Font Size: a A A

Disturbance Attenuation Control Scheme For Constrained Systems And Its Application

Posted on:2010-12-03Degree:DoctorType:Dissertation
Country:ChinaCandidate:M M MaFull Text:PDF
GTID:1118360272496805Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Time-domain hard constraints, including actuator saturations, state and output con-straints are probably the most widely encountered and most dangerous nonlinearity inpractice control systems. It is well recognized that constraint violation might result inenvironment pollution and safety accidents such as damages on equipments and persons.If these constraints can not be ignored we must consider its in?uence on the control per-formance, otherwise, the expectation control objectives can not be obtained. In addition,external disturbances also widely exist in practical control systems and we expect that thedesigned closed-loop system can guarantee a disturbance attenuation performance for ex-ternal disturbances. However, good control performance for control systems, such as fasttransient response, high level of disturbance attenuation, requires large control actions (orhigh-gain controllers), which easily leads to constraint violations. All of these result inthe con?ict between high performance and satisfaction of time-domain hard constraints."Higher"gain will be derived if we ignore these constraints and only performance is con-sidered while designing state-feedback controller. Control action will exceed the boundgreatly even though the system errors are very small, especially the system is encounteredby abrupt reference vary or imposed external disturbances. In such case, the control ac-tion needs to be clipped to the bounded level, but performance of closed-loop system cannot be guaranteed, even loss of stability. In a word, time-domain hard constraints mustbe handled in some control system design and controller design for constrained system isactually to manage the trade-o? between high performance and satisfaction of constraints.One objective of a control system is to improve the performance by using limited controlaction. Recently, there is a growing interest in the disturbance attenuation problem ofconstrained systems, which has both theoretical and engineering application significance.The objective of this paper is to present systematic disturbance attenuation controlscheme for constrained system. Firstly, we propose the state-feedback disturbance at-tenuation strategy to guarantee the disturbance attenuation performance of closed-loopsystem with constraints. we bring forward two approaches to deal with hard constraints: ellipsoidal invariable sets control and generalized H2 control. The main idea of the el-lipsoidal invariant sets is to seek an ellipsoidal invariant set that contains all possiblestate trajectories, and then su?cient conditions dependent on this set that guaranteessatisfaction of constraints are derived. In addition, generalized H2 norm is defined asa measurement from energy-bounded inputs to peak amplitudes of the outputs in timedomain, we consider to capture the hard constraints by generalized H2 performance. Inthe framework of LMI optimization and multi-objective control, several state-feedbackdisturbance attenuation control schemes, which combining either of the constraints con-trol methods and H∞performance or H2 performance are formulated, respectively, whichcan guarantee the performance of closed-loop system and time-domain hard constraintsare satisfied meanwhile. Then, we apply the proposed H2 /generalized H2 state-feedbackcontrol strategy to design active suspension systems, which achieve the best possible ridecomfort and keep hard constraints within bounds.The state-feedback control strategy requires system states as feedback, which are ingeneral not fully measurable. Thus, we present the output feedback disturbance attenua-tion control scheme for constrained system. Considering that the ellipsoidal invariant setmethod is not suitable to the output-feedback case. Several output-feedback disturbanceattenuation control schemes, which combining general H2 constraint control method andH∞performance or H2 performance are formulated, respectively. In addition, the pro-posed H2 /generalized H2 output-feedback control strategy is applied to design activesuspension systems. Analysis results show an obvious improvement to the performanceof the active suspension once again.Combining the moving horizon principle of model predictive control, a new movinghorizon H∞control scheme with performance adaptation for linear systems with con-straints has been discussed. By minimizing the performance index on-line, the less con-servative algorithm is presented, which is able to manage the trade-o? between requiredhigh performance and satisfying hard constraints (i.e., relax the performance when thecontrol inputs reach their bounds and improve the performance when the control inputsare far away from there bound), and to enable the most of the limited control to achievegood performance. In order to validate the e?ciencies of the proposed control scheme,we discussed their applications in active suspension control. In addition, we present themoving horizon H∞tracking scheme for nonlinear system and discuss the closed-loopproperties. In order to reduce the conservatism, an improved algorithm is suggested andapplied to the wheeled mobile robots tracking control problem in the simultaneous pres-ence of control constraints and external disturbances. Analysis an simulation results showthe moving horizon H∞tracking controller can manage automatically between achieving good tracking performance and satisfying control constraints.Finally, we present a nonlinear disturbance attenuation control approach in the com-bination of the LMI based constrained H∞control approach and backstepping techniques.We derive the controller into two steps: First we consider nonlinear term as fictitious con-trol for the linear subsystem with time-domain constraints and design a state feedbackH∞controller respecting time-domain constraints. In the second step, we apply the back-stepping technique to determine the control input such that the nonlinear term tracks thedesired reference well, while guaranteeing a bounded disturbance attenuation level for theclosed-loop system. Then, we apply the backstepping technique to deal with nonlinearactuator dynamics and to derive the control law. Analysis and simulation results have con-firmed the potential benefit of the proposed constrained nonlinear H∞active suspensionin achieving the best possible ride comfort and compensating for the hydraulic nonlinear-ity, in the precondition of keeping hard constraints within bounds. Hardware-in-the-loopsimulation results are provided to show the e?ectiveness of the designed controller.The formulation processes and the proof of mentioned approaches are presented indetail in this thesis. Moreover, in order to validate the e?ciencies of the proposed ap-proaches, we give simulation results for each approach, which is discussed in detail frommodelling, controller parameter choosing, disturbance attenuation testing. The simulationresults indicate that the control e?ects with the proposed approaches are satisfactory.Due to the limit of experiment equipment, we just give the theoretical analysis,numerical and hardware-in-loop simulation results. If the experiments are performed, webelieve that more valuable conclusion can be obtained. Deeper research work needs tobe done since some problems are still remain to be solved, such as how to reduce the therank of the output feedback controller, or how to speed up the online calculation etc.
Keywords/Search Tags:Time-domain hard constraints, ellipsoidal invariant set, Linear Matrix Inequality, H2 control, H_∞control, output-feedback, moving horizon optimization, disturbance attenuation
PDF Full Text Request
Related items