基于神经网络的非线性预测控制算法的研究
摘要
模型预测控制(简称预测控制)是20世纪70年代在长期的工业实践过程中产生和发展起来的一类新型计算机控制算法,由于预测控制对模型要求低、鲁棒性强,并能有效地处理时变时滞、多变量和带约束等问题,线性预测控制方法目前已经被广泛应用在石油、化工、冶金、机械、机器人、生物医学等领域,并且取得了非常明显的经济效益。但工业过程和现场非常复杂,实际控制系统往往具有强耦合性、强非线性、时滞时变等特性。面对复杂的控制对象和更高的控制性能要求,基于线性模型的预测控制很难实现实时而有效的控制,因此研究针对非线性系统的预测控制方法已经成为工业控制界非常关注的课题。
将人工智能引入控制系统是现代自动控制学科发展的趋势,作为多学科交叉领域的前沿技术,人工神经网络具有自学习、自适应不确定性系统动态特性和逼近任意复杂非线性系统的特点,这使得神经网络可以应用在解决非线性和不确定性系统的建模问题中。因此将神经网络和预测控制的优势结合起来的基于神经网络的非线性预测控制逐渐成为了解决非线性系统控制问题的重要方法。
基于上述理论,本文首先介绍非线性预测控制的产生发展、非线性预测控制的主要研究方法和神经网络非线性预测控制的研究现状,然后详细阐述了预测控制的产生发展和基本原理,并着重介绍广义预测控制的基本算法和参数选择原则,同时指出预测控制是针对线性系统提出的,对于非线性系统由于难以建立精确模型进行多步预测,预测控制的控制效果往往不理想。因此本文引入具有良好逼近能力的局部动态反馈网络Elman神经网络对非线性系统进行模型辨识,并在此基础上提出将改进Elman神经网络应用在广义预测控制中。首先利用神经网络作为预测模型进行多步预测,并输出未来输出值,然后利用优化算法使目标函数达到最优求出最优控制量。为了避免通过推导改进Elman神经网络的雅克比矩阵(Jacobian)来计算最优控制律算法的复杂性,本文采用动态自适应粒子群算法(APSO)作为优化算法来设计预测控制器。最后,选取典型的强非线性对象进行仿真研究,结果表明基于改进Elman神经网络的非线性预测控制取得了较好的控制效果。
Model Predictive Control (MPC) is a promising new predictive control algorithm with the development of industrial practice in1970s. MPC algorithm has the characteristics of low requirements of model, strong robustness and good control performance. MPC also can deal the problems effectively with varying delays, multi-variable and constraint.Now MPC has been widely applied in the fields of petroleum, chemical, metallurgy, machinery, robotics, biomedicine in which significant economic benefits has been achieved.The actual industrial production systems have characteristics of strongly coupled, nonlinear, time delay and varying due to the complexity of the industrial process. Facing with complex industrial subjects and higher control performance requirements of plants, Generalized Predictive Control only adapted to the linear systems is difficult to achieve rapid and effective control in nonlinear systems. Therefore, the researches on Model Predictive Control of nonlinear systems give rise to more concern about the issue.
Applying Artificial Intelligence Technology on control system becomes the trend of development of the subject of modern automatic control. As a cutting-edge technology of multidisciplinary field, Artificial Neural Network (ANN) can be used to approach any complex nonlinear systems and be able to learn and adapt to the uncertainty systems'dynamic characteristics. Taking into account these characteristics, ANN is can be applied in solving the problem of modeling and controlling of nonlinear and uncertain systems. So, Nonlinear Model Predictive Control based on neural network gradually which combines the advantages of MPC and ANN gradually becomes an important method of solving complex nonlinear system control problems in industrial process.
