基于音频信号的球磨机寻优控制
|
摘要
球磨机具有可磨煤种多,控制简单等特点,但也普遍存在着钢球消耗量大、系统运行自动化程度低、维护维修力度差等问题。如何优化制粉系统的自动控制以降低制粉系统的制粉单耗具有重要的实际意义。而中储式制粉系统是高度关联、大滞后、MIMO的非线性系统,仅以常规的PID控制已经难以取得理想的控制效果,本文提出一个系统的解决方案。对于本控制系统,需要解决的核心技术有两个,即球磨机的多变量控制系统的识别和解耦问题(系统识别和解耦),和球磨机负荷的识别问题。本文主要工作如下:
1.球磨机煤负荷的检测是球磨机实现自动控制、优化运行、节能降耗的关键。本文提出一种利用音频信号1/3倍频程谱识别存煤量的方法,该方法能够得到球磨机在不同煤负荷下噪音频谱的特征频带,从而反映出球磨机煤负荷的变化,为把球磨机音频信号引入煤负荷控制系统打下基础。
2.有效识别及控制钢球磨煤机的料位,是制粉系统实现优化运行、节能降耗的关键。针对钢球磨煤机的运行特点,建立基于信息融合的径向基神经网络模型对其运行中的各种工况进行识别,实现料位的识别。仿真结果表明,网络效果良好,为制粉系统优化运行和自动控制奠定了基础。
3.中储式制粉系统是高度关联、大滞后、MIMO的非线性系统,其动态特性随着运行工况的变化而大范围变化,仅以常规的PID控制已经难以取得理想的控制效果。本文采用模糊内模解耦控制技术对制粉系统的相关参数进行有效控制,确保了制粉系统的安全、高效运行。
4.介绍了PlantScape DCS分散控制系统,讨论了模糊内模控制算法在PlantScape DCS系统中的实现途径,优化了球磨机自动控制在DCS上的应用。
上述控制策略全部应用于中铝河南分公司制粉系统的自动控制,通过了实际运行的检验,确保了制粉系统的安全、高效运行。本文所研究开发的先进控制系统实现了制粉系统各个参数的自动控制,其研究成果在国内处于领先水平。
Ball mill has the characters of grinding with kinds of coals and controlling simple etc, but also exist widely high consumption of the ball, low degree of automation systems, and poor maintenance efforts. How to optimize milling system of automatic control systems to reduce the milling milling consumption has important practical significance. However the coal pulverizing storage system is a highly relevant, large time delay and MIMO nonlinear system and its dynamic properties change widely with circumstance, Only conventional PID control has been difficult to achieve ideal control effect. This paper presents a systematic solution. For the control system, there are two core technologies need to address,one is Ball mill multi-variable control system identification and decoupling problem(System identification and Decoupling )the other is Ball mill load identification .In this paper, the main job is as follows:
1. Measuring the coal load of ball mill is important for carrying out auto control,optimal operation,energy saving and reducing. A method that using audio singal 1/3 octave spectrum identifying coal load is put forward in this paper,which can get the characteristic frequency bands in different coal load,then reflecting the changes in coal load of ball mill and laying the groundwork for the introduction of audio signal ball mill to the coal load control system.
2. It is the key to identify effectively and control the material level of ball mill for the pulverizing system realizing optimal operation and energy saving and consumption reduction. The model of radial basis neural network (RBF)based on radial basis neural network (RBF) is built according to the running charactristics of the ball mill to recognize the various working conditions and the recognition level in the operation. The simulation results show that the network works well, which laid a foundation for the optimal operating and automatic control of the milling operation
3 The milling system are highly correlated, large time delay, MIMO nonlinear systems, and its dynamic characteristics changes with operating conditions on large scale,so conventional PID control has been difficult to achieve ideal control effect. In this paper the effective control using fuzzy and internal model decoupling control technology for the relevant parameters of milling system ensure the safety of the milling system and highly efficient operation.
4 A PlantScape DCS (distributed control system)is introduced and the fuzzy internal model control algorithm of the realization way in the PlantScape DCS system is discussed.Finally,the ball mill automation applications in the DCS is optimized.
All of the above-mentioned control strategy used in the milling of the automation control system in He'nan branch, China Aluminium Corp. ensured the milling system work well in safe and efficient operation. In this paper, the research and development of advanced control systems realize automatic control of the various parameters of the milling system, and the results of their research is the leading level at China and abroad.
