Pub Date : 1990-05-23DOI: 10.23919/ACC.1990.4790759
S. Rao, M. Westerheide, A. Das
The design and implementation of control strategies for large, flexible structures presents challenging problems. One of the difficulties in controller design arises from the incorrect knowledge of the structural parameters. This paper describes a procedure for designing and actual implementation of reduced order robust controllers for active vibration control on an experimental flexible grid structure. The experimental structure consists of a 5¿ × 5¿ grid made up of 2" wide, 1/8" thick aluminum strips. The grid hangs vertically down, being cantilevered at the top to a large I-beam anchored to a cinder block wall. The grid structure is represented by 75 degrees of freedom, finite element model and the controller uses three piezoelectric accelerometer sensors to command three non-collocated DC motor torquers. The modified balance-truncation model reduction method is used to derve a control synthesis model. These reduced order models preserve stability, controllability, observability and have a good frequency response match at low frequences. A 10-mode mathematical representation of the experimental grid structure has non-minimal phase zeros. The effects of non-minimum phase zeros on the performance of a closed loop LOG/LTR controller are investigated. The reduced order controllers are implemented on the structure using an ISI Max 100 computer. Experimental closed loop performance of the grid is obtained for various parameter variatons.
大型柔性结构控制策略的设计和实现提出了具有挑战性的问题。控制器设计的困难之一来自于对结构参数的不正确认识。本文介绍了一种用于柔性网架结构振动主动控制的降阶鲁棒控制器的设计和实际实现过程。实验结构由2英寸宽，1/8英寸厚的铝条组成的5 × 5网格组成。网格垂直向下悬挂，顶部悬挑在一根固定在煤渣砖墙上的大工字梁上。网格结构采用75自由度有限元模型表示，控制器采用3个压电加速度计传感器控制3个非配置的直流电机转矩器。采用改进的平衡截断模型约简法推导控制综合模型。这些降阶模型保持了稳定性、可控性、可观察性，并且在低频时有良好的频率响应匹配。实验网格结构的10模态数学表示具有非最小相位零。研究了非最小相位零对闭环LOG/LTR控制器性能的影响。用ISI Max 100计算机在结构上实现了降阶控制器。得到了不同参数变化下网格的实验闭环性能。
{"title":"Reduced Order Robust Controllers for an Experimental Flexible Grid","authors":"S. Rao, M. Westerheide, A. Das","doi":"10.23919/ACC.1990.4790759","DOIUrl":"https://doi.org/10.23919/ACC.1990.4790759","url":null,"abstract":"The design and implementation of control strategies for large, flexible structures presents challenging problems. One of the difficulties in controller design arises from the incorrect knowledge of the structural parameters. This paper describes a procedure for designing and actual implementation of reduced order robust controllers for active vibration control on an experimental flexible grid structure. The experimental structure consists of a 5¿ × 5¿ grid made up of 2\" wide, 1/8\" thick aluminum strips. The grid hangs vertically down, being cantilevered at the top to a large I-beam anchored to a cinder block wall. The grid structure is represented by 75 degrees of freedom, finite element model and the controller uses three piezoelectric accelerometer sensors to command three non-collocated DC motor torquers. The modified balance-truncation model reduction method is used to derve a control synthesis model. These reduced order models preserve stability, controllability, observability and have a good frequency response match at low frequences. A 10-mode mathematical representation of the experimental grid structure has non-minimal phase zeros. The effects of non-minimum phase zeros on the performance of a closed loop LOG/LTR controller are investigated. The reduced order controllers are implemented on the structure using an ISI Max 100 computer. Experimental closed loop performance of the grid is obtained for various parameter variatons.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126776251","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.1109/ACC.1990.4174213
R. Cristi, A. Healey, F. Papouias
Variable structure controllers for systems with nonlinearities can be designed in conjunction with robust observers to obtain regulation. The robust observer (of the Doyle Stein class) is designed on the basis of the nominal plant dynamics (assumed linear and minimum phase), and the estimated state is used in the variable structure controller. It is shown that the closed loop system is globally asymptoticvally stable and the desired regulation is obtained provided the nonlinear perturbation is expressed by an operator with bounded L2 and L¿ norms. Applications to the control of underwater vehicles in diving maneuvers show the effectiveness of the technique.
