正在加载图片...
第12卷第6期 智能系统学报 Vol.12 No.6 2017年12月 CAAI Transactions on Intelligent Systems Dec.2017 D0:10.11992/tis.201602015 网络出版t地址:http:/kns.cnki.net/kcms/detail/23.1538.TP.20170626.1739.004html 规则推理与神经计算智能控制系统改进及比较 刘经纬2,赵辉,周瑞,朱敏玲3,王普 (1,北京中医药大学中药学院,北京100029:2.首都经济贸易大学信息学院,北京100070:3.首都经济贸易大学计 算交通科学研究中心,北京100070:4.清华大学信息技术研究院,北京100084:5.北京工业大学电子信息与控制工 程学院,北京100124) 摘要:针对生产生活实践中的智能系统在实施控制过程中关键参数的实时在线智能整定与优化问题与需求,实现 将不同类型人工智能方法与经典的控制方法对接从而构成多种复合控制(A-CC)方法,提出改进算法并进行理论分 析与仿真对比研究。首先实现了基于规则与模糊推理机制的AICC方法,提出了增量式改进算法,进而提出基于小 波神经网络的AI-CC方法,进一步对两类智能系统的稳定性进行理论分析,提出稳定性保证算法,最后对比研究不 同类型的智能系统在智能程度与性能特征方面的差异。研究成果为该领域研究者提供了多种改进的智能控制算法 及其对比参照和理论分析,为该方法在工程实践中低成本地升级并稳定可靠地应用提供可操作方案。 关键词:智能系统;智能控制;先进控制;模糊PID;小波神经网络PID 中图分类号:U621:TP273 文献标志码:A文章编号:1673-4785(2017)06-0823-10 中文引用格式:刘经纬,赵辉,周瑞,等.规则推理与神经计算智能控制系统改进及比较J几.智能系统学报,2017,12(6):823-832. 英文引用格式:LIU Jingwei,,ZHAO Hui,,ZHOU Rui,,etal.Improvement and comparison research between intelligent control sys- tems based on rule based reasoning and neural computation AI methodsJl.CAAI transactions on intelligent systems,2017,12(6): 823-832. Improvement and comparison research between intelligent control systems based on rule based reasoning and neural computation al methods LIU Jingwei2,ZHAO Hui',ZHOU Rui',ZHU Minling',WANG Pu (1.School of Chinese Materia,Beijing University of Chinese Medicine,Beijing 100029,China;2.Information College,Capital Uni- versity of Economics and Business,Beijing 100070,China;3.Computational Transportation Science Center,Capital University of Economics and Business,Beijing 100070,China;4.Research Institute of Information Technology,Tsinghua University,Beijing 100084,China;5.School of Computer Science,Beijing Information Science &Technology University,Beijing 100124,China) Abstract:To solve problems,enable real-time online tuning,and to optimize intelligent system parameters during pro- duction and daily life usage,different artificial intelligent-classical(AI-CC)control methods and systems are proposed using a combination of different types of artificial intelligent methods and classical control methods.Algorithm im- provements are made,and a theoretical analysis comparing the stability and simulation of the Al-CC methods is also im- plemented.This research achieves the following.Implementation of a fuzzy classical-based intelligent control,and pro- posal of an incremental improvement algorithm and further adaptive wavelet neural network classical-based intelligent control (AI-CC).A theoretical analysis and stability ensuring method is also proposed and a comparative study under- taken.This research provides results of different types of improved AI-CC methods and a comparative study for use in further academic research,and is expected to enable low cost upgrades and a reliable solution(theoretical guarantee method)for engineering practitioners. Keywords:intelligent system;intelligent control;advanced control;fuzzy PID;wavelet neural network PID 收稿日期:2016-02-27.网络出版日期:2017-06-26 基金项目:国家自然科学基金项目(71371128.11402006):北京社 2015年,谷歌公司研究人员在Nature"”杂志发 科基金研究基地项目(16 JDYJB028):北京市科技计划 中央引导地方科技发展专项(Z171100004717002):首经 表了一篇关于将人工智能方法应用于49个不同的 贸学术骨干培养计划(00791754840263):首经贸研究生 教学改革项目(00791754310106):北京市属高校高水平 游戏控制系统的研究报告,报告中展示了在这 教师队伍建设支持计划高水平创新团队建设计划 49个智能系统中,计算机经过反复地学习与控制过 (00791762300501). 通信作者:周瑞.E-mail:r.zhou@bucm.edu.cn 程,不但可以学会上述系统的运行规则,还在绝大DOI: 10.11992/tis.201602015 网络出版地址: http://kns.cnki.net/kcms/detail/23.1538.TP.20170626.1739.004.html 规则推理与神经计算智能控制系统改进及比较 刘经纬1,2,3,赵辉4 ,周瑞1 ,朱敏玲3 ,王普5 (1. 