一体化核动力装置广义预测控制方法研究
|
摘要
预测控制是基于预测模型、滚动优化、反馈校正三项基本原理基础上的。由于实际上不可避免地存在着模型误差和环境干扰,这种建立在实际反馈信息基础上的反复优化,能不断顾及不确定性的影响,并及时加以校正,反而要比只依靠模型的一次优化更适应实际过程,有更强的鲁棒性。所以预测控制是针对传统最优控制在工业过程中的不适用性而进行修正的一种新型优化控制算法。
广义预测控制是预测控制中最具代表性的算法之一。它具有如下的特点：(1)基于传统的参数模型,模型参数少。(2)是在自适应控制研究中发展起来的,保留了自适应控制方法的优点,但比自适应控制方法更具有鲁棒性。(3)更适用于工业生产过程的控制。
本文对一体化压水堆的结构特点和动态特性进行了分析,对反应堆-回路进行了必要合理的简化,在此基础上建立了一体化压水堆功率控制系统的动态数学模型,包括中子动力学模型、氙毒动力学模型、反应堆热工水力模型等。根据建立的模型,运用广义预测控制算法对反应堆功率控制进行了研究。
在研究了直流蒸汽发生器的结构和换热等特点基础上,应用可移动边界,集总参数法,根据质量、能量和动量守恒方程,建立了直流蒸汽发生器的数学模型。
由于核动力汽轮机是较复杂的蒸汽能量转换设备和被控对象。将系统的数学模型分成六个部分来描述,它们分别是：进气量计算、调节级后汽室压力的计算、膨胀做功计算、转子模型、转动惯量的计算和负载模型。
最后根据所建立的反应堆、直流蒸汽发生器和汽轮机模型进行一体化核动力装置一、二回路的广义预测控制方法研究。
Predictive control is based on predictive model, circle-optimization and feedback emendation. Because there are model error and environment disturbance, the optimization based on practice feedback can consider the unpredictable effect, and proofread in time. It has more robustness than the once optimization rely on model. So, predictive control is a new optimal arithmetic aim at traditional optimal control can't adapt in the industrial process.
Generalized predictive control is one of the most representative arithmetic in predictive control. It has characteristics as follows:(1) based on the traditional models, model parameter is few. (2) It developed on the self-adaptation control research, and reserved the excellences of the self-adaptation control method, but has more robustness than the self-adaptation control method. (3) It is a control which adapt to industrial produce process.
In this paper, the structure characteristics and the dynamic characteristics of integrated Pressurized Water Reactor are analyzed. The primary loop of the reactor was reasonably simplified, and based on the simplification, the mathematical models of power regulating system were established. Those models included Neutron dynamics model, dynamics model of xenon poison, reactor thermal technology hydraulic model etc. Generalized predictive control was used to control the power of reactor based on the model.
In the basis of researching the structure and heat transfer features of once-through steam generator, and then according to the mass, energy and momentum balance equation, the steam generator's model is built by lumped parameters with moving boundary.
The nuclear-powered turbine is a complex steam-energy conversion equipment and controlled object. So, the math model is compartmentalized six parts, they are air input calculation, regulating stage back chamber pressure calculation, doing work by expansion calculation, rotor model, moment of inertia and load model.
At last, according to the model of reactor, steam generator and turbine were built, the primary and secondary loop system are researched with the generalized predictive control method.