{"title":"Dynamic Output Feedback by Robust Observer and Variable Structure Control","authors":"R. Cristi, A. Healey, F. Papouias","doi":"10.1109/ACC.1990.4174213","DOIUrl":"https://doi.org/10.1109/ACC.1990.4174213","url":null,"abstract":"Variable structure controllers for systems with nonlinearities can be designed in conjunction with robust observers to obtain regulation. The robust observer (of the Doyle Stein class) is designed on the basis of the nominal plant dynamics (assumed linear and minimum phase), and the estimated state is used in the variable structure controller. It is shown that the closed loop system is globally asymptoticvally stable and the desired regulation is obtained provided the nonlinear perturbation is expressed by an operator with bounded L2 and L¿ norms. Applications to the control of underwater vehicles in diving maneuvers show the effectiveness of the technique.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114942364","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.23919/ACC.1990.4790766
S. Lin
In this paper, we propose a textured decomposition based algorithm with parallel processing capability for solving large scale multicommodity network flow problem (MFP), which is popular in communication networks and transportation systems. We show that the algorithm will always converge to one of the stationary points characterized by the textured decomposition. Moreover, a modification over the MFP and sufficient conditions are developed to guarantee that the convergent stationary point is the unique minimum solution of the modified MFP; and it is the global minimum solution of the MFP in its limit. To avoid ill-condition, a practical two-stage textured decomposition based algorithm is proposed to get a more elaborate approximate global minimum solution. A simulation example for this two-stage algorithm was given in which only few iterations are needed to converge. Assume two processors available, a speed up ratio around 10:1 is estimated based on a projected Newton method if it is incorporated with this two-stage algorithm.
{"title":"A Textured Decomposition based Algorithm for Large-Scale Multicommodity Network Flow Problems","authors":"S. Lin","doi":"10.23919/ACC.1990.4790766","DOIUrl":"https://doi.org/10.23919/ACC.1990.4790766","url":null,"abstract":"In this paper, we propose a textured decomposition based algorithm with parallel processing capability for solving large scale multicommodity network flow problem (MFP), which is popular in communication networks and transportation systems. We show that the algorithm will always converge to one of the stationary points characterized by the textured decomposition. Moreover, a modification over the MFP and sufficient conditions are developed to guarantee that the convergent stationary point is the unique minimum solution of the modified MFP; and it is the global minimum solution of the MFP in its limit. To avoid ill-condition, a practical two-stage textured decomposition based algorithm is proposed to get a more elaborate approximate global minimum solution. A simulation example for this two-stage algorithm was given in which only few iterations are needed to converge. Assume two processors available, a speed up ratio around 10:1 is estimated based on a projected Newton method if it is incorporated with this two-stage algorithm.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"315 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115222549","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.23919/ACC.1990.4791122
Qingling Zhang
In this paper, the concept of structurally impulsive controllablity is defined for descriptor systems of the form Ep¿ = Aqx + Bru, where Ep, Aq and Br are parameterized matrices, which is an extended form of structured descriptor systems [1]. Some necessary and sufficient conditions for the system to be structurally impulsive controllable are formulated algebraically. Consequently, we give an answer for structurally strong controllablity. These may be thought of as a counterpart of structurally complete (reachable) controllablity.
