Made available in DSpace on 2016-06-02T20:06:05Z (GMT). No. of bitstreams: 1 4180.pdf: 1039417 bytes, checksum: d09a61a4895fb47d1c2456468800fc2f (MD5) Previous issue date: 2012-02-29 / Financiadora de Estudos e Projetos / In this work we consider a new estimator proposed by Ferrari & Yang (2010), called the maximum Lq-likelihood estimator (MLqE), to estimate the parameters of the measurement error models, in particular, the structural model. The new estimator extends the classical maximum likelihood estimator (MLE) and its based on the minimization, by means of the Kullback-Leibler (KL) divergence, of the discrepancy between a distribuiton in a family and one that modifies the true distribution by the degree of distortion q. Depending on the choice of q, the transformed distribution can diminish or emphasize the role of extreme observations, unlike the ML method that equally weights each observation. For small and moderate sample sizes, the MLqE can trade bias for precision, causing a reduction of the mean square error (MSE). The structural model has the characteristic of non-identifiability. For this reason, we must make assumptions on the parameters to overcome the non-identifiability. We perform a analytical study and a simulation study to compare MLqE and MLE. To gauge performance of the estimators, we compute measures of overall performance, bias, standard deviation, standard error, MSE, probability of coverage and length of confidence intervals. / Neste trabalho utilizaremos um novo estimador proposto por Ferrari & Yang (2010), denominado de estimador de máxima Lq-verossimilhança (EMLqV), na estimação dos parâmetros de modelos com erros de medição estruturais normais. O novo estimador é uma generalização do estimador de máxima verossimilhança (EMV) usual e sua construção baseia-se na comparação, utilizando divergência de Kullback-Leibler (KL), entre duas distribuições, a distribuição inalterada e a distribuição modificada pelo grau de distorção da função de verossimilhança (q). Conforme a escolha para q, a distribuição modificada poderá atenuar ou exaltar o papel das observações extremas, diferentemente do EMV usual que atribui os mesmos pesos a todas as observações. Na comparação entre as duas distribuições pela divergência de KL é inserida certa quantidade de viés no estimador resultante, que é controlada pelo parâmetro q. O aumento do viés do estimador MLqV pode ser compensado com a redução de sua variância, pela escolha apropriada de q. O modelo estrutural possui a característica de ser inidentificável. Para torná-lo identificável faremos suposições sobre os parâmetros do modelo, analisando cinco casos de identificabilidade do modelo. A comparação entre os métodos MLqV e MV na estimação dos parâmetros do modelo será baseada em resultados analíticos e em simulações, sendo calculadas medidas de desempenho global, viés, desvio padrão (DP), erro padrão estimado (EP), erro quadrático médio (EQM), probabilidade de cobertura e amplitude dos intervalos de confiança.

