Problems of robust residual generation and evaluation are investigated with the aid of frequency-domain approaches and H/sub infinity /-optimization techniques. Based on the parameterization of achievable residual dynamics the robust residual generation is defined as an optimization problem that can be solved by H/sub infinity /-techniques. To increase the robustness of residual evaluation, a frequency-domain residual measure is introduced, under which fault thresholds are derived.<>
{"title":"An approach to robust residual generation and evaluation","authors":"X. Ding, P. Frank","doi":"10.1109/CDC.1991.261391","DOIUrl":"https://doi.org/10.1109/CDC.1991.261391","url":null,"abstract":"Problems of robust residual generation and evaluation are investigated with the aid of frequency-domain approaches and H/sub infinity /-optimization techniques. Based on the parameterization of achievable residual dynamics the robust residual generation is defined as an optimization problem that can be solved by H/sub infinity /-techniques. To increase the robustness of residual evaluation, a frequency-domain residual measure is introduced, under which fault thresholds are derived.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131352086","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}
A class of nth-order linear systems with a variable structure is discretized and analyzed in detail. The occurrence and structure of pseudo-sliding modes give insight into corresponding sliding modes for the continuous system, and enable the relation between system parameters and the step size to be explored. A discretization can be used which gives a mathematical model revealing details of the chattering along a surface. A sufficient condition on the step size and the system and switching parameters to ensure the existence of a pseudo-sliding mode is derived. The analysis is illustrated with a simulation study of second-order systems.<>
{"title":"A class of discrete variable structure systems","authors":"X. Yu, R. B. Potts","doi":"10.1109/CDC.1991.261623","DOIUrl":"https://doi.org/10.1109/CDC.1991.261623","url":null,"abstract":"A class of nth-order linear systems with a variable structure is discretized and analyzed in detail. The occurrence and structure of pseudo-sliding modes give insight into corresponding sliding modes for the continuous system, and enable the relation between system parameters and the step size to be explored. A discretization can be used which gives a mathematical model revealing details of the chattering along a surface. A sufficient condition on the step size and the system and switching parameters to ensure the existence of a pseudo-sliding mode is derived. The analysis is illustrated with a simulation study of second-order systems.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132065398","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}
Linear dynamic errors-in-variables (or factor) models in the framework of stationary processes are considered. The noise process is assumed to have a diagonal spectral density. The relation between the (population) second moments of the observations and the system and noise characteristics is analyzed; of particular interest are the number of equations (or the number of factors) and a description of the set of all systems compatible with the second moments of the observations. Emphasis is placed on the case which can be reduced to a single factor. The problems considered arise in the context of identification and precede estimation.<>
{"title":"Identification of dynamic systems from noisy data: the case m*=n-1","authors":"B. Anderson, M. Deistler","doi":"10.1109/CDC.1991.261692","DOIUrl":"https://doi.org/10.1109/CDC.1991.261692","url":null,"abstract":"Linear dynamic errors-in-variables (or factor) models in the framework of stationary processes are considered. The noise process is assumed to have a diagonal spectral density. The relation between the (population) second moments of the observations and the system and noise characteristics is analyzed; of particular interest are the number of equations (or the number of factors) and a description of the set of all systems compatible with the second moments of the observations. Emphasis is placed on the case which can be reduced to a single factor. The problems considered arise in the context of identification and precede estimation.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132382270","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}
The problem of frequency-weighted optimal model order reduction for discrete-time systems is considered. Necessary conditions completely characterizing the reduced-order model are given. The solution consists of a set of one generalized Riccati equation and two generalized Lyapunov equations all coupled by a projection. In the case when the frequency-weighting transfer function is strictly proper an additional projection appears in the solution.<>
{"title":"Discrete-time frequency weighted model order reduction","authors":"Y. Halevi","doi":"10.1109/CDC.1991.261761","DOIUrl":"https://doi.org/10.1109/CDC.1991.261761","url":null,"abstract":"The problem of frequency-weighted optimal model order reduction for discrete-time systems is considered. Necessary conditions completely characterizing the reduced-order model are given. The solution consists of a set of one generalized Riccati equation and two generalized Lyapunov equations all coupled by a projection. In the case when the frequency-weighting transfer function is strictly proper an additional projection appears in the solution.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132420192","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}
Continuous time linear stochastic systems with unknown bilinear parameters are considered. A specific approximation to the optimal nonlinear filter used as a recursive parameter estimator is derived by retaining third-order moments and using a Gaussian approximation for higher-order moments. With probability one, the specific approximation is proven to converge to a minimum of the likelihood function. The proof uses the ordinary differential equation technique and requires that the slow system is bounded on finite time intervals and the fixed-parameter fast system is asymptotically stable. The fixed parameter fast system is proven asymptotically stable if the parameter update gain is small enough. Essentially, the specific approximation is asymptotically equivalent to the recursive prediction error method, thus inheriting its asymptotic rate of convergence. A numerical simulation for a simple example indicates that the specific approximation has better transient response than other commonly used parameter estimators.<>