Based on the above theories and ideas, this paper firstly introduces the development and main methods of Nonlinear Model Predictive Control (NMPC)as well as the research situation of NMPC. Then the emergence reason and development of Model Predictive Control is described, especially emphasis on Generalized Predictive Control algorithm and parameter selection principle. The paper also points out that MPC proposed for linear system and it is difficult to establish accurate models for nonlinear system for multi-step prediction and obtain satisfactory control effect Therefore, the neural network is applied in the nonlinear system identification in the paper. Neural network chooses Elman neural network which is a local dynamic feedback network with good approximation performance. On the basis of theories, the new method of applying modified Elman as a predictive model of multi-step prediction in the GPC is proposed. First of all the neural network is used as a predictive model to output multi-step prediction value, then optimum control is obtained by optimizing the objective function using optimization algorithms. Then Particle Swarm Optimization is adopted to avoid complexity of the recursive algorithm in the GPC controller. Finally, the results by the simulation of nonlinear system show that the new algorithm has a better control effect.
将人工智能引入控制系统是现代自动控制学科发展的趋势,作为多学科交叉领域的前沿技术,人工神经网络具有自学习、自适应不确定性系统动态特性和逼近任意复杂非线性系统的特点,这使得神经网络可以应用在解决非线性和不确定性系统的建模问题中。因此将神经网络和预测控制的优势结合起来的基于神经网络的非线性预测控制逐渐成为了解决非线性系统控制问题的重要方法。
基于上述理论,本文首先介绍非线性预测控制的产生发展、非线性预测控制的主要研究方法和神经网络非线性预测控制的研究现状,然后详细阐述了预测控制的产生发展和基本原理,并着重介绍广义预测控制的基本算法和参数选择原则,同时指出预测控制是针对线性系统提出的,对于非线性系统由于难以建立精确模型进行多步预测,预测控制的控制效果往往不理想。因此本文引入具有良好逼近能力的局部动态反馈网络Elman神经网络对非线性系统进行模型辨识,并在此基础上提出将改进Elman神经网络应用在广义预测控制中。首先利用神经网络作为预测模型进行多步预测,并输出未来输出值,然后利用优化算法使目标函数达到最优求出最优控制量。为了避免通过推导改进Elman神经网络的雅克比矩阵(Jacobian)来计算最优控制律算法的复杂性,本文采用动态自适应粒子群算法(APSO)作为优化算法来设计预测控制器。最后,选取典型的强非线性对象进行仿真研究,结果表明基于改进Elman神经网络的非线性预测控制取得了较好的控制效果。
Model Predictive Control (MPC) is a promising new predictive control algorithm with the development of industrial practice in1970s. MPC algorithm has the characteristics of low requirements of model, strong robustness and good control performance. MPC also can deal the problems effectively with varying delays, multi-variable and constraint.Now MPC has been widely applied in the fields of petroleum, chemical, metallurgy, machinery, robotics, biomedicine in which significant economic benefits has been achieved.The actual industrial production systems have characteristics of strongly coupled, nonlinear, time delay and varying due to the complexity of the industrial process. Facing with complex industrial subjects and higher control performance requirements of plants, Generalized Predictive Control only adapted to the linear systems is difficult to achieve rapid and effective control in nonlinear systems. Therefore, the researches on Model Predictive Control of nonlinear systems give rise to more concern about the issue.
Applying Artificial Intelligence Technology on control system becomes the trend of development of the subject of modern automatic control. As a cutting-edge technology of multidisciplinary field, Artificial Neural Network (ANN) can be used to approach any complex nonlinear systems and be able to learn and adapt to the uncertainty systems'dynamic characteristics. Taking into account these characteristics, ANN is can be applied in solving the problem of modeling and controlling of nonlinear and uncertain systems. So, Nonlinear Model Predictive Control based on neural network gradually which combines the advantages of MPC and ANN gradually becomes an important method of solving complex nonlinear system control problems in industrial process.
Based on the above theories and ideas, this paper firstly introduces the development and main methods of Nonlinear Model Predictive Control (NMPC)as well as the research situation of NMPC. Then the emergence reason and development of Model Predictive Control is described, especially emphasis on Generalized Predictive Control algorithm and parameter selection principle. The paper also points out that MPC proposed for linear system and it is difficult to establish accurate models for nonlinear system for multi-step prediction and obtain satisfactory control effect Therefore, the neural network is applied in the nonlinear system identification in the paper. Neural network chooses Elman neural network which is a local dynamic feedback network with good approximation performance. On the basis of theories, the new method of applying modified Elman as a predictive model of multi-step prediction in the GPC is proposed. First of all the neural network is used as a predictive model to output multi-step prediction value, then optimum control is obtained by optimizing the objective function using optimization algorithms. Then Particle Swarm Optimization is adopted to avoid complexity of the recursive algorithm in the GPC controller. Finally, the results by the simulation of nonlinear system show that the new algorithm has a better control effect.