Syftet med föreliggande studie var att validera ett instrument (LQ-test) som avser att mäta ledarintelligens baserat på teorin om ledarintelligens utifrån chefs- och ledarrollen. Ledarintelligens, skapat av Ronthy (Ronthy 2006; 2013) är ett holistiskt perspektiv på ledarskap och omfattar färdigheter från tre intelligenser; rationell, emotionell och själslig intelligens med större betoning på ledarrollen med färdigheter tillhörande emotionell och själslig intelligens även kallat medmänskligt ledarskap (ML). LQ-testet bestod av 71 item och konstruerades av Ronthy. Majoriteten av dessa item är inhämtade från Larsson et al. (2003). Den rationella skalan tillskrivs chefsrollen och de övriga två skalorna tillskrivs ledarrollen. Ytterligare ett syfte var att utifrån instrumentet undersöka ledarskapets fördelning mellan kvinnor och män. Studiens deltagare var 307 chefer (68% kvinnor). Utöver LQ-testet fick även deltagarna besvara hur mycket arbetstid per månad som de använde för olika uppgifter utifrån ledarintelligensbegreppet. Enligt teorin om ledarintelligens ska tiden användas mer till att leda sig själv och andra vilket ingår i ledarrollen till skillnad från chefsrollen som ska tillgodose uppgifter som omfattar verksamhetens administration och styrning. Resultatet visade att 20% uppfattade sig som chefer med betoning på administrativa uppgifter och 80% uppfattade sig som ledare med betoning på ett medmänskligt ledarskap. Vid en genomgång av chefernas tidsåtgång för de olika arbetsuppgifterna som indelats utifrån de tre skalorna, framkom att 72% av arbetstiden användes till chefsuppgifter och 28% av arbetstiden ägnades åt ledaruppgifter såsom coachning, personlig utveckling och reflektion. Studien visade inga signifikanta skillnader mellan kvinnor och män avseende uppdelningen chef och ledare. Den relationella aspekten av ledarskapet betonas alltmer i dagens organisationer och dess innebörd har förtydligats i denna studie. / The aim of the present study was to validate an instrument called Leader Intelligence Questionnaire (LQ-test) based on the theory of leader intelligence out of being a manager and a leader. Leader intelligence is a concept created by Ronthy observing skills within the rational, emotional and spiritual intelligence with the main focus on skills within the latter two which also is a human leadership. The questionnaire of leader intelligence is created by Ronthy and has 71 item, and the majority are from Larsson et al. (2003). The instrument has three scales to study all three intelligences. The scale for the rational intelligence studies the role as a manager and the emotional and spiritual intelligence studies the role as a leader. A second aim was to examine the gender distribution among managers and leaders. The participants in the study were 307 leaders (68% women). Furthermore, the participants also responded how much time they spent per month on tasks related to the management role as well as the leadership role. According to the leader intelligence theory the manager should spend more of his/her time in the leader role leading him / herself as well as others and less time in the managerial role managing administrative tasks. The result shows that 20% perceive themselves as managers with the emphasis on administrative tasks and 80% perceive themselves as leaders with the emphasis on a human leadership. When analysing the time spent on various tasks divided within the three different scales the study showed that 72% of the time was spent on managerial tasks and 28% on leader tasks as coaching, personal development and reflection. The study showed no significant differences between men and women. The relational aspect of leadership is being empathized more and more in today´s organizations which will be clarified in this study.

This master's thesis considers adaptive optimal controllers. It shows principles of optimal controllers, recursive identification using least-mean squares method and identification based on neural network.

Master’s thesis deals with using a linear matrix inequality (LMI) in control of a dynamic systems. We can define a stability of a dynamic system with a LMI. We can use a LMI for research if the poles of a system are in a given regions in the left half-plane of the complex plane with a LMI or we can use a LMI for a state feedback control. In the work we describe a desing of a controller minimizing a norm from an input to an output of the system. There is also a desing of a LQ controller with a LMI. In the end of the work, there are two examples of a design a LQ controller, which minimize the norm from the input to the output of the system and moves a poles of a dynamic system in a given regions in the complex plane, with the LMI. We use a LMI for a design a continuos LQ controller in the first example. In the second example we use a LMI for a design a discrete LQ controller.

Electro-Hydraulic (EH) systems are commonly used in the industry for applications that require high power-weight ratios and large driving forces. The EH system studied in this master thesis have recently been upgraded with new hardware components and as a part of this upgrade a new controller was requested. The system consists of a controller that computes a control signal for an electric motor. The motor drives a gear pump that generates a flow of hydraulic fluid. The flow is then directed to a cylinder. The movements of a piston in the cylinder is affected by the flow and the piston position can be measured. The measured piston position is then fed back to the controller and the control loop is complete. The system was previously controlled using a Proportional-Integral-Derivative (PID) controller and the purpose of this thesis is to compare the old controller with alternative control strategies suitable for this application. The evaluation of the controllers is based on both software and hardware simulations and results in a recommendation for final implementation of the best suited controller. The control strategies chosen for investigation are: a retuned PID controller, a PID controller with feed forward from reference, a PID based cascade controller, a Linear Quadratic (LQ) controller, and a Model Predictive Controller (MPC). To synthesize the controllers an approximate model of the system is formed and implemented in the software environment Matlab Simulink. The model is tuned to fit recorded data and provides a decent estimation of the actual system. The proposed control strategies are then simulated and evaluated in Simulink with the model posing as the real system. These simulations resulted in the elimination of the cascade controller as a possible candidate since it proved unstable for large steps in the reference signal. The remaining four controllers were all selected for simulation on the real hardware system. Unfortunately the MPC was never successfully implemented on the hardware due to some unknown compatibility error and hence eliminated as a possible candidate. The three remaining control strategies, PID, PID with feed forward from reference and the LQ controller, were all successfully implemented and simulated on hardware. The results from the hardware simulations compared to simulations made with the old controller, as well as the results from the software simulations, were then evaluated. Depending on the purpose one of two control strategies is recommended for this application. The LQ controller achieved the best overall performance and is presented as the control strategy best suited for this application.

