Thesis (Ph.D.(Electronic Engineering))--University of Pretoria, 2009. / Summaries in Afrikaans and English. Includes bibliographical references.

Η παρούσα διπλωματική εργασία ασχολείται με τη μελέτη και τον έλεγχο ενός συστήματος γερανού με προηγμένες τεχνικές. Το σύστημα αποτελείται από έναν κινούμενο φορέα στον οποίο είναι συνδεδεμένο ένα εκκρεμές φορτίο με μεταβαλλόμενο μήκος. Σκοπός είναι η μεταφορά του φορτίου σε μια διαφορετική θέση και ύψος σε ελάχιστο χρόνο με μηδενική υπολειπόμενη ταλάντωση. Αρχικά παρουσιάζεται ο έλεγχος του συστήματος με τη μέθοδο ανατροφοδότησης κατάστασης. Στη συνέχεια αναλύεται η βασική θεωρία του input shaper και προτείνεται μια προσαρμοστική τεχνική για την εφαρμογή του σε σύστημα με μεταβαλλόμενο μήκος σχοινιού. Έπειτα, η τεχνική αυτή βελτιώνεται συνδυάζοντάς την με τον έλεγχο ανατροφοδότησης κατάστασης και τέλος εφαρμόζεται ένας μη γραμμικός ελεγκτής για το πλήρες μη γραμμικό σύστημα. Τα αποτελέσματα παρουσιάζονται με εκτενείς προσομοιώσεις καθώς και με εργαστηριακές μετρήσεις σε κατάλληλη πειραματική διάταξη. / This diploma thesis includes the analysis and control of a crane system using advanced control techniques. The system consists of a carrier mass connected with a playload with a hoisting mechanism. The aim of this work is to transport the payload to a different position and a different level within minimum time period while simutaneously minimizing the residual vibration. At first, a state feedback control is presented. Furthermore, the input shaping technique is analyzed and an adaptive version of the input shaper is proposed including hoisting cases. Moreover, this technique is significally improved when combined with a state feedback control law and finally a nonlinear controller is proposed for the total nonlinear system. The above techniques are justified with both simulation and experimental results from a suitable experimental setup.

Made available in DSpace on 2016-03-07T19:21:17Z (GMT). No. of bitstreams: 0 Previous issue date: 2015-11-27. Added 1 bitstream(s) on 2016-03-07T19:25:28Z : No. of bitstreams: 1 000859488.pdf: 1084542 bytes, checksum: b2c792e2ab102fae0335228868786b51 (MD5) / Neste trabalho são apresentadas duas novas propostas de condições de estabilidade para sistemas não lineares descritos através de modelos fuzzy Takagi-Sugeno (TS) baseadas em funções de Lyapunov fuzzy (FLF). A primeira proposta e baseada na determinação algébrica do hiper-retângulo composto pelo conjunto fechado das derivadas temporais das funções de pertinência de sistemas fuzzy TS e é formulada em termos de desigualdades matriciais lineares, em inglês Linear Matrix Inequalities (LMIs). A segunda proposta é baseada na obtenção também algébrica, de um politopo formado pela interseção do hiper-plano das derivadas temporais das funções de pertinência com o hiper-retângulo. são propostos dois teoremas que fornecem condições necessárias e suficientes para esse problema de estabilidade. Exemplos, utilizando o software MATLAB, ilustram os resultados numéricos, mostrando que o hiper-retângulo fornece resultados mais relaxados, no sentido de reduzir o conservadorismo, em relação aos da literatura e a estrutura proposta através de politopos apresenta os mesmos resultados com tempos de execução computacionais consideravelmente menores. Portanto, as propostas apresentadas generalizam as já existentes na literatura, para o estudo da estabilidade assintótica de sistemas não lineares baseados em FLF, para sistemas fuzzy TS / This thesis proposes two new stability conditions regarding nonlinear systems des- cribed by Takagi-Sugeno (TS) fuzzy models, based on Fuzzy Lyapunov Functions (FLF). The first result is based on the hyper-rectangle defined by the closed set of the time deriva- tives of the membership functions of TS fuzzy systems, and is described by a set of Linear Matrix Inequalities (LMIs). The second result is based on the obtention of a polytope composed by the intersection of the hyperplane of the time derivatives of the membership functions with the hyper-rectangle. Two proposed theorems offer necessary and sufficient conditions for this stability problem. Examples, using the MATLAB software, illustrate the numerical results and show that the hyper-rectangle allows the reduction of the con- servatism of the stability analysis considering similar results available in the literature and with the polytope it is possible to obtain the same results with a smaller computati- onal time. Therefore, the proposed methods generalize previous results about asymptotic stability available in the literature, for nonlinear systems described by TS fuzzy models, based on FLF