广义预测控制是预测控制中最具代表性的算法之一。它具有如下的特点：(1)基于传统的参数模型,模型参数少。(2)是在自适应控制研究中发展起来的,保留了自适应控制方法的优点,但比自适应控制方法更具有鲁棒性。(3)更适用于工业生产过程的控制。
本文对一体化压水堆的结构特点和动态特性进行了分析,对反应堆-回路进行了必要合理的简化,在此基础上建立了一体化压水堆功率控制系统的动态数学模型,包括中子动力学模型、氙毒动力学模型、反应堆热工水力模型等。根据建立的模型,运用广义预测控制算法对反应堆功率控制进行了研究。
在研究了直流蒸汽发生器的结构和换热等特点基础上,应用可移动边界,集总参数法,根据质量、能量和动量守恒方程,建立了直流蒸汽发生器的数学模型。
由于核动力汽轮机是较复杂的蒸汽能量转换设备和被控对象。将系统的数学模型分成六个部分来描述,它们分别是：进气量计算、调节级后汽室压力的计算、膨胀做功计算、转子模型、转动惯量的计算和负载模型。
最后根据所建立的反应堆、直流蒸汽发生器和汽轮机模型进行一体化核动力装置一、二回路的广义预测控制方法研究。
Predictive control is based on predictive model, circle-optimization and feedback emendation. Because there are model error and environment disturbance, the optimization based on practice feedback can consider the unpredictable effect, and proofread in time. It has more robustness than the once optimization rely on model. So, predictive control is a new optimal arithmetic aim at traditional optimal control can't adapt in the industrial process.
Generalized predictive control is one of the most representative arithmetic in predictive control. It has characteristics as follows:(1) based on the traditional models, model parameter is few. (2) It developed on the self-adaptation control research, and reserved the excellences of the self-adaptation control method, but has more robustness than the self-adaptation control method. (3) It is a control which adapt to industrial produce process.
In this paper, the structure characteristics and the dynamic characteristics of integrated Pressurized Water Reactor are analyzed. The primary loop of the reactor was reasonably simplified, and based on the simplification, the mathematical models of power regulating system were established. Those models included Neutron dynamics model, dynamics model of xenon poison, reactor thermal technology hydraulic model etc. Generalized predictive control was used to control the power of reactor based on the model.
In the basis of researching the structure and heat transfer features of once-through steam generator, and then according to the mass, energy and momentum balance equation, the steam generator's model is built by lumped parameters with moving boundary.
The nuclear-powered turbine is a complex steam-energy conversion equipment and controlled object. So, the math model is compartmentalized six parts, they are air input calculation, regulating stage back chamber pressure calculation, doing work by expansion calculation, rotor model, moment of inertia and load model.
At last, according to the model of reactor, steam generator and turbine were built, the primary and secondary loop system are researched with the generalized predictive control method.
