Mixtures of Controllers for Jump Linear and Non-Linear Plants
نویسندگان
چکیده
Steven J. Nowlan Synaptics, Inc. 2698 Orchard Parkway San Jose, CA 95134 We describe an extension to the Mixture of Experts architecture for modelling and controlling dynamical systems which exhibit multiple modes of behavior. This extension is based on a Markov process model, and suggests a recurrent network for gating a set of linear or non-linear controllers. The new architecture is demonstrated to be capable of learning effective control strategies for jump linear and non-linear plants with multiple modes of behavior.
منابع مشابه
Stability and Performance Attainment with Fixed Order Controller Using Frequency Response
Recently, a new data driven controller synthesis is presented for calculating the family of stabilizing first, second and fixed order controller using frequency respons. However, this method is applicable just for plants that can guarantee some smoothness at the boundary of the resulted high dimension LMI. This paper solve that issue and extends the approach to fixed order controllers guarantee...
متن کاملAdaptive Predictive Controllers Using a Growing and Pruning RBF Neural Network
An adaptive version of growing and pruning RBF neural network has been used to predict the system output and implement Linear Model-Based Predictive Controller (LMPC) and Non-linear Model-based Predictive Controller (NMPC) strategies. A radial-basis neural network with growing and pruning capabilities is introduced to carry out on-line model identification.An Unscented Kal...
متن کاملSoft Switching MBC Controller for MIMO Linear Hybrid Systems
Switching supervisory is the most important section of a feedback control process inMIMO hybrid systems. By choosing a non-compatible controller, system may go tounstable mode or high overshoot response. In this paper, a new method of switchingfor selecting MBC controllers is discussed. Results of the simulation show the MIMO(2-in 2-out) linear hybrid system can be switched stable and low overs...
متن کاملLinear Dynamically Varying Versus Jump Linear Systems
The connection between linear dynamically varying (LDV) systems and jump linear systems is explored. LDV systems have been shown to be useful in controlling systems with "complicated dynamics". Some systems with complicated dynamics, for example Axiom A systems, admit Markov partitions and can be described, up to finite resolution, by a Markov chain. In this case, the control system for these s...
متن کاملA short overview of the electrical machines control based on Flatness-technique
Optimal linear controllers and high computational non-linear controllers are normally applied to control the nonlinear systems. Flatness control method is a control technique for linear systems as well as nonlinear systems by static and dynamic feedback namely as endogenous dynamic feedback. This method takes into account the non-linear behavior of the process while preventing complicated compu...
متن کامل