This thesis work contains a stabilization and a motion planning strategy for a truck and trailer system. A dynamical model for a general 2-trailer with two rigid free joints and a kingpin hitching has been derived based on previous work. The model holds under the assumption of rolling without slipping of the wheels and has been used for control design and as a steering function in a probabilistic motion planning algorithm. A gain scheduled Linear Quadratic (LQ) controller with a Pure pursuit path following algorithm has been designed to stabilize the system around a given reference path. The LQ controller is only used in backward motion and the Pure pursuit controller is split into two parts which are chosen depending on the direction of motion. A motion planning algorithm called Closed-Loop Rapidly-exploring Random Tree (CL-RRT) has then been used to plan suitable reference paths for the system from an initial state configuration to a desired goal configuration with obstacle-imposed constraints. The motion planning algorithm solves a non-convex optimal control problem by randomly exploring the input space to the closed-loop system by performing forward simulations of the closed-loop system. Evaluations of performance is partly done in simulations and partly on a Lego platform consisting of a small-scale system. The controllers have been used on the Lego platform with successful results. When the reference path is chosen as a smooth function the closed-loop system is able to follow the desired path in forward and backward motion with a small control error. In the work, it is shown how the CL-RRT algorithm is able to plan non-trivial maneuvers in simulations by combining forward and backward motion. Beyond simulations, the algorithm has also been used for planning in open-loop for the Lego platform. / <p>Links to movies:</p><p>Reference tracking on Lego platform:</p><p>https://www.dropbox.com/s/ebtfgfo7aq9ij8w/reference_tracking.m4v?dl=0</p><p></p><p>Motion planning simulation with CL-RRT:</p><p>https://www.dropbox.com/s/z9kk27cxdxc1llp/CL_RRT_motion_planning.wmv?dl=0</p>

Na základě Monte Carlo simulace bylo provedeno porovnání estimátorů klasických a některých nově navržených měr variability, relativní variability, šikmosti a kurtozy. Z celkem 40 různých populací bylo pořízeno vždy 16 000 výběrů o rozsahu n = {7; 15; 23; 31; 47; 63; 100; 200; 350; 500; 1000}. Z každého výběru byly vypočteny hodnoty estimátorů měr založených na momentech, kvantilech, L momentech a jejich modifikacích a robustních měr založených na mediánu funkce lineární kombinace pořádkových statistik. Na základě experimentu bylo provedeno porovnání výběrových vlastností jednotlivých estimátorů z hlediska variability, vychýlení a rychlosti konvergence jejich výběrových rozdělení k normalitě. U estimátorů vybraných charakteristik byla dále na základě experimentálně odhadnuté průměrné síly testu posouzena vhodnost jejich použití jako testového kritéria při testu o rozdělení, z něhož výběr pochází. Zároveň byla porovnána síla závislosti estimátorů nově navržených charakteristik s estimátory momentovými s cílem posoudit, zda je možné danou charakteristiku skutečně považovat za vhodnou alternativu charakteristiky momentové. Výsledky ukazují, že estimátory momentových měr jsou vyhovující pro popis souboru pouze při výběrech z populací nepříliš odlišných od normálního rozdělení. S rostoucí odlišností od normality rychle roste relativní variabilita i vychýlení jejich estimátorů a projevují se různé anomálie v jejich chování. Vhodnou alternativou k mírám založeným na klasických momentech i kvantilech by se mohly stát míry založené na L momentech, jejichž estimátory vykazují ve většině případů nejlepší výběrové vlastnosti a zároveň vykazují vysokou míru závislosti s hodnotami estimátorů momentových charakteristik. Modifikace L-momentů, tzv. LQ- a TL-momenty, nepřinášejí oproti mírám založeným na L-momentech žádné zlepšení, v některých ohledech vykazují výrazně horší vlastnosti.