{"title":"Structurally Impulsive Controllablity for Descriptor Systems","authors":"Qingling Zhang","doi":"10.23919/ACC.1990.4791122","DOIUrl":"https://doi.org/10.23919/ACC.1990.4791122","url":null,"abstract":"In this paper, the concept of structurally impulsive controllablity is defined for descriptor systems of the form E<inf>p</inf>¿ = A<inf>q</inf>x + B<inf>r</inf>u, where E<inf>p</inf>, A<inf>q</inf> and B<inf>r</inf> are parameterized matrices, which is an extended form of structured descriptor systems [1]. Some necessary and sufficient conditions for the system to be structurally impulsive controllable are formulated algebraically. Consequently, we give an answer for structurally strong controllablity. These may be thought of as a counterpart of structurally complete (reachable) controllablity.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115348221","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.23919/ACC.1990.4791205
M. Jacobus, M. Jamshidi, C. Abdallah, P. Dorato, D. Bernstein
This paper considers the problem of stabilizing a plant using a suboptimal stable compensator of fixed order. The resulting equations are a modified form of the optimal projection equations, with the separation principle not holding in either the full- or reduced-order case.
{"title":"Suboptimal Strong Stabilization Using Fixed-Order Dynamic Compensation","authors":"M. Jacobus, M. Jamshidi, C. Abdallah, P. Dorato, D. Bernstein","doi":"10.23919/ACC.1990.4791205","DOIUrl":"https://doi.org/10.23919/ACC.1990.4791205","url":null,"abstract":"This paper considers the problem of stabilizing a plant using a suboptimal stable compensator of fixed order. The resulting equations are a modified form of the optimal projection equations, with the separation principle not holding in either the full- or reduced-order case.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"874 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116157650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.23919/ACC.1990.4791129
E. Abed, Hsien-Chiarn Lee
We consider the problem of designing stabilizing control laws for flight over a broad range of angles-of-attack which also serve to signal the pilot of impending stall. The paper employs bifurcation stabilization coupled with more traditional linear control system design. To focus the discussion, a detailed analysis is given for a model of the longitudinal dynamics of an F-8 Crusader.
{"title":"Nonlinear Stabilization of High Angle-of-Attack Flight Dynamics using Bifurcation Control","authors":"E. Abed, Hsien-Chiarn Lee","doi":"10.23919/ACC.1990.4791129","DOIUrl":"https://doi.org/10.23919/ACC.1990.4791129","url":null,"abstract":"We consider the problem of designing stabilizing control laws for flight over a broad range of angles-of-attack which also serve to signal the pilot of impending stall. The paper employs bifurcation stabilization coupled with more traditional linear control system design. To focus the discussion, a detailed analysis is given for a model of the longitudinal dynamics of an F-8 Crusader.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122393574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.23919/ACC.1990.4790971
Martin Gullaian, J. Lane
High performance pH control requires a good model of the titration curve, whih the described the nonlinear gain of the process. In many applications, the titration curve changes significantly over short periods of time, and some form of online titration curve estimation is essential. Unfortunately, the titration curve is a complex nonlinear function of process composition, and direct estimation of its parameters is difficult. This paper presents an estimation technique which uses as parameters a basis set of entire titration curves. The parameter choice linearizes the estimation problem, and has proven flexible and robust. An adaptive pH controller is described that uses the estimation technique to identify a process model online, and uses the model to provide a gain schedule for a PI controller. Results of pilot plant testing of this controller are presented.
{"title":"Titration Curve Estimation for Adaptive pH Control","authors":"Martin Gullaian, J. Lane","doi":"10.23919/ACC.1990.4790971","DOIUrl":"https://doi.org/10.23919/ACC.1990.4790971","url":null,"abstract":"High performance pH control requires a good model of the titration curve, whih the described the nonlinear gain of the process. In many applications, the titration curve changes significantly over short periods of time, and some form of online titration curve estimation is essential. Unfortunately, the titration curve is a complex nonlinear function of process composition, and direct estimation of its parameters is difficult. This paper presents an estimation technique which uses as parameters a basis set of entire titration curves. The parameter choice linearizes the estimation problem, and has proven flexible and robust. An adaptive pH controller is described that uses the estimation technique to identify a process model online, and uses the model to provide a gain schedule for a PI controller. Results of pilot plant testing of this controller are presented.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122760582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.1109/ACC.1990.4173840
W. Grossman, F. Khorrami, B. Friedland
In this paper, we describe a feedback linearization design for a class of nonlinear systems. The control system is an observer-based (full or reduced-order) design that utilizes estimates of some of the states both for the exact feedback linearization and the linear outer-loop controller. In particular, the approach advocated is applied to the dynamics of multi-link robot manipulators. Simulation studies and experimental results on a two-link planar arm are given and contrasted to show the applicability of the proposed methodology. Some practical issues for real-time implementation of the proposed feedback control algorithms are also presented.