Thesis (MTech (Electrical Engineering))--Cape Peninsula University of Technology, 2009 / More than 70% of the earth's surface is covered by water, only a small part of which is suitable for either human consumption or agricultural use. Due to pollution from agriculture, households and industry reaching rivers, lakes and seas it is Important for wastewater to be properly treated in order to remove harmful substances before it reaches the environment. Strict environmental and health regulations together with a demand for cost effective ways of wastewater treatment have made control technology in wastewater Treatment Plants an important priority. Dissolved oxygen (DO) is the amount of oxygen in the effluent and it plays a vital role of controlling VV\YTP. Oxygen dissolves in water through mixing water surface with the atmosphere, The dissolved oxygen concentration in the aerobic part of an activated sludge process should be sufficiently high to supply enough oxygen to the microorganisms in the sludge. an excessive high DO leads to high energy consumption and may also deteriorate the sludge quality, A high DO concentration in the internally recirculated water also makes the denitrification less efficient Hence, both for economical and process reasons, it is of interest to control the DO. The used controllers are normally linear controllers, proportional integral (PI) or proportional integral derivative (PID) ones. The work of these controllers leads to bad system performance, because, the process of dissolving oxygen into the wastewater is a nonlinear process and requires nonlinear control. The aim of the research project is to develop methods for design of linear and nonlinear controllers of the concentration of the DO in the aeration tank of the WWTP and to implement the designed controllers in the frameworks of PLC. The nonlinear linearizing controller based on a reference model and Lyapunov second method is designed. Additionally a linear controller is developed in a form of PI controller based on pole placement method to improve, the performance of the closed loop system. The resultant controller is to be on a PLC as a part of Adroit SCADA system. The developed programmes are used to control the wastewater treatment process in laboratory scale plant and can be applied as a part of SCADA software for control of the wastewater treatment plants.

Thesis (DTech (Electrical Engineering))--Cape Peninsula University of Technology, 2016. / Majority of the industrial systems encountered are significantly non-linear in nature, so if they are synthesised and designed by linear methods, then some of salient features characterising of their performance may not be captured. Therefore designing a control system that captures the nonlinearities is important. This research focuses on the control design strategies for the Continuous Stirred Tank Reactor (CSTR) process. To control such a process a careful design strategy is required because of the nonlinearities, loop interaction and the potentially unstable dynamics characterizing the system. In these systems, linear control methods alone may not perform satisfactorily. Three different control design strategies (Dynamic decoupling, Decentralized and Input-output feedback linearization controller) are proposed and implemented .in the Matlab/Simulink platform and the developed strategies are then deployed to the design of distributed automation control system configuration using the IEC 61499 standard based functional block programming language. Twin CAT 3.1 system real-time and Matlab/Simulink (www.mathworks.com) environment are used to test the effectiveness of the models The simulation results from the investigation done between Simulink and TwinCAT 3 software (Beckhoff Automation) platforms in the case of the model transformation and closed loop simulation of the process for the considered cases have shown the suitability and the potentials of merging the Matlab/Simulink control function blocks into the TwinCAT 3.1 function blocks in real-time. The merits derived from such integration imply that the existing software and software components can be re-used. This is in line with one of the IEC 6144 standard requirements such as portability and interoperability. Similarly, the simplification of programming applications is greatly achieved. The investigation has also shown that the integration the of Matlab/Simulink models running in the TwinCAT 3.1 PLC do not need any modification, hence confirming that the TwinCAT 3.1 development platform can be used for the design and implementation of controllers from different platforms. Also, based on the steps required for model transformation the between the Matlab/Simulink to the TwinCAT 3 functional blocks, the algorithms of the control design methodologies developed, simulation results are used to verify the suitability of the controls to find whether the effective set-point tracking control and disturbance effect minimisation for the output variables can be achieved in real-time using the transformed Simulink blocks to the TwinCAT 3 functional blocks, then downloaded to the Beckhoff CX5020 PLC for real-time execution. Good set-point tracking control is achieved for the MIMO closed loop nonlinear CSTR process for the considered cases of the developed control methodologies. Similarly, the effects of disturbances are investigated. TwinCAT functional modules achieved good set-point tracking with these disturbances minimization under all the cases considered.