引文
[1]王剑明.多区域变风量空调系统的神经网络广义预测控制研究[D].哈尔滨：哈尔滨工业大学,2008.
[2]郭红戈.基于神经网络多变量非线性系统广义预测控制[D].太原：太原理工大学,2005.
[3]王伟.基于RBF神经网络的滞后系统广义预测控制[D].大连：大连理工大学,2005.
[4]凌呼君,朱俊峰,李晓明.单元机组的小波神经网络广义预测控制[J].控制工程,2010,17(3)：316-319.
[5]张日东.非线性预测控制及应用[D].杭州：浙江大学,2007.
[6]安树.基于T-S模糊模型的广义预测控制方法研究[D].秦皇岛： 燕山大学,2006.
[7]张晓北.预测控制在热工中的应用[D].北京：华北电力大学,2008.
[8]张济民.基于神经网络的预测控制在摆式客车倾摆系统的应用研究[D].西安：西安交通大学,2004.
[9]赵敏.约束非线性系统预测控制算法设及稳定性分析[D].上海：上海交通大学,2009.
[10]Pedro Rodriguez, Didier Dumur. Generalized Predictive Control Robustification Under Frequency and Time-Domain Constraints [J]. IEEE Transactions on control systems technology,2005,13(4):577-579.
[11]范燕,申东日,陈义俊等.基于改进Elman神经网络的非线性预测控制[J].河南科技大学学报(自然科学),2007,28(1)：41-44.
[12]朱秋琴.基于神经网络的非线性GPC算法研究[D].兰州：兰州理工大学,2005.
[13]冯伟,张安年,丁酷.基于BP神经网络的非线性广义预测控制[J].河南科技大学学报(自然科学版),2009,30(2)：41-43.
[14]李楠.基于BP神经网络广义预测控制在气体压缩过程中的应用研究[D].哈尔滨：哈尔滨理工大学,2010.
[15]程辉,秦廷,陈宗海.液位对象的一种非线性广义预测控制策略研究[J].计算机仿真,2006,23(9)：333-335.
[16]陈薇.非线性预测控制快速算法的研究与应用[D].合肥：中国科技大学,2007.
[17]张建桃,张铁民,梁莉.超声电机非线性建模和广义预测控制[J].电机与控制学报, 2011,15(6)：50-54.
[18]Yasuki Kansha, Min-Sen Chiu. Adaptive generalized predictive control based on JITL technique[J]. Journal of Process Control,2009(19):1067-1072.
[19]庞中华,崔红.系统辨识与自适应控制MATLAB仿真[M].北京：北京航空航天大学出版社,2009：141-17.
[20]舒迪前.预测控制系统及其应用[M].北京：机械工业出版社,1996：176-189.
[21]董旸.复杂热工系统广义非线性预测控制及在线优化研究[D].北京：北京交通大学,2010.
[22]Rui Huang, Victor M. Zavala, Lorenz T. Biegler. Advanced step nonlinear model predictive control for air separation units [J]. Journal of Process Control,2009 (19):678-685.
[23]赵群.非线性系统的模糊广义预测控制算法研究[D].兰州：兰州理工大学,2007.
[24]王智琴.复杂热工系统的T-S模糊模型辨识研究[D].北京：北京交通大学,2009.
[25]迟庆雷.基于模糊神经网络的玉米干燥广义预测控制理论研究[D].沈阳：东北大学,2008.
[26]王喜庆.预测控制在电厂过热蒸汽温度控制器中的应用[D].杭州：浙江大学,2010.
[27]赵阳立.RBF网络直接广义预测控制及其收敛性研究[D].秦皇岛：燕山大学,2007.
[28]朱俊峰.小波神经网络广义预测控制在火电机协调控制系统中的应用[D].内蒙古工业大学,2007.
