none

No description available.

No description available.

O controle e otimização da operação da suspensão eletromagnética de um protótipo de veículo MAGLEV são apresentados neste trabalho. Este tipo de sistema é inerentemente instável e altamente não linear, sendo um excelente exemplo para o estudo e comparação de diferentes metodologias de controle. Nos últimos anos, experiências significativas com veículos levitados têm ganhado força. A arquitetura do veículo consiste de quatro atuadores eletromagnéticos, quatro sensores de entreferros e quatro acelerômetros, todos instalados nos cantos do veículo. A dinâmica do sistema é descrita através de três modelos matemáticos diferentes: modelo de corpo rígido com três graus de liberdade (movimentos vertical, de rolagem e de arfagem), modelo de corpo não rígido, onde está presente mais um grau de liberdade (movimento de torção), e modelo SISO com quatro movimentos verticais independentes. Blocos de compatibilização de variáveis são utilizados devido à diferença no número de variáveis do modelo matemático com o número de variáveis medidas e de atuação disponíveis no veículo. Os acelerômetros, usados para estimar as velocidades do veículo, foram instalados com o objetivo de estudar-se o controle ótimo da suspensão por realimentação de estados com o emprego da técnica do Regulador Linear Quadrático (RLQ) e utilização do método do lugar das raízes para definição dos pólos de malha fechada. Para efeito de comparação foi implementado um controle H'IND.2'/H'INFINITO' por realimentação da saída. Controladores independentes para cada grau de liberdade foram simulados no software MATLAB e depois implementados e testados no veículo utilizando uma placa de aquisição de dados instalada no microcomputador. A preocupação principal do projeto dos controladores foi a estabilidade do sistema na ocorrência de perturbações que desviem o entreferro de seu valor nominal.Exceto para o sistema com controle RLQ e modelo de corpo rígido, os resultados experimentais obtidos com o protótipo mostram que o sistema em malha fechada é estável e apresenta uma resposta transiente satisfatória. O sistema com controle RLQ e modelo de corpo não rígido apresentou a melhor performance entre todas as alternativas testadas. / The control and optimization of the electromagnetic suspension operation of a MAGLEV vehicle prototype are presented in this work. This kind of system is inherently unstable and highly nonlinear, being an excellent example for the study and comparison of different control methodologies. In the last years significant experience has been gained with levitated vehicles. The vehicle architecture consists of four electromagnetic actuators, four air gap sensors and four accelerometers, all located at the corners of the vehicle. The vehicle dynamics is described through three different mathematical models: rigid body model, with three degrees of freedom (heave, roll and pitch motions), not-rigid body model, with an additional degree of freedom (torsion motion) and SISO model, with four independent heave motions. Blocks to make variables compatible are used in based on the difference between the number of variables of the mathematical model and the number of measured and actuation variables available in the vehicle. The accelerometers, used to estimate the speeds of the vehicle, were installed in order to study the optimal control of the suspension for states feedback. The Linear Quadratic Regulator (LQR) technique combined with the root locus method for definition of the closed-loop poles were used. The H'IND.2'/H'INFINITE' controller with output feedback was implemented for sake of comparison. Independent controllers for each degree of freedom were simulated with MATLAB then implemented and tested in the vehicle using AD/DA converter installed in the microcomputer. The first goal of the controllers' design was the system stability when subject to disturbances that drive the air away from its nominal value. Except for the system with RLQ controller and rigid body model, the experimental results obtained with the prototype show that the closed-loop system is stable and exhibits a satisfactory transient response.The system with the RLQ controller and not-rigid body model exhibits the best performance among all the alternatives tested.