In this work we address the problem of friction compensation in a pneumatic control valve. It is proposed a nonlinear control law that uses algebraic estimators in its structure, in order to adapt the controller to the aging of the valve. For that purpose we estimate parameters related to the valve\'s Karnopp model, necessary to friction compensation, online. The estimators and the controller are validated through simulations. / Nessa pesquisa, estudamos o problema de compensação de atrito em válvulas pneumáticas. É proposta uma lei de controle não linear que tem estimadores algébricos em sua estrutura, para adaptar o controlador ao envelhecimento da válvula. Para isso, estimam-se os valores de parâmetros relacionados ao modelo de Karnopp da válvula, necessários à compensação do atrito, de maneira online. Os estimadores e o controlador são validados através de simulações.

Neste trabalho estuda-se a aplicação de um controlador preditivo não linear em reatores de policondensação visando à segurança, qualidade do produto, capacidade de “scale-up" e a produtividade. Nos processos de policondensação em batelada há formação de um produto volátil conhecido como condensado e a remoção deste torna-se necessária favorecer o avanço da reação a fim de obter alto grau de conversão. Na maioria dos trabalhos envolvendo a policondensação a separação não é levada em conta. Diante disso, desenvolveu-se um modelo de policondensação acoplado com a coluna de destilação a fim de verificar através das simulações em malha aberta o comportamento das variavéis de operação no sistema. O modelo de policondensação foi obtido na literatura (Shin et al.,1999) e em sua implementação foi levada em conta a inercia térmica entre as paredes do reator e as interações com a coluna destilação. A reação de policondensação estudada aqui corresponde à produção de polietileno-tereftalato (PET). Nesta reação, Tereftalato de dimetila (DMT) e Etilenoglicol (EG) geram como produtos grupos ésteres e metanol (M). O reator modelado corresponde ao reator piloto instalado no Laboratório de Simulação e Controle de Processos. A técnica de controle preditivo não linear (NMPC) foi implementada e testada em dois modelos de polimerização por radicais livres (sistema SISO e MIMO) e comparado com o desempenho de um controlador PID. O controlador preditivo não linear (NMPC) foi aplicado no modelo do reator de policondensação e foi utilizado um controlador PID como comparativo, nesta etapa realizou-se várias simulações em malha fechada e análises de sensibilidades. Para as análises de sensibilidades o controlador NMPC foi denominado de controlador NMPC com “feedback” e foi comparado com um controlador NMPC com retroalimentação de estado. Através dos estudos mostrou que o controlador NMPC tem um desempenho melhor do que o PID. Nos estudos de sensibilidade o controlador NMPC com “feedback” apresenta um desempenho superior ao do controlador NMPC com retroalimentação de estado. O controlador NMPC com “feedback” manteve o processo sob controle nas simulações que foram realizadas . / In this work the application of a non-linear predictive controller to policondensation reactors aiming the safety, quality of the product, capacity of scale-up and productivity is studied. In batch policondensation processes, a volatile product called condensate is generated, and its removal becomes necessary in order to favor the reaction progress as a mean to obtain high conversion rates. In most of the works involving policondensation the separation is not taken into account. Therefore, a policondensation model was developed which was coupled to a distillation column in order to verify the behavior of the variables in the system through open-loop simulations. The polycondensation model was obtained from the literature (Shin et al. ,1999) and was implemented taking into account the thermal inertia between the walls of the reactor and the interactions with the distillation column. The polycondensation reaction studied here corresponds to the production of polyethylene-terephtalate (PET). In this reaction, Dimethyl Terephtalate (DMT) and Etylene Glycol (EG) generate as products ester groups and methanol (M). The reactor that was modeled corresponds to a pilot reactor installed in the Laboratory of Simulation and Processes Control. A non linear predictive control (NMPC) was implemented and tested in two free radical polymerization models (SISO and MIMO systems) and compared with PID controllers performance. The NMPC was applied to the polycondensation model and a PID controller was also applied for comparison. In this stage several closed-loop simulations and sensitivity analyses were performed. For the sensitivity analyses the NMPC controller was denominated NMPC controller with "feedback" for comparison with an NMPC with full state feedback. The studies showed that NMPC controller has a better performance than PID. NMPC with "feedback" presents a superior performance to the NMPC with full state feedback in the presence of modeling gaps. In the simulations that were performed, the NMPC with "feedback" has kept the process under control.