1.球磨机煤负荷的检测是球磨机实现自动控制、优化运行、节能降耗的关键。本文提出一种利用音频信号1/3倍频程谱识别存煤量的方法,该方法能够得到球磨机在不同煤负荷下噪音频谱的特征频带,从而反映出球磨机煤负荷的变化,为把球磨机音频信号引入煤负荷控制系统打下基础。
2.有效识别及控制钢球磨煤机的料位,是制粉系统实现优化运行、节能降耗的关键。针对钢球磨煤机的运行特点,建立基于信息融合的径向基神经网络模型对其运行中的各种工况进行识别,实现料位的识别。仿真结果表明,网络效果良好,为制粉系统优化运行和自动控制奠定了基础。
3.中储式制粉系统是高度关联、大滞后、MIMO的非线性系统,其动态特性随着运行工况的变化而大范围变化,仅以常规的PID控制已经难以取得理想的控制效果。本文采用模糊内模解耦控制技术对制粉系统的相关参数进行有效控制,确保了制粉系统的安全、高效运行。
4.介绍了PlantScape DCS分散控制系统,讨论了模糊内模控制算法在PlantScape DCS系统中的实现途径,优化了球磨机自动控制在DCS上的应用。
上述控制策略全部应用于中铝河南分公司制粉系统的自动控制,通过了实际运行的检验,确保了制粉系统的安全、高效运行。本文所研究开发的先进控制系统实现了制粉系统各个参数的自动控制,其研究成果在国内处于领先水平。
Ball mill has the characters of grinding with kinds of coals and controlling simple etc, but also exist widely high consumption of the ball, low degree of automation systems, and poor maintenance efforts. How to optimize milling system of automatic control systems to reduce the milling milling consumption has important practical significance. However the coal pulverizing storage system is a highly relevant, large time delay and MIMO nonlinear system and its dynamic properties change widely with circumstance, Only conventional PID control has been difficult to achieve ideal control effect. This paper presents a systematic solution. For the control system, there are two core technologies need to address,one is Ball mill multi-variable control system identification and decoupling problem(System identification and Decoupling )the other is Ball mill load identification .In this paper, the main job is as follows:
1. Measuring the coal load of ball mill is important for carrying out auto control,optimal operation,energy saving and reducing. A method that using audio singal 1/3 octave spectrum identifying coal load is put forward in this paper,which can get the characteristic frequency bands in different coal load,then reflecting the changes in coal load of ball mill and laying the groundwork for the introduction of audio signal ball mill to the coal load control system.
2. It is the key to identify effectively and control the material level of ball mill for the pulverizing system realizing optimal operation and energy saving and consumption reduction. The model of radial basis neural network (RBF)based on radial basis neural network (RBF) is built according to the running charactristics of the ball mill to recognize the various working conditions and the recognition level in the operation. The simulation results show that the network works well, which laid a foundation for the optimal operating and automatic control of the milling operation
3 The milling system are highly correlated, large time delay, MIMO nonlinear systems, and its dynamic characteristics changes with operating conditions on large scale,so conventional PID control has been difficult to achieve ideal control effect. In this paper the effective control using fuzzy and internal model decoupling control technology for the relevant parameters of milling system ensure the safety of the milling system and highly efficient operation.
4 A PlantScape DCS (distributed control system)is introduced and the fuzzy internal model control algorithm of the realization way in the PlantScape DCS system is discussed.Finally,the ball mill automation applications in the DCS is optimized.
All of the above-mentioned control strategy used in the milling of the automation control system in He'nan branch, China Aluminium Corp. ensured the milling system work well in safe and efficient operation. In this paper, the research and development of advanced control systems realize automatic control of the various parameters of the milling system, and the results of their research is the leading level at China and abroad.