O controle e otimização da operação da suspensão eletromagnética de um protótipo de veículo MAGLEV são apresentados neste trabalho. Este tipo de sistema é inerentemente instável e altamente não linear, sendo um excelente exemplo para o estudo e comparação de diferentes metodologias de controle. Nos últimos anos, experiências significativas com veículos levitados têm ganhado força. A arquitetura do veículo consiste de quatro atuadores eletromagnéticos, quatro sensores de entreferros e quatro acelerômetros, todos instalados nos cantos do veículo. A dinâmica do sistema é descrita através de três modelos matemáticos diferentes: modelo de corpo rígido com três graus de liberdade (movimentos vertical, de rolagem e de arfagem), modelo de corpo não rígido, onde está presente mais um grau de liberdade (movimento de torção), e modelo SISO com quatro movimentos verticais independentes. Blocos de compatibilização de variáveis são utilizados devido à diferença no número de variáveis do modelo matemático com o número de variáveis medidas e de atuação disponíveis no veículo. Os acelerômetros, usados para estimar as velocidades do veículo, foram instalados com o objetivo de estudar-se o controle ótimo da suspensão por realimentação de estados com o emprego da técnica do Regulador Linear Quadrático (RLQ) e utilização do método do lugar das raízes para definição dos pólos de malha fechada. Para efeito de comparação foi implementado um controle H'IND.2'/H'INFINITO' por realimentação da saída. Controladores independentes para cada grau de liberdade foram simulados no software MATLAB e depois implementados e testados no veículo utilizando uma placa de aquisição de dados instalada no microcomputador. A preocupação principal do projeto dos controladores foi a estabilidade do sistema na ocorrência de perturbações que desviem o entreferro de seu valor nominal.Exceto para o sistema com controle RLQ e modelo de corpo rígido, os resultados experimentais obtidos com o protótipo mostram que o sistema em malha fechada é estável e apresenta uma resposta transiente satisfatória. O sistema com controle RLQ e modelo de corpo não rígido apresentou a melhor performance entre todas as alternativas testadas. / The control and optimization of the electromagnetic suspension operation of a MAGLEV vehicle prototype are presented in this work. This kind of system is inherently unstable and highly nonlinear, being an excellent example for the study and comparison of different control methodologies. In the last years significant experience has been gained with levitated vehicles. The vehicle architecture consists of four electromagnetic actuators, four air gap sensors and four accelerometers, all located at the corners of the vehicle. The vehicle dynamics is described through three different mathematical models: rigid body model, with three degrees of freedom (heave, roll and pitch motions), not-rigid body model, with an additional degree of freedom (torsion motion) and SISO model, with four independent heave motions. Blocks to make variables compatible are used in based on the difference between the number of variables of the mathematical model and the number of measured and actuation variables available in the vehicle. The accelerometers, used to estimate the speeds of the vehicle, were installed in order to study the optimal control of the suspension for states feedback. The Linear Quadratic Regulator (LQR) technique combined with the root locus method for definition of the closed-loop poles were used. The H'IND.2'/H'INFINITE' controller with output feedback was implemented for sake of comparison. Independent controllers for each degree of freedom were simulated with MATLAB then implemented and tested in the vehicle using AD/DA converter installed in the microcomputer. The first goal of the controllers' design was the system stability when subject to disturbances that drive the air away from its nominal value. Except for the system with RLQ controller and rigid body model, the experimental results obtained with the prototype show that the closed-loop system is stable and exhibits a satisfactory transient response.The system with the RLQ controller and not-rigid body model exhibits the best performance among all the alternatives tested.

No description available.

This thesis presents the design and fabrication of a novel magnetically levitated (maglev) device with six-degree-of-freedom motion capability at nanometer precision. The applications of this device are manufacture of nanoscale structures, assembly of microparts, vibration isolation of delicate instrumentation, and telerobotics. In this thesis, a single-moving stage is levitated by six maglev actuators. The total mass of the moving stage is 0.2126 kg. Three laser interferometers and three capacitance sensors are used to gather the position information. User interface and real-time control routines are implemented digitally on a VME PC and a digital-signal-processor (DSP) board. The underlying mechanical design and fabrication, electrical system setup, control system design, noise analysis, and test results are presented in this thesis. Test results show a quick step response in all six axes and a resolution of 2.5 nm rms in horizontal motion and 25 nm rms in vertical motion.

This thesis presents the design and fabrication of a novel magnetically levitated (maglev) device with six-degree-of-freedom motion capability at nanometer precision. The applications of this device are manufacture of nanoscale structures, assembly of microparts, vibration isolation of delicate instrumentation, and telerobotics. In this thesis, a single-moving stage is levitated by six maglev actuators. The total mass of the moving stage is 0.2126 kg. Three laser interferometers and three capacitance sensors are used to gather the position information. User interface and real-time control routines are implemented digitally on a VME PC and a digital-signal-processor (DSP) board. The underlying mechanical design and fabrication, electrical system setup, control system design, noise analysis, and test results are presented in this thesis. Test results show a quick step response in all six axes and a resolution of 2.5 nm rms in horizontal motion and 25 nm rms in vertical motion.

