欢迎来到天天文库
浏览记录
ID:58140154
大小:1.31 MB
页数:7页
时间:2020-04-24
《基于双粒子群协同优化的ECT图像重建算法-论文.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、计算机研究与发展DOI:10.7544/issnlO00—1239.2014.20131006JournalofComputerResearchandDevelopment51(9):2094—2100,2014基于双粒子群协同优化的ECT图像重建算法赵玉磊郭宝龙吴宪祥王湃。(西安电子科技大学空间科学与技术学院西安710071)。(西安科技大学电气工程与控制学院西安710054)(zhaoyulei一2008@aliyun.corn)ImageReconstructionAlgorithmforECTBasedonDualParticleSwarmColl
2、aborativeOptimizationZhaoYulei,GuoBaolong,WuXianxiang,andWangPai(SchoolofAerospaceScience&Technology.XidianUniversity,Xi'an710071)。(SchoolofElectricandContro1.Xi'anUniversityofScience&Technology,Xi’an710054)AbstractSincethesensitivityfieldinthecapacitancesensorofelectricalcapacita
3、ncetomographysystemis“softfield”.andthe“softfield’’natureisignoredbythetraditionalimagereconstructionalgorithms,thereisbottleneckinimprovingtheimagingaccuracyforthealgorithms.Tosolvetheproblem,basedontheanalysisofthedistributionofsensitivityfieldandthediscussionofthe“softfield”eff
4、ectanditsimpactontheimagereconstruction,anovelimagereconstructionalgorithmisproposed,whichisdualparticleswarmcollaborativeoptimization.Inthealgorithm,toeliminatetheimpactgeneratedbyignoringthe“softfield”nature.aprioriconditionisusedtoconstructthefitnessfunctionofparticleswarmoptim
5、ization.Theprioriconditionsunderthedifferentflowpatternsareobtainedbytheleastsquaresupportvectormachine.Meanwhile,byintroducingtheIotka—Volterramodel,anewcooperative—competitiveschemeisdiscussed.Thediversityofparticlesisincreasedbyintraspecificandinterspecificlearningandcompetitio
6、n.Sothealgorithmimprovestheglobalconvergenceandconvergencerate.Theexperimentalresultsillustratethatthisalgorithmnotonlyhashigherimageprecisionandstrongerconvergence,butalsoisresistanttotheinterferenceofnoiseinthemeasurementsigna1.Keywordselectricalcapacitancetomography;imagerecons
7、tructi0n;particleswarmoptimization;Lotka—Vloterramodel;prioriconditions摘要由于电容层析成像(electricalcapacitancetomography,ECT)系统中电容传感器的敏感场是“软场”,然而传统的图像重建算法是在忽略“软场”效应的条件下构建的,因此在提高成像精度上存在瓶颈.针对该问题,在分析敏感场分布,并讨论“软场”效应及其对图像重建产生的影响的基础上,提出一种基于双粒子群协同优化的图像重建算法.该算法通过引入用于构造粒子群优化适应度函数的先验条件,消除了因忽略“软场”效
8、应而产生的影响,并通过最小二乘支持向量机得到不同流型下的先验条件.
此文档下载收益归作者所有