引文
[1]球磨机负荷检测方法综述.沙亚红,常太华,常建平.现代电力.2006:66-68
[2]基于振动频谱分析的球磨机存煤量测控与优化.朱明.东南大学硕士学位论文,2004
[3]钢球磨煤机料位的软测量及其动态过程建模与控制.王颖洁.东南大学硕士学位论文,2005
[4]火电厂球磨机筒内载煤量料位测量的现状与发展.程启明,郭瑞青,杜许峰,郑勇.华尔电力.2008:112-114
[5]解耦控制的现状及发展.马平,杨金芳,崔长春,胡胜坤.控制工程.2005:97-99
[6]球磨机智能控制算法研究及其在DCS中的应用.吴永存.华北电力大学硕士学位论文,2001
[7]球磨机负荷的现状与发展趋势.王泽红,陈炳辰.中国粉体技术.2001,:20-22
[8]钢球磨煤机负荷检测方法的研究及实现.孙丽华,曲莹军,张彦斌,司刚全.热力发电.24-28
[9]韩捷,张瑞林等编著.旋转机械故障机理及诊断技术.北京:机械工业出版社,1996.25-36..
[10]基于振动频谱分析的球磨机存煤量测控与优化.朱明.东南大学硕士学位论文,2004:15-18
[11]声学试验三分之一倍频程控制技术研究.晏廷飞,方贵前.航天器环境工程.2008:463-465
[12]1/3倍频程频谱分析系统的数字化设计与实现.沈秋霞,姚青,陈淑敏,童基均.工业控制计算机.75-78
[13]基于振动频谱分析的球磨机存煤量测控与优化.朱明.东南大学硕士学位论文,2004:7-9
[14]胡广书 数字信号处理--理论、算法与实现[M]北京:清华大学出版社
[15]MATLAB在振动信号处理中的应用.王济 中国水利水电出版社.2006
[16]张明照,刘政波等编著.应用MATLAB实现信号分析和处理.北京:科学出版社,2006.123-176
[17]Kolacz J.Measurement system of the Mill charge in grinding ball mill circuits[J]..Minerals Engineering.,1997.10(12):1329-1338
[18]Zeng Yigen,Forssberg E.Technical note monitoring grinding Parameters by vibration signal measurement---a primary application[J].Minerals Engineering,1994
[19]朱明,吕震中,王建波.基于振动信号频谱分析的球磨机料位监测仪的设计.测控技术,2003.22(12):51-55.
[20]嵌入式球磨机料位检测系统的设计与实现.冯凯.大连海事大学硕士学位论文.2006
[21]磨音检测与处理方法研究.张莲,陈丽.基础自动化.2002.27-30
[22]基于振动信号频谱分析的球磨机料位监测仪表的设计.朱明,吕震中,王建波,张志.测控技术.51-53 gongshi
[23]张福学.传感器应用及其电路精选.北京:电子工业出版社,1992
[24]基于信息融合技术的球磨机三因素负荷检测研究.孙景敏,李世厚.云南冶金.16-19
[25]钢球磨煤机料位的软测量及其动态过程建模与控制.王颖沾.尔南大学硕士学位论文,2005
[26]人工神经网络技术及其应用.覃光华.四川大学博士学位论文。2003
[27]基于径向基函数神经网络的图像识别研究.黄锋.太原理工大学硕士学位论文.2007
[28]BP神经网络在磨煤机料位监测中的应州.陶泯,李英,汪思义.热力发电.2006(09)
[29]径向基函数神经网络学习算法研究.苏美娟.苏州大学硕士学位论文.2007
[30]基于径向基函数神经网络的语音识别.夏妍妍,黄健,尹丽华.大连海事大学学报.2007:157-159
[31]基于信息融合技术的球磨机三因素负荷检测研究.孙景敏,李世厚.云南冶金.16-19
[32]基于数据融合的球磨机最佳负荷工作点判断.田亮,曾德良,刘鑫屏,刘吉臻.热能动力工程.2004.198-200
[33]神经网络信息融合及其在球磨机测量中的应用.边宝峰,马平.仪器仪表用户.63-65
[34]MATLAB在RBF径向基神经网络仿真中的应用.王阳萍,朱正平.甘肃科技.2004
[35]David L.ball and Tames Linas.An introduction to multisensor data Fusion.proceedings of the IEEE,1997
[36]Maria Huffman.Liquid source misted chemical deposition(LSMCD)-a critical review.Integrated Ferroelectrics,1995,(10):35-39.