北京中医药大学 中药学院,北京 100029; 2. 首都经济贸易大学 信息学院,北京 100070; 3. 首都经济贸易大学 计 算交通科学研究中心,北京 100070; 4. 清华大学 信息技术研究院,北京 100084; 5. 北京工业大学 电子信息与控制工 程学院,北京 100124) 摘 要:针对生产生活实践中的智能系统在实施控制过程中关键参数的实时在线智能整定与优化问题与需求,实现 将不同类型人工智能方法与经典的控制方法对接从而构成多种复合控制(AI-CC)方法,提出改进算法并进行理论分 析与仿真对比研究。首先实现了基于规则与模糊推理机制的 AI-CC 方法,提出了增量式改进算法,进而提出基于小 波神经网络的 AI-CC 方法,进一步对两类智能系统的稳定性进行理论分析,提出稳定性保证算法,最后对比研究不 同类型的智能系统在智能程度与性能特征方面的差异。研究成果为该领域研究者提供了多种改进的智能控制算法 及其对比参照和理论分析,为该方法在工程实践中低成本地升级并稳定可靠地应用提供可操作方案。 关键词:智能系统;智能控制;先进控制;模糊 PID;小波神经网络 PID 中图分类号:U621;TP273 文献标志码:A 文章编号:1673−4785(2017)06−0823−10 中文引用格式:刘经纬, 赵辉, 周瑞, 等. 规则推理与神经计算智能控制系统改进及比较[J]. 智能系统学报, 2017, 12(6): 823–832. 英文引用格式:LIU Jingwei, ZHAO Hui, ZHOU Rui, et al. Improvement and comparison research between intelligent control sys￾tems based on rule based reasoning and neural computation AI methods[J]. CAAI transactions on intelligent systems, 2017, 12(6): 823–832. Improvement and comparison research between intelligent control systems based on rule based reasoning and neural computation AI methods LIU Jingwei1,2,3 ,ZHAO Hui4 ,ZHOU Rui1 ,ZHU Minling3 ,WANG Pu5 (1. School of Chinese Materia, Beijing University of Chinese Medicine, Beijing 100029, China; 2. Information College, Capital Uni￾versity of Economics and Business, Beijing 100070, China; 3. Computational Transportation Science Center, Capital University of Economics and Business, Beijing 100070, China; 4. Research Institute of Information Technology, Tsinghua University, Beijing 100084, China; 5. School of Computer Science, Beijing Information Science &Technology University, Beijing 100124, China) Abstract: To solve problems, enable real-time online tuning, and to optimize intelligent system parameters during pro￾duction and daily life usage, different artificial intelligent-classical (AI-CC) control methods and systems are proposed using a combination of different types of artificial intelligent methods and classical control methods . Algorithm im￾provements are made, and a theoretical analysis comparing the stability and simulation of the AI-CC methods is also im￾plemented. This research achieves the following. Implementation of a fuzzy classical-based intelligent control, and pro￾posal of an incremental improvement algorithm and further adaptive wavelet neural network classical-based intelligent control (AI-CC). A theoretical analysis and stability ensuring method is also proposed and a comparative study under￾taken. This research provides results of different types of improved AI-CC methods and a comparative study for use in further academic research, and is expected to enable low cost upgrades and a reliable solution (theoretical guarantee method) for engineering practitioners. Keywords: intelligent system; intelligent control; advanced control; fuzzy PID; wavelet neural network PID 2015 年,谷歌公司研究人员在“Nature”杂志发 表了一篇关于将人工智能方法应用于 49 个不同的 游戏控制系统的研究报告[ 1 ] ,报告中展示了在这 49 个智能系统中,计算机经过反复地学习与控制过 程,不但可以学会上述系统的运行规则,还在绝大 收稿日期:2016−02−27. 网络出版日期:2017−06−26. 基金项目:国家自然科学基金项目(71371128,11402006);北京社 科基金研究基地项目(16JDYJB028);北京市科技计划 中央引导地方科技发展专项(Z171100004717002);首经 贸学术骨干培养计划(00791754840263);首经贸研究生 教学改革项目(00791754310106);北京市属高校高水平 教师队伍建设支持计划高水平创新团队建设计划 (00791762300501). 通信作者:周瑞. E-mail: r.zhou@bucm.edu.cn. 第 12 卷第 6 期 智 能 系 统 学 报 Vol.12 No.6 2017 年 12 月 CAAI Transactions on Intelligent Systems Dec. 2017
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有