[29]阮定龙.基于人工神经网络的生物膜滴滤塔废气处理过程的建模和预测控制[D].重庆：重庆大学,2004.
[30]高萍.基于神经网络的水下机器人广义预测控制技术研究[D].哈尔滨：哈尔滨工程大学,2007.
[31]李国勇.智能预测控制及其MATLAB实现[M].北京：北京电子工业出版社,2010：262-265.
[32]A.R. Neshasteriz, A. Khaki Sedigh, H. Sadjadian. Generalized predictive control and tuning of industrial processes with second order plus dead time models. Journal of Process Control,2010(20):63-72.
[33]L.T. Biegler, V.M. Zavala. Large-scale nonlinear programming using IPOPT:An integrating framework for enterprise-wide dynamic optimization. Computers and Chemical Engineering[J]. Computers and Chemical Engineering,2009 (33):575-582.
[34]徐建安.水下机器人动力学模型辨识与广义预测控制研究[D].哈尔滨：哈尔滨工程 大学,2006.
[35]陈建华.基于神经网络的非线性预测控制的研究及其应用[D].北京：华北电力大学,2007.
[36]洪瑞,陈志旺,李建雄.非线性系统参数自适应直接广义预测控制[J].自动化学报,2007,10(2)：1110-1113.
[37]钱积新.预测控制[M].北京：化学工业出版社,2007：30,64.
[38]Wang Youqing, Zhou Donghua, Gao Furong.Generalized predictive control of linear systems with actuator arrearage faults [J]. Journal of Process Control,2009 (19):803-815.
[39]席裕庚,李德伟.预测控制定性综合理论的基本思路和研究现状[J].自动化学报,2008,34(10)：1225：1227.
[40]Zhonghua Denga, Hongliang Caoa, Xi Lia etc. Generalized predictive control for fractional order dynamic model of solid oxide fuel cell output power [J]. Journal of Power Sources,2010 (195):8097-8103.
[41]赵建军.模糊预测在火电厂的应用研究[D].保定：华北电力大学,2005.
[42]夏晓华,刘波,栾志业.基于APSO的非线性预测控制及在PH中和反应中的应用[J].化工自动化及仪表,2006,33(1)：24-27.
[43]夏晓华,刘波,栾志业等.基于PSO的预测控制及在聚丙烯中的应用[J].控制工程,2006,3(5)：401-403.
[44]Maciej Lawryn'czuk. Training of neural models for predictive control[J]. Neuro computing,2010(73):1332-1343.
[45]李国勇.神经网络模糊控制理论及其应用.北京：电子工业出版社,2009：11,44.
[46]黄西平,李睿,刘军.一种改进的神经网络非线性预测控制[J].计算机仿真,2006,23(4)：154-156.
[48]Vincent A. Akpan, George D. Hassapis, Nonlinear model identification and adaptive model predictive control using neural networks[J]. ISA Transactions.2011 (50):177-194.
[49]温淑焕,王科平.基于Elman网络的鲁棒型广义预测控制方法研究[J].系统工程学报,2004,19(5)：503-506.
[50]秦鹏,林中达.基于改进型Elman神经网络和遗传算法的锅炉在线燃烧优化[J].锅炉技术,2005,36(5)：40-41.
[51]李德伟.预测控制在线优化策略的研究[D].上海：上海交通大学,2009.
[52]Victor M. Zavala 1, Lorenz T. Biegler. Optimization-based strategies for the operation of low-density polyethylene tubular reactors:nonlinear model predictive control [J]. Computers and Chemical Engineering,2009 (33):1735-1746.
[53]范燕.基于改进Elman神经网络的非线性预测控制[J].河南科技大学学报,2007,28(1)：41-43.
[54]陈进东,王鲜芳,潘丰.基于PSO滚动优化的LS-SVM预测控制[J].计算机应用研究,2009,4(3)：1381-1382.
[55]柴毅,周海林,付东莉等.基于ESN和PSO的非线性模型预测控制[J].控制工程,2011,18(6)：865-867.
[56]张坚.神经网络控制技术在现代电站中的应用[D].重庆：重庆大学,2003.