引文
[1]刘聚奎.法国一体化压水堆CAP述评.核动力工程,1990,11(1)：42-47页
[2]Tsutao Hoshi, Masaaki Ochiai.日本一体化反应堆MAX的研究和开发现状.国外核动力,1996,(6)：11页
[3]陈炳德.日本小型核动力反应堆及其技术特点.核动力工程.2004,25(3)：192-197
[4]陈世君.国外舰船一体化压水堆技术发展趋势及其应用前景.国外核动力,2003,(1)：2-11页
[5]李满昌.世界一体化压水堆发展技术现状.国外核动力.1998(4)：1-12页
[6]李满昌,唐传宝.一体化压水堆发展前景.核动力工程.1998,19(4)：370-374页
[7]王成孝.核能与核技术应用.原子能出版社,2002
[8]H. Levent Akin, Vural.Akin. Rule-based fuzzy logic controller for a PWR-type nuclear power plant. IEEE Trance on Nuclear Science and Engineering,1993,114(1):42-54P
[9]Moon Ghu Park; Nam Zin Cho. Self-tuning control of a nuclear reactor using aGaussian function neural network. Nuclear Technology,1995, 110(2):285-293P
[10]Man Gyun Na. Amodel prddictive controller for the water level of nuclear steam generators. Journal of the Korean Nuclear Society,2001,33(1):102-110P
[11]Duk-Hyun Son, Chang-Goo Lee. Design of model predictive controller for water level control in the steam generator of a nuclear power plant. Transactions of the Korean Institute of Electrical Engineers,2001,50(8):376-383P
[12]孙增圻等编著.智能控制理论与技术.北京：清华大学出版社,1997：133-135页,195-196页
[13]赵兆颐,朱瑞安编著.反应堆热工流体力学.北京：清华大学出版社,1992：30-32页
[14]凌备备,杨延洲编著.核反应堆工程原理.第二版.北京：原子能出版社,1982：231-234页
[15]张建民.核反应堆控制.西安：西安交通大学出版社,2002：38-41页,147-157页
[16]于涛,罗璋琳等.压水堆核电站堆芯集总参数模型的微机仿真.核电子学与探测技术.2001,21(6)：457-460页
[17]张法邦,吴清泉编著.核反应堆运行物理.北京：原子能出版社,2000：146-157页
[18]黄祖洽著.核反应堆动力学基础.北京：原子能出版社,1983：480-495页,181-186页
[19]于涛.压水堆核电站运行堆芯物理过程的PC仿真和演示平台研制.中国原子能科学研究院硕士研究生学位论文.2001：7-15页
[20]张玉峰.堆芯动态仿真中子动力学方程刚性解法的改进.核工程研究与设计.40(2000)：33-40页
[21]Jian Su, Renato M. Cotta. Improved lumped parameter formulation for simplified LWR thermohydraulic analysis. Annals of Nuclear Energy.28(2001):1019-1031P
[22]易维竞,李长顺,魏仁杰等.固定边界与移动边界直流蒸汽发生器模型的比较.核科学与工程,2002,20(4)：314-317页
[23]Kuridan, R. M, Beynon, T. D.. Analysis of the steam generator for the safe integral reactor concept:I. steady state. Progress in Nuclear Energy,1997,31(3):272-287P
[24]Kuridan, R. M, Beynon, T. D. Analysis of the steam generator for the safe integral reactor concept:Ⅱ. A linearized non-steady state model. Progress in Nuclera Energy,1998,31(3):289-301P
[25]杨世铭,陶文栓.传热学.高等教育出版社,1998
[26]Hassan Y. A. Thermal hydraulic predictions of a19-tube once-through steam generator tses using TRAC-PF1. The Winter Annual Meetiong of ASME, New York:asme,1985:92-100P
[27]Collier J. G对流沸腾和凝结.魏先英等译.北京：科学出版社,1982
[28]阎昌琪.汽液两相流.哈尔滨：哈尔滨工程大学出版社,1995
[29]M. A. Abdalla. A nonlinear dynamic model of a once-through, helical-coil steam generator. Oak Ridge Associated Universituy Master Paper,1993
