Robust low-complexity control laws design is the ultimate goal which Computer-Aided controlsystem Design (CACSD) aims for. We recognize seven classes of services CACSD tools and environments have to support w.r.t. th...
Robust low-complexity control laws design is the ultimate goal which Computer-Aided controlsystem Design (CACSD) aims for. We recognize seven classes of services CACSD tools and environments have to support w.r.t. this view: physical system modelling with modelling uncertainty descriptions; mathematical system modelling with suited uncertainty models; system model analysis in state space, time- and frequency domain; control law synthesis; multiobjective (multicriteria and multimodel) synthesis-parameter tuning; nonlinear assessment via simulation; control code generation. We deal with state of the art aspects of such services and their integration within an interactively controlled automatic design feedback loop. This is illustrated by experience gained during the two-years GARTEUR a design challenge project on Robust Aircraft Flight control, where 12 teams from 6 European countries demonstrated 11 different CACSD synthesis methods to design a landing-approach autopilot controller based on a Beluga-type aircraft model from Airbus Industries.
The performance benefits of an aerodynamically unstable aircraft are obtained at the expense of the necessity of an automatic controlsystem. For reliability requirements controllers without gain scheduling are looked...
详细信息
The performance benefits of an aerodynamically unstable aircraft are obtained at the expense of the necessity of an automatic controlsystem. For reliability requirements controllers without gain scheduling are looked for. To achieve this a complex ”robust control” design problem has to be solved. Fuzzy-controllers are claimed to have robustness properties. On the other hand the fuzzy-control methodology can be seen as a systematic way to specify controllers with nonlinear gain surfaces. To properly tune such controllers, computer-aided techniques can be applied provided appropriate models are available such as in more classical design approaches. A fuzzy-controller is developed, which keeps an aerodynamically unstable aircraft stationary for a sufficient long range of time and providing good flying qualities over the entire flight envelope. The results are compared with a completely linear controller design, done earlier.
Integrated systems engineering is essential in design-concept validation. In particular, computer-aided procedures for systemdynamics integration beyond mere functional integration are necessary in order to neither u...
详细信息
Integrated systems engineering is essential in design-concept validation. In particular, computer-aided procedures for systemdynamics integration beyond mere functional integration are necessary in order to neither underestimate nor to overestimate design-safety margins. At hand of the open system-dynamics integration environment ANDECS (ANalysis & DEsign of controlled systems), we deal with concepts and realized features of a framework for computational experimenting in system-dynamics design and evaluation. This paper is the lead paper to four companion papers which together form a case study on applied multidisciplinary dynamics design experimenting.
The robot trajectory optimization features of the ANDECS environment are described. Robot trajectory optimization problems are formulated as multi-objectives parameter optimization problems in search for a good compro...
详细信息
The robot trajectory optimization features of the ANDECS environment are described. Robot trajectory optimization problems are formulated as multi-objectives parameter optimization problems in search for a good compromise among conflicting criteria, such as “minimal” time and “minimal” power consumption. For that the robot dynamics is analyzed by time-simulation of a robot model with variable structure, which is generated by the object-oriented modeling language Dymola.
Nonlinear vehicle systemdynamics plays an important role in the development and the assessment of active suspension controlsystems. The current way of designing active suspension controlsystems mostly uses simplifi...
详细信息
Nonlinear vehicle systemdynamics plays an important role in the development and the assessment of active suspension controlsystems. The current way of designing active suspension controlsystems mostly uses simplified models, in order to to apply analytical synthesis methods or to achieve low complexity of the controller when using such methods. However, advanced vehicle design require methodologies, that can deal with more complex vehicle models, as a step towards integrated design. Multi-objective control design is such a methodology, that is open to cope with any controller structure and arbitrary kind of models. In this case study the usefulness of this methodology is investigated by designing an actively controlled vehicle suspension with and without preview for the IAVSD benchmark model ILTIS. The full nonlinear model is used for design. The ’nonlinear’ results are compared to the results of a ’linear’ design, based on linearized vehicle models.
For a low speed Maglev system of the single stage suspension type a control design study is described where a ’classical’ dynamic compensator control structure is compared with a ’modern’ LQG control structure. To...
详细信息
For a low speed Maglev system of the single stage suspension type a control design study is described where a ’classical’ dynamic compensator control structure is compared with a ’modern’ LQG control structure. To properly tune the free parameters in the respective controller synthesis procedures the ANDECS-MATLAB design environment is used. The multi-objective design approach of ANDECS allows to simultaneously satisfy the suspension requirements corresponding to both random and deterministic inputs.
In this paper, the numerical stability of the well known gain-normalized fast transversal filters (GFTF) algorithm is analytically investigated. Starting from an analysis of the error propagation, suitable round-off e...
详细信息
In this paper, the numerical stability of the well known gain-normalized fast transversal filters (GFTF) algorithm is analytically investigated. Starting from an analysis of the error propagation, suitable round-off errors difference equations (REDE) are derived. The stabilization of the GFTF algorithm can be achieved by stabilizing the REDE. The resulting algorithm is numerically stable, having only 7N + 27 multiplications and divisions per recursion (MADPR). The numerical stability of the proposed algorithm is demonstrated by the simulation results.
作者：
G. GrübelDLR Oberpfaffenhofen
Control-Design Engineering Group Institute for Robotics and System Dynamics D-82234 Wessling Germany
For computer-aided controlengineering education a scalable software technology is proposed, which is both useful for professional work in controlengineering theory and practice and for novice training. The idea is t...
详细信息
For computer-aided controlengineering education a scalable software technology is proposed, which is both useful for professional work in controlengineering theory and practice and for novice training. The idea is to provide a common environment for computational experimenting in all stages of controlengineering. This proposal is based on ANDECS (ANalysis & Design of controlled systems), a modular interactive control- and systemdynamics analysis and design system developed at the German Aerospace Research Establishment DLR.
Features of an available computation environment for robot-dynamics design are described. These are environments for object-oriented mechatronics/robot dynamics modeling, for executing dynamics simulation experiments,...
详细信息
Features of an available computation environment for robot-dynamics design are described. These are environments for object-oriented mechatronics/robot dynamics modeling, for executing dynamics simulation experiments, for multiobjective parameter and trajectory optimization, and for multivariate result visualization. The environments have a common look and feel and are coherently integrated using an engineering database system.< >
The ANDECS-Matlab environment provides two alternatives for databased computational experimenting using Matlab: The module Amatlab is a re-implementation of the classic matlab syntax to be used for interactive specifi...
详细信息
The ANDECS-Matlab environment provides two alternatives for databased computational experimenting using Matlab: The module Amatlab is a re-implementation of the classic matlab syntax to be used for interactive specification of matrix calculations in modular computation sequences. This syntax is extended to allow database access for data communication with other databased modules. Alternatively, the module PROMAT allows one to use commercial MATLAB via process communication with online data transfer between the ANDECS database and the MATLAB workspace in both directions. This allows a convenient use of the matlab syntax and a coherent integration of M-file Toolboxes in any module sequence of the control-engineering software environment ANDECS.< >
暂无评论