[57]马量.基于Elman神经网络的预测控制研究及应用[D].哈尔滨：哈尔滨工程大学,2008.
[58]满红,邵诚.基于Hammerstein-Wiener模型的连续搅拌反应釜神经网络预测控制[J].化工学报,2011,62(8)：2275-2276.
[2]郭红戈.基于神经网络多变量非线性系统广义预测控制[D].太原：太原理工大学,2005.
[3]王伟.基于RBF神经网络的滞后系统广义预测控制[D].大连：大连理工大学,2005.
[4]凌呼君,朱俊峰,李晓明.单元机组的小波神经网络广义预测控制[J].控制工程,2010,17(3)：316-319.
[5]张日东.非线性预测控制及应用[D].杭州：浙江大学,2007.
[6]安树.基于T-S模糊模型的广义预测控制方法研究[D].秦皇岛： 燕山大学,2006.
[7]张晓北.预测控制在热工中的应用[D].北京：华北电力大学,2008.
[8]张济民.基于神经网络的预测控制在摆式客车倾摆系统的应用研究[D].西安：西安交通大学,2004.
[9]赵敏.约束非线性系统预测控制算法设及稳定性分析[D].上海：上海交通大学,2009.
[10]Pedro Rodriguez, Didier Dumur. Generalized Predictive Control Robustification Under Frequency and Time-Domain Constraints [J]. IEEE Transactions on control systems technology,2005,13(4):577-579.
[11]范燕,申东日,陈义俊等.基于改进Elman神经网络的非线性预测控制[J].河南科技大学学报(自然科学),2007,28(1)：41-44.
[12]朱秋琴.基于神经网络的非线性GPC算法研究[D].兰州：兰州理工大学,2005.
[13]冯伟,张安年,丁酷.基于BP神经网络的非线性广义预测控制[J].河南科技大学学报(自然科学版),2009,30(2)：41-43.
[14]李楠.基于BP神经网络广义预测控制在气体压缩过程中的应用研究[D].哈尔滨：哈尔滨理工大学,2010.
[15]程辉,秦廷,陈宗海.液位对象的一种非线性广义预测控制策略研究[J].计算机仿真,2006,23(9)：333-335.
[16]陈薇.非线性预测控制快速算法的研究与应用[D].合肥：中国科技大学,2007.
[17]张建桃,张铁民,梁莉.超声电机非线性建模和广义预测控制[J].电机与控制学报, 2011,15(6)：50-54.
[18]Yasuki Kansha, Min-Sen Chiu. Adaptive generalized predictive control based on JITL technique[J]. Journal of Process Control,2009(19):1067-1072.
[19]庞中华,崔红.系统辨识与自适应控制MATLAB仿真[M].北京：北京航空航天大学出版社,2009：141-17.
[20]舒迪前.预测控制系统及其应用[M].北京：机械工业出版社,1996：176-189.
[21]董旸.复杂热工系统广义非线性预测控制及在线优化研究[D].北京：北京交通大学,2010.
[22]Rui Huang, Victor M. Zavala, Lorenz T. Biegler. Advanced step nonlinear model predictive control for air separation units [J]. Journal of Process Control,2009 (19):678-685.
[23]赵群.非线性系统的模糊广义预测控制算法研究[D].兰州：兰州理工大学,2007.
[24]王智琴.复杂热工系统的T-S模糊模型辨识研究[D].北京：北京交通大学,2009.
[25]迟庆雷.基于模糊神经网络的玉米干燥广义预测控制理论研究[D].沈阳：东北大学,2008.
[26]王喜庆.预测控制在电厂过热蒸汽温度控制器中的应用[D].杭州：浙江大学,2010.
[27]赵阳立.RBF网络直接广义预测控制及其收敛性研究[D].秦皇岛：燕山大学,2007.
[28]朱俊峰.小波神经网络广义预测控制在火电机协调控制系统中的应用[D].内蒙古工业大学,2007.
[29]阮定龙.基于人工神经网络的生物膜滴滤塔废气处理过程的建模和预测控制[D].重庆：重庆大学,2004.
[30]高萍.基于神经网络的水下机器人广义预测控制技术研究[D].哈尔滨：哈尔滨工程大学,2007.