[30]徐济鋆,贾斗南.沸腾传热和汽液两相流.北京：原子能出版社,1993
[31]刘铭.船用单缸汽轮机动态仿真及控制方法研究,硕士论文,哈尔滨工程大学,2005
[32]王增辉,贾斗南,刘瑞兰.狭缝通道两相流强化换热研究综述.热能动力工程,2002,17(100)：329-334
[33]于达仁,张宏光,徐基豫.中间再热抽汽式汽轮机的数学模型.汽轮机技术,1993,35(3)：9-12页
[34]倪维斗,王中泽.双抽汽式汽轮机的动态数学模型和仿真研究.动力工程,1991,11(3)：45-50页
[35]郑艳.300MW汽轮机数学模型的建立.南京工程学院,2002,2(4)：29-32页
[36]朱明善,刘颖,林兆庄等.工程热力学.北京：清华大学出版社,1993
[37]沈维道,郑佩芝,蒋炎安.工程热力学.北京：高等教育出版社,1983
[38]于瑞侠.核动力汽轮机.哈尔滨：哈尔滨工程大学出版社,2000
[39]Elbert Hendricks, Thomas Vesterholm. The analysis of mean value SI engine models. SAE 920682
[40]Minghui Kao, John J. Moskwa. Turbocharged diesel engine modeling for nonlinear engine control and state estimation. Transactions of the ASME, 1995(3):20-30P
[41]宋继武,彭义恒等.二回路汽轮机组.武汉：中国人民解放军工程学院,1982
[42]于明义,翁史烈,黄善衡.舰船核动力装置蒸汽轮机系统数模混合实时仿真.动力工程,1997,17(1)：45-49页
[43]B.A库兹涅佐夫编著.船舶核动力装置.中船重工集团第719研究所,2003
[44]丁宗华.一体化压水堆核动力装置协调控制技术研究,硕士论文,哈尔滨工程大学,2006
[45]熊淑燕,李国勇.一种改进的隐式广义预测自校正控制算法.太原工业大学学报,1992
[46]李国勇.输入受限的广义预测控制算法的稳定性.华北工学院学报,2004(4)
[47]舒迪前,预测控制系统及其应用.北京：机械工业出版社,1996：66-70页
[48]刘吉臻.协调控制与给水全程控制.北京：水利电力出版社,1995
[49]陈增强,袁著祉.工业锅炉的加权预测自校正控制.自动化学报,1993
[50]李国勇,谢克明.广义预测多变量隐式自校正控制算法研究.中国控制会议论文集.1994
[51]Dion J M, et al. MIMO Adaptive Generalized Predictive Control with Input-Output Constrains. Proceedings of ADCHEM'88 IFAC International Symposium Series Cepenhagen Denmark,1988
[2]Tsutao Hoshi, Masaaki Ochiai.日本一体化反应堆MAX的研究和开发现状.国外核动力,1996,(6)：11页
[3]陈炳德.日本小型核动力反应堆及其技术特点.核动力工程.2004,25(3)：192-197
[4]陈世君.国外舰船一体化压水堆技术发展趋势及其应用前景.国外核动力,2003,(1)：2-11页
[5]李满昌.世界一体化压水堆发展技术现状.国外核动力.1998(4)：1-12页
[6]李满昌,唐传宝.一体化压水堆发展前景.核动力工程.1998,19(4)：370-374页
[7]王成孝.核能与核技术应用.原子能出版社,2002
[8]H. Levent Akin, Vural.Akin. Rule-based fuzzy logic controller for a PWR-type nuclear power plant. IEEE Trance on Nuclear Science and Engineering,1993,114(1):42-54P
[9]Moon Ghu Park; Nam Zin Cho. Self-tuning control of a nuclear reactor using aGaussian function neural network. Nuclear Technology,1995, 110(2):285-293P
[10]Man Gyun Na. Amodel prddictive controller for the water level of nuclear steam generators. Journal of the Korean Nuclear Society,2001,33(1):102-110P
[11]Duk-Hyun Son, Chang-Goo Lee. Design of model predictive controller for water level control in the steam generator of a nuclear power plant. Transactions of the Korean Institute of Electrical Engineers,2001,50(8):376-383P
[12]孙增圻等编著.智能控制理论与技术.北京：清华大学出版社,1997：133-135页,195-196页
[13]赵兆颐,朱瑞安编著.反应堆热工流体力学.北京：清华大学出版社,1992：30-32页
[14]凌备备,杨延洲编著.核反应堆工程原理.第二版.北京：原子能出版社,1982：231-234页
[15]张建民.核反应堆控制.西安：西安交通大学出版社,2002：38-41页,147-157页