The objective of this dissertation is to provide a framework for constructing a transitional behavior, connecting any two trajectories from a set with a particular characteristic, in such a way that the transition is as inconspicuous as possible. By this we mean that the connection is such that the characteristic behavior persists during the transition. These special classes include stationary solutions, limit cycles etc. We call this framework the Gluskabi raccordation. This problem is motivated from physical applications where it is often desired to steer a system from one stationary solution or periodic orbit to another in a ̒smooth̕ way. Examples include motion control in robotics, chemical process control and quasi-stationary processes in thermodynamics, etc. Before discussing the Gluskabi raccordations of periodic behaviors, we first study several periodic phenomena. Specifically, we study the self- propulsion of a number of legless, toy creatures based on differential friction under periodic excitations. This friction model is based on viscous friction which is predominant in a wet environment. We investigate the effects of periodic and optimal periodic control on locomotion. Subsequently, we consider a control problem of a stochastic system, under the basic constraint that the feedback control signal and the observations from the system cannot use the communication channel simultaneously. Hence, two modes of operation result: an observation mode and a control mode. We seek an optimal periodic regime in a statistical steady state by switching between the observation and the control mode. For this, the duty cycle and the optimal gains for the controller and observer in either mode are determined. We then investigate the simplest special case of the Gluskabi raccordation, namely the quasi-stationary optimal control problem. This forces us to revisit the classical terminal controller. We analyze the performance index as the control horizon increases to infinity. This problem gives a good example where the limiting operation and integration do not commute. Such a misinterpretation can lead to an apparent paradox. We use symmetrical components (the parity operator) to shed light on the correct solution. The main part of thesis is the Gluskabi raccordation problem. We first use several simple examples to introduce the general framework. We then consider the signal Gluskabi raccordation or the Gluskabi raccordation without a dynamical system. Specifically, we present the quasi-periodic raccordation where we seek the maximally ̒smooth̕ transitions between two periodic signals. We provide two methods, the direct and indirect method, to construct these transitions. Detailed algorithms for generating the raccordations based on the direct method are also provided. Next, we extend the signal Gluskabi raccordation to the dynamic case by considering the dynamical system as a hard constraint. The behavioral modeling of dynamical system pioneered by Willems provides the right language for this generalization. All algorithms of the signal Gluskabi raccordation are extended accordingly to produce these ̒smooth̕ transition behaviors.

Dans le contexte très général de la détection radar, les détecteurs classiques, basés sur l'hypothèse d'un bruit Gaussien, sont souvent mis en défaut dès lors que l'environnement (fouillis de sol, de mer) devient inhomogène, voire impulsionnel, s'écartant très vite du modèle Gaussien. Des modèles physiques de fouillis basés sur les modèles de bruit composé (SIRP, Compound Gaussian Processes) permettent de mieux représenter la réalité (variations spatiales de puissance et nature de fouillis, transitions, ...). Ces modèles dépendent cependant de paramètres (matrice de covariance, loi de texture, paramètres de "disturbance") qu'il devient nécessaire d'estimer. Une fois ces paramètres estimés, il est possible de construire des détecteurs radar optimaux (Generalized Likelihood Ratio Test - Linear Quadratic) pour ces environnements. Cette thèse, qui s'appuie sur ces modèles, propose une analyse complète de diverses procédures d'estimation de matrices de covariance, associées à ce problème de détection. Une étude statistique des principaux estimateurs de matrice de covariance, utilisés actuellement, est réalisée. De plus, un nouvel estimateur est proposé: l'estimateur du point fixe, très attractif grâce à ses bonnes propriétés statistiques et "radaristiques".<br />Elle décrit également les performances et les propriétés théoriques (SIRV-CFAR) du détecteur GLRT-LQ construits avec ces nouveaux estimateurs. En particulier, on montre l'invariance du détecteur à la loi de la texture mais également à la matrice de covariance régissant les propriétés spectrales du fouillis. Ces nouveaux détecteurs sont ensuite analysés sur des données simulées mais également testés sur des données réelles de fouillis de sol.

No description available.