[37]球磨机新型自动加球机的研制与应用.贾永红.金属矿山.2005
[38]球磨机装球率自动检测系统.凌永发.昆明理工大学博士学位论文.2003
[39]苏卫霞.基于模糊内模的球磨机制粉系统优化控制研究.郑州大学硕士学位论文.2009
[40]TANAKAK,IKEDAT,WANG H O.Fuzzy regulations and fuzzy observers:Relaxed stability conditions and LMI2based designs[J].IEEE Transactions on Fuzzy System s,1998,6(2):250-264.
[41]LEE H J,PARK J B,CHEN G R.Robust fuzzy control of nonlinear system with parametric uncertainties[J].IEEE Transations on Fuzzy System s,2001,9(2):369-379.
[42]陈娟等.多变量时滞系统的解耦模糊内模控制.电机与控制学报,2006,10(2):203-207.
[43]TANAKA K,IKEDA T,WANG H O.Robust stabilization of a class of uncertain nonlinear systems via fuzzy control:quadratic stability,H_∞ control theory and Linear Matrix Inequalities[J].IEEE Transactions on Fuzzy System s,1996,4(1):1-13.
[44]TAKAGI T,SUGENO M.Fuzzy identification of systems and its application to modeling and control[J].IEEE Trans on System s,M an and Cybem etics,1985,15(1):116-132.
[45]基于内模的球磨机控制系统仿真研究.尚雪莲,王东风,韩璞,梁保峰.电力科学与工程.2004
[46]模糊控制及自适应内模控制在电厂中储式制粉系统控制中的应用研究.于国强.东南大学硕士学位论文.2004
[47]电站制粉系统的多变量控制与模糊神经网络的智能控制研究.方斌中国科学院离子体物理研究所博士论文
[48]MATLAB/Simulink与控制系统仿真.王正林等.电子工业出版社.2008
[49]多变量解耦控制方法.闵娟,黄之初.控制工程2005:125-128
[50]先进控制技术在DCS中的应用与研究.李炜.太原理工大学硕士学位论文.2004
[51]球磨机智能控制算法研究及其在DCS中的应用.吴永存.华北电力大学
[52]集散控制系统在火电厂中的应用.张广辉.重庆电力高等专科学校学报.1998
[2]基于振动频谱分析的球磨机存煤量测控与优化.朱明.东南大学硕士学位论文,2004
[3]钢球磨煤机料位的软测量及其动态过程建模与控制.王颖洁.东南大学硕士学位论文,2005
[4]火电厂球磨机筒内载煤量料位测量的现状与发展.程启明,郭瑞青,杜许峰,郑勇.华尔电力.2008:112-114
[5]解耦控制的现状及发展.马平,杨金芳,崔长春,胡胜坤.控制工程.2005:97-99
[6]球磨机智能控制算法研究及其在DCS中的应用.吴永存.华北电力大学硕士学位论文,2001
[7]球磨机负荷的现状与发展趋势.王泽红,陈炳辰.中国粉体技术.2001,:20-22
[8]钢球磨煤机负荷检测方法的研究及实现.孙丽华,曲莹军,张彦斌,司刚全.热力发电.24-28
[9]韩捷,张瑞林等编著.旋转机械故障机理及诊断技术.北京:机械工业出版社,1996.25-36..
[10]基于振动频谱分析的球磨机存煤量测控与优化.朱明.东南大学硕士学位论文,2004:15-18
[11]声学试验三分之一倍频程控制技术研究.晏廷飞,方贵前.航天器环境工程.2008:463-465
[12]1/3倍频程频谱分析系统的数字化设计与实现.沈秋霞,姚青,陈淑敏,童基均.工业控制计算机.75-78
[13]基于振动频谱分析的球磨机存煤量测控与优化.朱明.东南大学硕士学位论文,2004:7-9
[14]胡广书 数字信号处理--理论、算法与实现[M]北京:清华大学出版社
[15]MATLAB在振动信号处理中的应用.王济 中国水利水电出版社.2006
[16]张明照,刘政波等编著.应用MATLAB实现信号分析和处理.北京:科学出版社,2006.123-176
[17]Kolacz J.Measurement system of the Mill charge in grinding ball mill circuits[J]..Minerals Engineering.,1997.10(12):1329-1338
[18]Zeng Yigen,Forssberg E.Technical note monitoring grinding Parameters by vibration signal measurement---a primary application[J].Minerals Engineering,1994
[19]朱明,吕震中,王建波.基于振动信号频谱分析的球磨机料位监测仪的设计.测控技术,2003.22(12):51-55.