[31]李国勇.智能预测控制及其MATLAB实现[M].北京：北京电子工业出版社,2010：262-265.
[32]A.R. Neshasteriz, A. Khaki Sedigh, H. Sadjadian. Generalized predictive control and tuning of industrial processes with second order plus dead time models. Journal of Process Control,2010(20):63-72.
[33]L.T. Biegler, V.M. Zavala. Large-scale nonlinear programming using IPOPT:An integrating framework for enterprise-wide dynamic optimization. Computers and Chemical Engineering[J]. Computers and Chemical Engineering,2009 (33):575-582.
[34]徐建安.水下机器人动力学模型辨识与广义预测控制研究[D].哈尔滨：哈尔滨工程 大学,2006.
[35]陈建华.基于神经网络的非线性预测控制的研究及其应用[D].北京：华北电力大学,2007.
[36]洪瑞,陈志旺,李建雄.非线性系统参数自适应直接广义预测控制[J].自动化学报,2007,10(2)：1110-1113.
[37]钱积新.预测控制[M].北京：化学工业出版社,2007：30,64.
[38]Wang Youqing, Zhou Donghua, Gao Furong.Generalized predictive control of linear systems with actuator arrearage faults [J]. Journal of Process Control,2009 (19):803-815.
[39]席裕庚,李德伟.预测控制定性综合理论的基本思路和研究现状[J].自动化学报,2008,34(10)：1225：1227.
[40]Zhonghua Denga, Hongliang Caoa, Xi Lia etc. Generalized predictive control for fractional order dynamic model of solid oxide fuel cell output power [J]. Journal of Power Sources,2010 (195):8097-8103.
[41]赵建军.模糊预测在火电厂的应用研究[D].保定：华北电力大学,2005.
[42]夏晓华,刘波,栾志业.基于APSO的非线性预测控制及在PH中和反应中的应用[J].化工自动化及仪表,2006,33(1)：24-27.
[43]夏晓华,刘波,栾志业等.基于PSO的预测控制及在聚丙烯中的应用[J].控制工程,2006,3(5)：401-403.
[44]Maciej Lawryn'czuk. Training of neural models for predictive control[J]. Neuro computing,2010(73):1332-1343.
[45]李国勇.神经网络模糊控制理论及其应用.北京：电子工业出版社,2009：11,44.
[46]黄西平,李睿,刘军.一种改进的神经网络非线性预测控制[J].计算机仿真,2006,23(4)：154-156.
[48]Vincent A. Akpan, George D. Hassapis, Nonlinear model identification and adaptive model predictive control using neural networks[J]. ISA Transactions.2011 (50):177-194.
[49]温淑焕,王科平.基于Elman网络的鲁棒型广义预测控制方法研究[J].系统工程学报,2004,19(5)：503-506.
[50]秦鹏,林中达.基于改进型Elman神经网络和遗传算法的锅炉在线燃烧优化[J].锅炉技术,2005,36(5)：40-41.
[51]李德伟.预测控制在线优化策略的研究[D].上海：上海交通大学,2009.
[52]Victor M. Zavala 1, Lorenz T. Biegler. Optimization-based strategies for the operation of low-density polyethylene tubular reactors:nonlinear model predictive control [J]. Computers and Chemical Engineering,2009 (33):1735-1746.
[53]范燕.基于改进Elman神经网络的非线性预测控制[J].河南科技大学学报,2007,28(1)：41-43.
[54]陈进东,王鲜芳,潘丰.基于PSO滚动优化的LS-SVM预测控制[J].计算机应用研究,2009,4(3)：1381-1382.
[55]柴毅,周海林,付东莉等.基于ESN和PSO的非线性模型预测控制[J].控制工程,2011,18(6)：865-867.
[56]张坚.神经网络控制技术在现代电站中的应用[D].重庆：重庆大学,2003.
[57]马量.基于Elman神经网络的预测控制研究及应用[D].哈尔滨：哈尔滨工程大学,2008.
[58]满红,邵诚.基于Hammerstein-Wiener模型的连续搅拌反应釜神经网络预测控制[J].化工学报,2011,62(8)：2275-2276.