Navios aliviadores com sistemas de posicionamento dinâmico (SPD) são dispendiosos. Dessa forma, é desejável que se otimize sua utilização, por exemplo, através da redução da quantidade de viagens entre as plataformas em alto-mar e os terminais na costa. Quando o petróleo produzido é destinado à exportação, uma solução é transferi-lo diretamente do aliviador ao navio exportador (geralmente um VLCC) através das operações conhecidas como ship-to-ship, ainda em alto-mar e próximo aos campos de produção. No entanto, uma vez que VLCCs raramente são providos de SPD, é necessário que se desenvolvam formas de se assegurar a controlabilidade de ambos os navios durante a operação de transferência. Nesse sentido, duas diferentes configurações de operações ship-to-ship são propostas. Uma delas consiste em realizar a transferência com os navios avançando lado a lado. As embarcações são amarradas e então o VLCC desenvolve potência avante, rebocando o aliviador que mantém suas máquinas ociosas. A outra solução consiste em transferir o petróleo enquanto os navios mantêm uma formação em comboio, com o VLCC seguindo uma trajetória arbitrária e sendo seguido pelo aliviador, que mantém uma distância relativa através de uma estratégia específica de piloto automático. Os modelos dinâmicos para ambas as operações são desenvolvidos e implementados em simuladores numéricos. Os resultados das simulações são discutidos e utilizados para que se avalie a viabilidade operacional das operações sob diferentes condições ambientais e de carregamento dos navios. / Shuttle tankers with dynamic positioning (DP) systems are expensive ships. Therefore, it is desirable to optimize their usage by, e.g., eliminating the travels between offshore platforms and the terminals in the coast. When the oil is intended to exportation, an attractive idea is to transfer it between the shuttle tanker and the exporter ship (usually a VLCC) in open sea, close to the oil fields. However, since a VLCC is rarely provided with a DP system, it is necessary to develop alternative ways of attaining controllability of both vessels while the transfer operation is performed. In this sense, two configurations for the so-called ship-to-ship operations are proposed. One of them consists in performing the transfer with underway ships, arranged side-by-side. The vessels are moored together and the VLCC develops power ahead, towing the idle shuttle tanker. Another alternative is to transfer the oil while the ships maintain a convoy formation, with the VLCC trailing a given trajectory and being followed by the shuttle tanker, which keeps a constant relative position by means of an appropriate autopilot strategy. Dynamic models are developed for both operations and implemented in numerical simulators. The simulations results are discussed and used to outline the operational viability under different combinations of environment and loading conditions.