Many electrically propelled, and magnetically levitated and guided actuation systems (maglev) use either linear induction or synchronous machine topologies. From the cost, reliability, fault tolerance, and phase independence points of view, linear switched reluctance topologies are attractive for transportation application. This thesis investigates a novel topology in which a linear switched reluctance machine (LSRM) propulsion drive is incorporated in the magnetically levitated and guided vehicle. Designs of the LSRM and dc electromagnet, analytical aspects of modeling and dynamics of the vehicle, and closed loop control of propulsion, levitation, and guidance systems are discussed with comprehensive simulations and experimental results. Due to the lack of standard design procedure for LSRM, a novel design procedure is proposed using the current knowledge and design procedure of rotating switched reluctance machines. Analysis procedures for the phase winding inductance, propulsion and normal forces with translator position are developed with a lumped-parameter magnetic circuit model and the results from it are verified with two-dimensional finite element analysis. Extensive experimental correlation of inductance, propulsion and normal forces to validate the analysis and design procedure is presented. For the stable operation of the electromagnetic levitation and guidance systems, which have inherent unstable characteristics, the air gap position and force/current control loops are designed using PID (or PD) and PI controllers, respectively, and implemented and tested. The step-by-step design procedures for each controller are systematically derived. A feedforward compensation strategy for the levitation air gap control is proposed to reject the external force disturbance mainly caused by the normal force component generated in the LSRM propulsion drive system. The reduction of mechanical vibration and hence the enhancement of ride quality is achieved. Extensive dynamic simulations and experimental results for the integrated maglev system are presented with a 6 m long prototype system. Experimental correlation proves the validity of the controller design procedure based on the single-input and single-output model, and shows the feasibility of the LSRM-propelled electromagnetic levitation and guidance systems. A novel maglev topology in which only two sets of LSRMs are utilized to control individually propulsion, levitation, and guidance forces is proposed. One set of the linear switched reluctance actuator produces the levitation and propulsion forces and the other set generates the propulsion and guidance forces. The proposed architecture, thereby, obviates the need for design, development, and implementation of separate actuation systems for individual control of propulsion, levitation, and guidance forces and in contrast to most of the present practice. Further, the proposed system utilizes each of the linear switched reluctance actuation system for producing the propulsion force, thereby giving an overall high force density package for the entire system. The feasibility of the proposed system by finite element analysis is demonstrated. / Ph. D.

A multidisciplinary design optimization (MDO) methodology is developed to link the aerodynamic shape design to the system costs for magnetically levitated (MAGLEV) vehicles. These railed vehicles can cruise at speeds approaching that of short haul aircraft and travel just inches from a guideway. They are slated for high speed intercity service of up to 500 miles in length and would compete with air shuttle services. The realization of this technology hinges upon economic viability which is the impetus for the design methodology presented here. This methodology involves models for the aerodynamics, structural weight, direct operating cost, acquisition cost, and life cycle cost and utilizes the DOT optimization software. Optimizations are performed using sequential quadratic programming for a 5 design variable problem. This problem is reformulated using 7 design variables to overcome problems due to non-smooth design space. The reformulation of the problem provides a smoother design space which is navigable by calculus based optimizers. The MDO methodology proves to be a useful tool for the design of MAGLEV vehicles. The optimizations show significant and sensible differences between designing for minimum life cycle cost and other figures of merit. The optimizations also show a need for a more sensitive acquisition cost model which is not based simply on weight engineering. As a part of the design methodology, a low-order aerodynamics model is developed for the prediction of 2-D, ground effect flow over bluff bodies. The model employs a continuous vortex sheet to model the solid surface, discrete vortices to model the shed wake, the Stratford Criterion to determine the location of the turbulent separation, and the vorticity conservation condition to determine the strength of the shed vorticity. The continuous vortex sheet better matches the mechanics of the flow than discrete singularities and therefore better predicts the ground effect flow. The predictions compare well with higher-order computational methods and experimental data. A 3-D extension to this model is investigated, although no 3-D design optimizations are performed. NOTE: An updated copy of this ETD was added on 05/29/2013. / Ph. D.