极大提高伤口感染检测的准确率
|
| 关 键 词： | 电子鼻 伤口感染 支持向量机 粒子群算法 传感器阵列优化 参数优化 |
Wound Infection Based on Electronic Nose Combined Support Vector Machine and Particle Swarm Optimization |
| |
| YAN Jia,TIAN Fengchun,HE Qinghua,FENG Jingwei,JIA Pengfei,SUN Cheng,FAN Shu.Wound Infection Based on Electronic Nose Combined Support Vector Machine and Particle Swarm Optimization[J].World Sci-tech R & D,2012(2):261-264. |
| |
| Authors: | YAN Jia TIAN Fengchun HE Qinghua FENG Jingwei JIA Pengfei SUN Cheng FAN Shu |
| |
| Institution: | 1. College of Communication Engineering, Chongqing University,Chongqing 400030,2. State Key Laboratory of Trauma, Burns and Combined Injury, Institute of Surgery Research, Daping Hospital. Third Military Medical University, Chongqing 400042) |
| |
| Abstract: | In order to solve the time-consuming and complicated operation problem in traditional diagnosis method of wound infection, a new method based on the electronic nose （enose） and support vector machine （SVM） is proposed to detect wound headspaee gases of rats noninfeeted and those infected by three types common pathogens respectively. Meafiwhile, owing to the strong impact of optimization of sensor array and parameters selection on the classification accuracy of SVM, an simultaneous optimization method of sensor array and parameters of SVM based on particle swarm optimization （PSO） is presented. The results show that SVM combined with PSO greatly improves the recognition accuracy rate of wound infection ,compared with the traditional neural networks and genetic algorithms. |
| |
| Keywords: | electronic nose wound infection support vector machine particle swarm optimization sensor array optimization parameters optimization |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