{"title":"An Observer-Based Design for Robust Control of Robot Manipulators","authors":"W. Grossman, F. Khorrami, B. Friedland","doi":"10.1109/ACC.1990.4173840","DOIUrl":"https://doi.org/10.1109/ACC.1990.4173840","url":null,"abstract":"In this paper, we describe a feedback linearization design for a class of nonlinear systems. The control system is an observer-based (full or reduced-order) design that utilizes estimates of some of the states both for the exact feedback linearization and the linear outer-loop controller. In particular, the approach advocated is applied to the dynamics of multi-link robot manipulators. Simulation studies and experimental results on a two-link planar arm are given and contrasted to show the applicability of the proposed methodology. Some practical issues for real-time implementation of the proposed feedback control algorithms are also presented.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"283 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122957718","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.1109/ACC.1990.4174182
Sam Chuan Yao, E. Zafiriou
This paper investigates the use of local receptive field networks (LRFN) in detecting sensor failures of a control system in the presence of model-plant mismatch. Simulation results indicate that LRFNs hold significant promise in sensor failure detection. Another issue discussed in this paper is a method to prune redundant nodes. A simple scheme which uses singular value decomposition (SVD) is developed to identify and remove excess nodes. Comparable classification performance is obtained using reduced and standard LRFN.
{"title":"Control System Sensor Failure Detection via Networks of Localized Receptive Fields","authors":"Sam Chuan Yao, E. Zafiriou","doi":"10.1109/ACC.1990.4174182","DOIUrl":"https://doi.org/10.1109/ACC.1990.4174182","url":null,"abstract":"This paper investigates the use of local receptive field networks (LRFN) in detecting sensor failures of a control system in the presence of model-plant mismatch. Simulation results indicate that LRFNs hold significant promise in sensor failure detection. Another issue discussed in this paper is a method to prune redundant nodes. A simple scheme which uses singular value decomposition (SVD) is developed to identify and remove excess nodes. Comparable classification performance is obtained using reduced and standard LRFN.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114278966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1990-05-23DOI: 10.1109/ACC.1990.4174055
Martin S. K. Leung, A. Calise
This paper presents an approximate solution for the maximum payload trajectory of a 2-stage launch vehicle using a regular perturbation technique. A zero-order solution for a two stage vehicle based on a flat Earth approximation and negligible atmospheric effect is obtained in closed form. High order correction terms are obtained from the solution of non-homogeneous, first-order linear differential equations by quadrature. This promises a capability for an on-board optimal guidance law implementation.
{"title":"An Approach to Optimal Guidance of an Advanced Lauch Vehicle Concept","authors":"Martin S. K. Leung, A. Calise","doi":"10.1109/ACC.1990.4174055","DOIUrl":"https://doi.org/10.1109/ACC.1990.4174055","url":null,"abstract":"This paper presents an approximate solution for the maximum payload trajectory of a 2-stage launch vehicle using a regular perturbation technique. A zero-order solution for a two stage vehicle based on a flat Earth approximation and negligible atmospheric effect is obtained in closed form. High order correction terms are obtained from the solution of non-homogeneous, first-order linear differential equations by quadrature. This promises a capability for an on-board optimal guidance law implementation.","PeriodicalId":307181,"journal":{"name":"1990 American Control Conference","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122035117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：