7th IFAC Symposium on Advances in Automotive Control, The International Federation of Automatic Control, September 4-7, 2013. Tokyo, Japan

No description available.

This dissertation consists of two topics: (1) structure-inspired disturbance observer design and (2) disturbance observer-based control/estimation. The disturbance is defined as the discrepancy between a model and the system the model represents. A disturbance observer is an algorithm that generates an estimate of the disturbance. The first topic illustrates a disturbance observer that provides a big class of nonlinear systems with a large basin of attraction, even ensuring global convergence. Such robustness is achieved by leveraging particular system nonlinearities in the observer design. The second topic discusses the usage of disturbance estimates to counteract or capture the effects of disturbances to recover the nominal controller/estimator performance. The main research results are theorems concerning stability analysis of the disturbance observer and the disturbance observer-based systems, whose practical aspects are supported by three application examples---a fixed-wing aircraft, an underwater vehicle, and a Furuta pendulum. / Doctor of Philosophy / This dissertation consists of two topics: (1) structure-inspired disturbance observer design and (2) disturbance observer-based control/estimation. Disturbances are the unknown signals entering the system; an external force, for example, such as the additional lift force due to turbulence surrounding an aircraft is a disturbance. A disturbance observer is an algorithm that estimates the mathematical value of disturbances. The first topic illustrates a disturbance observer whose convergence is guaranteed regardless of the initial condition. Such robustness is achieved by leveraging the system's special properties in the observer design. The second topic discusses the usage of disturbance observers to recover the nominal controller/estimator performance. Control is a study of how make systems behave ideally by properly designing the inputs, while estimation is about how to infer quantities that cannot be directly measured using the measurements that really are available; the solutions are correspondingly called controller and estimator. Disturbance estimates can be exploited by existing controllers and estimators as extra information to counteract or capture the effects of disturbances. The main research results are theorems about the conditions under which these algorithms perform as desired. Practical aspects are supported by three application examples---a fixed-wing aircraft, an underwater vehicle, and a Furuta pendulum.

No description available.

No description available.

One of the most common goals in manufacturing is to improve the quality and accuracy of the parts being fabricated without reducing productivity. Aiming at this goal, many different manufacturing processes have been developed. Among them, machining plays a major role in increasing product accuracy. As an important machining process, grinding is a vital step that can produce both fine finish and dimensional accuracy for applications in which the workpiece material is either hard or brittle. Currently, the ball screw is the most frequently used setup for grinding machine tool feed drive. However, the existence of transmission components induces wear, high friction, backlash, and also lower system stiffness; therefore, applications of conventional feed drives for high speed and high accuracy machining are very limited. As a promising technology, a linear motor feed drive discards the transmission system; therefore, it eliminates transmission induced error, such as backlash and pitch error, and avoids stiffness reduction as well. As a result, a linear motor drive can achieve both high speed and high accuracy performance. A linear motor feed drive will be subject to external disturbances such as friction, force ripple and machining force. Due to the lack of a transmission unit, the tracking behavior of a linear motor feed drive is prone to be affected by external disturbances and model parameter variations. Thus, in order to deliver high performance, a controller should be capable of achieving high accuracy in the presence of external disturbance and parameter uncertainty. This dissertation proposes a general robust motion control framework for the CNC design of a linear motor feed drive to achieve high speed/high precision as well as low speed/high precision. An application to the linear motor feed drives in grinding machines was carried out. One of the developed algorithms is the HSMC, which combines the merits of a reaching law based sliding mode control and a modified disturbance observer for precision tracking to address the practical issues of friction, force ripple, and grinding force disturbances. Another algorithm presented is ASMC, which combines the reaching law based sliding mode control with adaptive disturbance estimation to achieve an adaptive robust motion control.

No description available.

The eciency of electrical machines is of major concern due to their widespread usage and the globally increasing awareness of energy consumption issues. Iron losses have a signicant impact on the total and thus researchers and manufacturers of electrical machines are developing dierent strategies in order to reduce them. The iron losses are highly dependent on the magnetic material that is used and thus it is necessary to identify its relevant characteristics. In this work, the development of a control system for inducing a pure sinusoidal magnetic ux density in the magnetic material is described. This is necessary in order to perform characterisation of the magnetic material. The main diculty is the highly non-linear and hysteretic relationship between the magnetic eld strength and the magnetic ux density. In order to mitigate the eect of the hysteresis, a mathematical inverse model was used in the control system. To nd a suitable model, an extensive study of literature was performed and discussed in this work. The Preisach model and its dynamic extension was chosen as the most suitable approach. A detailed description of both theory and implementation details is provided in this work. Furthermore, the model is validated by comparing simulation against measurement data for two dierent materials. In the last part of this work, the inverse model is combined with a controller to form a feedback control system. Two dierent control schemes are investigated: a simpler PI controller and a more elaborate disturbance observer (DOB) based control scheme. The DOB is used to observe the hysteresis inversion error and the observation is used to correct for the error. The controller's ability to produce a pure sinusoidal magnetic ux density was assessed by simulations with dierent magnetic materials at varying frequencies. / Verkningsgraden for elmotorer ar av okande intresse pa grund av deras omfattande anvandning och vaxande oro for globala energiforbrukningsfragor. Jarnforluster har ett stort inytande i de totala forlusterna och ar darfor ett viktigt omrade for forskare och tillverkare av elektriska motorer. Jarnforlusterna beror till stor del av det magnetiska materialet som anvands i konstruktion av elmotorer och det ar darfor nodvandigt att identiera materialets egenskaper. I det har arbetet beskrivs utvecklingen av ett reglersystem for att inducera en ren sinusformat magnetisk odestathet i ett magnetiskt material. Detta ar nodvandigt for att kunna bestamma det magnetiska materialets egenskaper under kontrollerade forhallanden. Huvudsvarigheten ar det icke-linjara sambandet mellan magnetiska faltstyrkan och odest atheten. Sambandet formar en hysteres och for att eliminera dess inytande anvandes en matematisk invers model. For att hitta en lamplig model genomfordes en literaturstudie och Preisach modellen och dess dynamiska utokning valdes. I detta arbete nns en detaljerad beskrivning av bade teorin bakom modellen och dess implementering. Modellen utvarderades genom att jamfora matvarden med simulationsresultat for olika magnetiska material. I sista delen av detta arbete kombineras inversmodellen med ett reglerssystem for att kunna uppna en sinusformat odestathet i det magnetiska materialet. Tva olika regleralgoritmer utvarderas, en enklare PI-regulator och en regulator som inkluderar en sa kallat "Disturbance Observer" (DOB). DOB:n anvandes for att observera felet som uppstar vid invertering av hysteresen och for att korrigera felet. De bada regulatorernas formaga attaterskapa en ren sinusformat magnetisk odestathet testas genom att genomfora simulationer for olika magnetiska material vid varierande frekvenser.