[16]于涛,罗璋琳等.压水堆核电站堆芯集总参数模型的微机仿真.核电子学与探测技术.2001,21(6)：457-460页
[17]张法邦,吴清泉编著.核反应堆运行物理.北京：原子能出版社,2000：146-157页
[18]黄祖洽著.核反应堆动力学基础.北京：原子能出版社,1983：480-495页,181-186页
[19]于涛.压水堆核电站运行堆芯物理过程的PC仿真和演示平台研制.中国原子能科学研究院硕士研究生学位论文.2001：7-15页
[20]张玉峰.堆芯动态仿真中子动力学方程刚性解法的改进.核工程研究与设计.40(2000)：33-40页
[21]Jian Su, Renato M. Cotta. Improved lumped parameter formulation for simplified LWR thermohydraulic analysis. Annals of Nuclear Energy.28(2001):1019-1031P
[22]易维竞,李长顺,魏仁杰等.固定边界与移动边界直流蒸汽发生器模型的比较.核科学与工程,2002,20(4)：314-317页
[23]Kuridan, R. M, Beynon, T. D.. Analysis of the steam generator for the safe integral reactor concept:I. steady state. Progress in Nuclear Energy,1997,31(3):272-287P
[24]Kuridan, R. M, Beynon, T. D. Analysis of the steam generator for the safe integral reactor concept:Ⅱ. A linearized non-steady state model. Progress in Nuclera Energy,1998,31(3):289-301P
[25]杨世铭,陶文栓.传热学.高等教育出版社,1998
[26]Hassan Y. A. Thermal hydraulic predictions of a19-tube once-through steam generator tses using TRAC-PF1. The Winter Annual Meetiong of ASME, New York:asme,1985:92-100P
[27]Collier J. G对流沸腾和凝结.魏先英等译.北京：科学出版社,1982
[28]阎昌琪.汽液两相流.哈尔滨：哈尔滨工程大学出版社,1995
[29]M. A. Abdalla. A nonlinear dynamic model of a once-through, helical-coil steam generator. Oak Ridge Associated Universituy Master Paper,1993
[30]徐济鋆,贾斗南.沸腾传热和汽液两相流.北京：原子能出版社,1993
[31]刘铭.船用单缸汽轮机动态仿真及控制方法研究,硕士论文,哈尔滨工程大学,2005
[32]王增辉,贾斗南,刘瑞兰.狭缝通道两相流强化换热研究综述.热能动力工程,2002,17(100)：329-334
[33]于达仁,张宏光,徐基豫.中间再热抽汽式汽轮机的数学模型.汽轮机技术,1993,35(3)：9-12页
[34]倪维斗,王中泽.双抽汽式汽轮机的动态数学模型和仿真研究.动力工程,1991,11(3)：45-50页
[35]郑艳.300MW汽轮机数学模型的建立.南京工程学院,2002,2(4)：29-32页
[36]朱明善,刘颖,林兆庄等.工程热力学.北京：清华大学出版社,1993
[37]沈维道,郑佩芝,蒋炎安.工程热力学.北京：高等教育出版社,1983
[38]于瑞侠.核动力汽轮机.哈尔滨：哈尔滨工程大学出版社,2000
[39]Elbert Hendricks, Thomas Vesterholm. The analysis of mean value SI engine models. SAE 920682
[40]Minghui Kao, John J. Moskwa. Turbocharged diesel engine modeling for nonlinear engine control and state estimation. Transactions of the ASME, 1995(3):20-30P
[41]宋继武,彭义恒等.二回路汽轮机组.武汉：中国人民解放军工程学院,1982
[42]于明义,翁史烈,黄善衡.舰船核动力装置蒸汽轮机系统数模混合实时仿真.动力工程,1997,17(1)：45-49页
[43]B.A库兹涅佐夫编著.船舶核动力装置.中船重工集团第719研究所,2003
[44]丁宗华.一体化压水堆核动力装置协调控制技术研究,硕士论文,哈尔滨工程大学,2006
[45]熊淑燕,李国勇.一种改进的隐式广义预测自校正控制算法.太原工业大学学报,1992
[46]李国勇.输入受限的广义预测控制算法的稳定性.华北工学院学报,2004(4)
[47]舒迪前,预测控制系统及其应用.北京：机械工业出版社,1996：66-70页
[48]刘吉臻.协调控制与给水全程控制.北京：水利电力出版社,1995
[49]陈增强,袁著祉.工业锅炉的加权预测自校正控制.自动化学报,1993
[50]李国勇,谢克明.广义预测多变量隐式自校正控制算法研究.中国控制会议论文集.1994
[51]Dion J M, et al. MIMO Adaptive Generalized Predictive Control with Input-Output Constrains. Proceedings of ADCHEM'88 IFAC International Symposium Series Cepenhagen Denmark,1988