Orientador: Marconi Kolm Madrid / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-19T15:15:21Z (GMT). No. of bitstreams: 1 LopesFilho_ClaudioHartkopf_M.pdf: 1796245 bytes, checksum: 6df0518799ef2b63443cdacb8d02837f (MD5) Previous issue date: 2011 / Resumo: Este trabalho desenvolve um modelo computacional para simulação do comportamento de uma célula a combustível do tipo PEM. O modelo foi desenvolvido usando equações físico-químicas e ajuste de parâmetros baseados em resultados experimentais - os quais foram tirados da literatura, pois não foram conduzidos testes práticos em laboratório - sendo capaz de determinar o valor de variáveis de saída importantes como potencia elétrica, tensão de saída e rendimento energético, assim com outros parâmetros de desempenho, como a razão de excesso de oxigênio, a partir de variáveis de entrada. O sistema completo foi dividido em diversos subsistemas, permitindo a análise dinâmica dos principais estados internos. O sistema de controle proposto e desenvolvido via inversão do modelo matemático. da célula. Dada a complexidade do modelo completo, sua divisão em diversos sub-sistemas permite a inversão de cada bloco isoladamente, de acordo com princípios da representação energética macroscópica (energetic macroscopic representation, EMR) e a construção de um controlador completo. No caso de subsistemas que possuem equacionamento não linear, o processo de inversão e obtido iterativamente via um algoritmo de busca repetitiva usando uma função Lyapunovcandidata construída a partir da equação que descreve o subsistema. Este controlador mostra excelentes resultados em simulação frente as não linearidades inerentes do sistema / Abstract: In this work we develop a computational model for the dynamical behavior simulation of a PEM fuel cell. The model was developed using physical and chemical equations and parameter fitting based on experimental data - which were obtained from literature. - being capable of predicting the value of important output variables as power, output tension and system efficiency, as well as other parameters like the oxygen excess ratio. The complete system was divided in several sub-systems, allowing dynamical analysis of the most important internal states. The proposed control system is obtained by the inversion of the fuel cell mathematical model. Given the model's complexity, each internal sub-system is inverted according the principles of energetic macroscopic representation (EMR) and all inverted controllers are integrated into the full system controller. In order to invert the mathematical model of non-linear sub-systems a interactive repetitive search algorithm is proposed, using Lyapunov-candidate functions based on the dynamic equations created for each sub-system. This control system shows good results in simulation when facing the fuel cell system non-linearity / Mestrado / Automação / Mestre em Engenharia Elétrica

Actuator saturation is ubiquitous in physical plants. In closed-loop systems limits imposed on the actuators may result in degraded performance of the control law and, ultimately, instability of the system. When other non-linearities, such as deadzone, backlash or stiction, are also present in a system’s input, the analysis and design procedures become more involved. The core of this thesis is a new structure based on the right inverse approach for deadzone and backlash, which is extended to linear plants that exhibit a combination of saturation and either deadzone, backlash or stiction, in the actuator. It is shown that, for this type of system, the inclusion of the right inverse nonlinearity results in the linear plant being subject to a new input saturation. Then, one can design standard controllers such as anti-windup or input constrained MPC around this saturation. This simplifies the analysis and design processes, in spite of the presence of complex nonlinearities. The results for deadzone and backlash are extended to stiction by proposing an approximate stiction nonlinearity, and then introducing a right inverse to this approximation. It is demonstrated that the systems studied can be compensated by a standard input constrained MPC which can be solved by a convex quadratic program. Additionally, a simple anti-windup structure is used to demonstrate the applicability of the proposed structure using existing control strategies.