As ondulações na corrente do estator e o forte impacto de erros em parâmetros são apontados como desvantagens do controle preditivo finite control-set (FCS) quando aplicado no controle de máquinas elétricas trifásicas. Neste contexto, o objetivo deste trabalho é propor métodos para aprimorar tanto o desempenho em regime permanente quanto à robustez frente a erros paramétricos do controle preditivo de corrente, conhecido como model predictive current control (MPCC). No presente trabalho são propostas as abordagens MPCC baseada na aplicação de dois e três vetores de tensão durante o mesmo período de controle com intuito de aprimorar o desempenho do controle preditivo de corrente, sem a necessidade de elevar a frequência de amostragem do sistema. Os princípios para escolha dos vetores ótimos e seus respectivos tempos de aplicação são baseados nas expressões para controle desacoplado de torque e fluxo do FOC e no princípio de funcionamento do deadbeat. A robustez paramétrica é garantida por meio de um observador de distúrbios discreto incorporado ao método de controle MPCC. Com a estimação dos impactos causados por variações de parâmetros e dinâmicas não modeladas é possível corrigir os erros dos sistema de controle, mas mantendo-se as suas características transitórias e de regime permanente. Resultados de simulação computacional bem como resultados obtidos em bancada de ensaios experimentais são apresentados para avaliação e comprovação das propostas. / Ripples in the stator current and the strong impact of errors in parameters are pointed as disadvantages of the predictive control when applied in the control of three-phase electric machines. In this context, the objective of this work is to propose methods to improve both the steady-state performance and the robustness against parametric errors of the current predictive control, known as MPCC. In the present work the MPCC approaches are proposed based on the application of two and three voltage vectors during the same control period in order to improve the performance of the current predictive control, without the need to raise the sampling frequency of the system. The principles for choosing the optimal vectors and their respective duration are based on the expressions for FOC method and on the concept of deadbeat operation. In addition, the impact of errors on parameters is minimized by means of a discrete disturbance observer embedded in the MPCC control method. With the estimation of the impacts caused by changes of parameters and not modeled dynamics it is possible to correct the errors of the control system, but keeping their transient and steady state characteristics. Results of computational simulation as well as results obtained in experimental tests are presented for evaluation and proof of the proposals.

The research in this thesis is undertaken by observing that modern systems are becoming more and more complex and safety-critical due to the increasing requirements on system smartness and autonomy, and as a result health monitoring system needs to be developed to meet the requirements on system safety and reliability. The state-of-the-art approaches to monitoring system status are model based Fault Diagnosis (FD) systems, which can fuse the advantages of system physical modelling and sensors' characteristics. A number of model based FD approaches have been proposed. The conventional residual based approaches by monitoring system output estimation errors, however, may have certain limitations such as complex diagnosis logic for fault isolation, less sensitiveness to system faults and high computation load. More importantly, little attention has been paid to the problem of fault diagnosis system verification which answers the question that under what condition (i.e., level of uncertainties) a fault diagnosis system is valid. To this end, this thesis investigates the design and verification of fault diagnosis algorithms. It first highlights the differences between two popular FD approaches (i.e., residual based and fault estimation based) through a case study. On this basis, a set of uncertainty estimation algorithms are proposed to generate fault estimates according to different specifications after interpreting the FD problem as an uncertainty estimation problem. Then FD algorithm verification and threshold selection are investigated considering that there are always some mismatches between the real plant and the mathematical model used for FD observer design. Reachability analysis is drawn to evaluate the effect of uncertainties and faults such that it can be quantitatively verified under what condition a FD algorithm is valid. First the proposed fault estimation algorithms in this thesis, on the one hand, extend the existing approaches by pooling the available prior information such that performance can be enhanced, and on the other hand relax the existence condition and reduce the computation load by exploiting the reduced order observer structure. Second, the proposed framework for fault diagnosis system verification bridges the gap between academia and industry since on the one hand a given FD algorithm can be verified under what condition it is effective, and on the other hand different FD algorithms can be compared and selected for different application scenarios. It should be highlighted that although the algorithm design and verification are for fault diagnosis systems, they can also be applied for other systems such as disturbance rejection control system among many others.