{"title":"A convergent approximation of the optimal parameter estimator","authors":"D. Wiberg, D.C. DeWolf","doi":"10.1109/CDC.1991.261771","DOIUrl":"https://doi.org/10.1109/CDC.1991.261771","url":null,"abstract":"Continuous time linear stochastic systems with unknown bilinear parameters are considered. A specific approximation to the optimal nonlinear filter used as a recursive parameter estimator is derived by retaining third-order moments and using a Gaussian approximation for higher-order moments. With probability one, the specific approximation is proven to converge to a minimum of the likelihood function. The proof uses the ordinary differential equation technique and requires that the slow system is bounded on finite time intervals and the fixed-parameter fast system is asymptotically stable. The fixed parameter fast system is proven asymptotically stable if the parameter update gain is small enough. Essentially, the specific approximation is asymptotically equivalent to the recursive prediction error method, thus inheriting its asymptotic rate of convergence. A numerical simulation for a simple example indicates that the specific approximation has better transient response than other commonly used parameter estimators.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"2147 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130015793","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}
A unified generalized model-based predictive control (GMBPC) technique is presented. This technique combines in an efficient way the key properties of several previous MBPC-like algorithms. The multiple-input-multiple-output (MIMO) state-space is employed, and state and control constraints are included in the system formulation. For better accuracy a second-order model is employed for each output variable, while a first-order model is always used in the available algorithms. For the unconstrained case a simple explicit control law is obtained. A particular feature of the proposed technique is that the predictive functional control principle is used to reduce the computational complexity of the resulting GMBP controller. Extensive simulation examples in industrial and managerial system models support the effectiveness of the present GMBP controller in terms of meeting the desired specifications and having reduced computational requirements.<>
{"title":"A new generalized model-based predictive control algorithm","authors":"S. Tzafestas, G. Vagelatos, G. Capsiotis","doi":"10.1109/CDC.1991.261473","DOIUrl":"https://doi.org/10.1109/CDC.1991.261473","url":null,"abstract":"A unified generalized model-based predictive control (GMBPC) technique is presented. This technique combines in an efficient way the key properties of several previous MBPC-like algorithms. The multiple-input-multiple-output (MIMO) state-space is employed, and state and control constraints are included in the system formulation. For better accuracy a second-order model is employed for each output variable, while a first-order model is always used in the available algorithms. For the unconstrained case a simple explicit control law is obtained. A particular feature of the proposed technique is that the predictive functional control principle is used to reduce the computational complexity of the resulting GMBP controller. Extensive simulation examples in industrial and managerial system models support the effectiveness of the present GMBP controller in terms of meeting the desired specifications and having reduced computational requirements.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130024250","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}
A Petri net model is described that is consistent with the control theory for discrete event systems. The purpose is to apply the theoretical approach of supervisory control to a model that permits an efficient solution of the control problem. The focus is at the model level. The authors show how Petri nets may be used to design a supervisor. The design requires two steps. In the first step, a coarse structure for a supervisor is synthesized by means of a concurrent composition of different modules. In the second step, the structure is refined to avoid reaching forbidden markings. The refinement procedure may always be applied when the net is conservative. In both steps, the use of Petri nets allows the structure of the model to be small.<>
{"title":"Supervisory design using Petri nets","authors":"Alessandro Giua, F. DiCesare","doi":"10.1109/CDC.1991.261262","DOIUrl":"https://doi.org/10.1109/CDC.1991.261262","url":null,"abstract":"A Petri net model is described that is consistent with the control theory for discrete event systems. The purpose is to apply the theoretical approach of supervisory control to a model that permits an efficient solution of the control problem. The focus is at the model level. The authors show how Petri nets may be used to design a supervisor. The design requires two steps. In the first step, a coarse structure for a supervisor is synthesized by means of a concurrent composition of different modules. In the second step, the structure is refined to avoid reaching forbidden markings. The refinement procedure may always be applied when the net is conservative. In both steps, the use of Petri nets allows the structure of the model to be small.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130107562","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}