[20]嵌入式球磨机料位检测系统的设计与实现.冯凯.大连海事大学硕士学位论文.2006
[21]磨音检测与处理方法研究.张莲,陈丽.基础自动化.2002.27-30
[22]基于振动信号频谱分析的球磨机料位监测仪表的设计.朱明,吕震中,王建波,张志.测控技术.51-53 gongshi
[23]张福学.传感器应用及其电路精选.北京:电子工业出版社,1992
[24]基于信息融合技术的球磨机三因素负荷检测研究.孙景敏,李世厚.云南冶金.16-19
[25]钢球磨煤机料位的软测量及其动态过程建模与控制.王颖沾.尔南大学硕士学位论文,2005
[26]人工神经网络技术及其应用.覃光华.四川大学博士学位论文。2003
[27]基于径向基函数神经网络的图像识别研究.黄锋.太原理工大学硕士学位论文.2007
[28]BP神经网络在磨煤机料位监测中的应州.陶泯,李英,汪思义.热力发电.2006(09)
[29]径向基函数神经网络学习算法研究.苏美娟.苏州大学硕士学位论文.2007
[30]基于径向基函数神经网络的语音识别.夏妍妍,黄健,尹丽华.大连海事大学学报.2007:157-159
[31]基于信息融合技术的球磨机三因素负荷检测研究.孙景敏,李世厚.云南冶金.16-19
[32]基于数据融合的球磨机最佳负荷工作点判断.田亮,曾德良,刘鑫屏,刘吉臻.热能动力工程.2004.198-200
[33]神经网络信息融合及其在球磨机测量中的应用.边宝峰,马平.仪器仪表用户.63-65
[34]MATLAB在RBF径向基神经网络仿真中的应用.王阳萍,朱正平.甘肃科技.2004
[35]David L.ball and Tames Linas.An introduction to multisensor data Fusion.proceedings of the IEEE,1997
[36]Maria Huffman.Liquid source misted chemical deposition(LSMCD)-a critical review.Integrated Ferroelectrics,1995,(10):35-39.
[37]球磨机新型自动加球机的研制与应用.贾永红.金属矿山.2005
[38]球磨机装球率自动检测系统.凌永发.昆明理工大学博士学位论文.2003
[39]苏卫霞.基于模糊内模的球磨机制粉系统优化控制研究.郑州大学硕士学位论文.2009
[40]TANAKAK,IKEDAT,WANG H O.Fuzzy regulations and fuzzy observers:Relaxed stability conditions and LMI2based designs[J].IEEE Transactions on Fuzzy System s,1998,6(2):250-264.
[41]LEE H J,PARK J B,CHEN G R.Robust fuzzy control of nonlinear system with parametric uncertainties[J].IEEE Transations on Fuzzy System s,2001,9(2):369-379.
[42]陈娟等.多变量时滞系统的解耦模糊内模控制.电机与控制学报,2006,10(2):203-207.
[43]TANAKA K,IKEDA T,WANG H O.Robust stabilization of a class of uncertain nonlinear systems via fuzzy control:quadratic stability,H_∞ control theory and Linear Matrix Inequalities[J].IEEE Transactions on Fuzzy System s,1996,4(1):1-13.
[44]TAKAGI T,SUGENO M.Fuzzy identification of systems and its application to modeling and control[J].IEEE Trans on System s,M an and Cybem etics,1985,15(1):116-132.
[45]基于内模的球磨机控制系统仿真研究.尚雪莲,王东风,韩璞,梁保峰.电力科学与工程.2004
[46]模糊控制及自适应内模控制在电厂中储式制粉系统控制中的应用研究.于国强.东南大学硕士学位论文.2004
[47]电站制粉系统的多变量控制与模糊神经网络的智能控制研究.方斌中国科学院离子体物理研究所博士论文
[48]MATLAB/Simulink与控制系统仿真.王正林等.电子工业出版社.2008
[49]多变量解耦控制方法.闵娟,黄之初.控制工程2005:125-128
[50]先进控制技术在DCS中的应用与研究.李炜.太原理工大学硕士学位论文.2004
[51]球磨机智能控制算法研究及其在DCS中的应用.吴永存.华北电力大学
[52]集散控制系统在火电厂中的应用.张广辉.重庆电力高等专科学校学报.1998