A method of compensating for friction in control systems is presented. The method entails the use of an observer to estimate the friction which is modeled as a constant times the sign of velocity. The purpose of the observer is to estimate this constant. The observer model is selected to ensure that the error in estimation of the friction constant converges asymptotically to zero. Simulation results verify the theory and show that the method can significantly improve the performance of a control system in which it is used. Although based on the assumption of a constant friction magnitude, the observer displays the ability to track friction whose magnitude depends on velocity.<>
{"title":"On adaptive friction compensation","authors":"Bernard Friedland, Y. Park","doi":"10.1109/CDC.1991.261068","DOIUrl":"https://doi.org/10.1109/CDC.1991.261068","url":null,"abstract":"A method of compensating for friction in control systems is presented. The method entails the use of an observer to estimate the friction which is modeled as a constant times the sign of velocity. The purpose of the observer is to estimate this constant. The observer model is selected to ensure that the error in estimation of the friction constant converges asymptotically to zero. Simulation results verify the theory and show that the method can significantly improve the performance of a control system in which it is used. Although based on the assumption of a constant friction magnitude, the observer displays the ability to track friction whose magnitude depends on velocity.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"05 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130437320","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}
The authors develop a methodology for reasoning about the state of the environment based on evidence received from some source. It is assumed that the evidence is expressed as a probability mass function defined on a discrete set of mutually exclusive hypotheses about the state of the environment. Given that the quality of the evidence is variable, it follows that the precision of the reasoning process must also vary. That is, the level of specificity and the certainty associated with decisions made at that level depend directly on the quality of the evidence. An indistinguishability measure is used to generate a core set of aggregate focal elements, each of which may consist of logical disjunctions of the basic hypothesis set. The measure takes into account both the differences in support levels for the hypotheses and the degree to which they are similar. Partial dominance is then used to associate a basic probability assignment on the core set. This approach makes it possible to apply simple, quantitative methods to express the variations in the precision associated with decisions. The result is a set of aggregate hypotheses and their support levels which become input to the classification process. In most cases, multiple sets of aggregate hypotheses will be used in an evidential classification scheme to produce a composite characterization of the environment.<>
{"title":"A quantitative treatment of multilevel specificity and certainty in variable precision reasoning","authors":"W. L. Perry, H. Stephanou","doi":"10.1109/CDC.1991.261744","DOIUrl":"https://doi.org/10.1109/CDC.1991.261744","url":null,"abstract":"The authors develop a methodology for reasoning about the state of the environment based on evidence received from some source. It is assumed that the evidence is expressed as a probability mass function defined on a discrete set of mutually exclusive hypotheses about the state of the environment. Given that the quality of the evidence is variable, it follows that the precision of the reasoning process must also vary. That is, the level of specificity and the certainty associated with decisions made at that level depend directly on the quality of the evidence. An indistinguishability measure is used to generate a core set of aggregate focal elements, each of which may consist of logical disjunctions of the basic hypothesis set. The measure takes into account both the differences in support levels for the hypotheses and the degree to which they are similar. Partial dominance is then used to associate a basic probability assignment on the core set. This approach makes it possible to apply simple, quantitative methods to express the variations in the precision associated with decisions. The result is a set of aggregate hypotheses and their support levels which become input to the classification process. In most cases, multiple sets of aggregate hypotheses will be used in an evidential classification scheme to produce a composite characterization of the environment.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134263026","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}
The authors present a case study of a two-input-two-output ill-conditioned plant using the quantitative feedback theory (QFT) approach. No iterative steps were needed to find the controller, and a good compatibility with time domain performance was achieved. The pre-filter, on the other hand, required five iterations because the performance was given in the time domain. The Horowitz design philosophy which emphasize tolerances at each frequency and design to the exact uncertain plant, i.e. no norm-bounded uncertainty, justified itself as a design tool for ill-conditioned plants to achieve an economical bandwidth solution.<>
{"title":"Ill conditioned plants: a case study","authors":"O. Yaniv, H. Isaac","doi":"10.1109/CDC.1991.261674","DOIUrl":"https://doi.org/10.1109/CDC.1991.261674","url":null,"abstract":"The authors present a case study of a two-input-two-output ill-conditioned plant using the quantitative feedback theory (QFT) approach. No iterative steps were needed to find the controller, and a good compatibility with time domain performance was achieved. The pre-filter, on the other hand, required five iterations because the performance was given in the time domain. The Horowitz design philosophy which emphasize tolerances at each frequency and design to the exact uncertain plant, i.e. no norm-bounded uncertainty, justified itself as a design tool for ill-conditioned plants to achieve an economical bandwidth solution.<<ETX>>","PeriodicalId":344553,"journal":{"name":"[1991] Proceedings of the 30th IEEE Conference on Decision and Control","volume":"119 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1991-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134178636","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：