In chemical processes, online measurements of all the process variables and parameters required for process control, monitoring and optimization are seldom available. The use of soft sensors or observers is, therefore, highly significant as they can estimate unmeasured state variables from available process measurements. However, for reliable estimation by a soft sensor, the process measurements have to be placed at locations that allow reconstruction of process variables by the soft sensors. This dissertation presents a new technique for computing an optimal measurement structure for state and parameter estimation of stable nonlinear systems. The methodology can compute locations for individual sensors as well as networks of sensors where a trade-off between process information, sensor cost, and information redundancy is taken into account. The novel features of the approach are (1) that the nonlinear behavior that a process can exhibit over its operating region can be taken into account, (2) that the technique is applicable for systems described by lumped or by distributed parameter models, (3) that the technique reduces to already established methods, if the system is linear and only some of the objectives are examined, (4) that the results obtained from the procedure can be easily interpreted, and (5) that the resulting optimization problem can be decomposed, resulting in a significant reduction of the computational effort required for its solution. The other issue addressed in this dissertation is designing soft sensors for a given measurement structure. In case of high-dimensional systems, the application of conventional soft sensor or observer designs may not always be practical due to the high computational requirements or the resulting observers being too sensitive to measurement noise. To address these issues, this dissertation presents reduced-order observer design techniques for state estimation of high-dimensional chemical processes. The motivation behind these approaches is that subspaces, which are close to being unobservable, cannot be correctly reconstructed in a realistic setting due to measurement noise and inaccuracies in the model. The presented approaches make use of this observation and reconstruct the parts of the system where accurate state estimation is possible.

In chemical processes, online measurements of all the process variables and parameters required for process control, monitoring and optimization are seldom available. The use of soft sensors or observers is, therefore, highly significant as they can estimate unmeasured state variables from available process measurements. However, for reliable estimation by a soft sensor, the process measurements have to be placed at locations that allow reconstruction of process variables by the soft sensors. This dissertation presents a new technique for computing an optimal measurement structure for state and parameter estimation of stable nonlinear systems. The methodology can compute locations for individual sensors as well as networks of sensors where a trade-off between process information, sensor cost, and information redundancy is taken into account. The novel features of the approach are (1) that the nonlinear behavior that a process can exhibit over its operating region can be taken into account, (2) that the technique is applicable for systems described by lumped or by distributed parameter models, (3) that the technique reduces to already established methods, if the system is linear and only some of the objectives are examined, (4) that the results obtained from the procedure can be easily interpreted, and (5) that the resulting optimization problem can be decomposed, resulting in a significant reduction of the computational effort required for its solution. The other issue addressed in this dissertation is designing soft sensors for a given measurement structure. In case of high-dimensional systems, the application of conventional soft sensor or observer designs may not always be practical due to the high computational requirements or the resulting observers being too sensitive to measurement noise. To address these issues, this dissertation presents reduced-order observer design techniques for state estimation of high-dimensional chemical processes. The motivation behind these approaches is that subspaces, which are close to being unobservable, cannot be correctly reconstructed in a realistic setting due to measurement noise and inaccuracies in the model. The presented approaches make use of this observation and reconstruct the parts of the system where accurate state estimation is possible.

No description available.

No description available.

This thesis focuses on soft sensor development based on fuzzy logic used for real time online monitoring of anaerobic digestion to improve methane output and for robust fermentation. Important process parameter indicators such as pH, biogas production, daily difference in pH and daily difference in biogas production were used to infer alkalinity, a reliable indicator of process stability. Additionally, a fuzzy logic and a rule-based controller were developed and tested with single stage anaerobic digesters operating with cow slurry and cellulose. Alkalinity predictions from the fuzzy logic algorithm were used by both controllers to regulate the organic loading rate that aimed to optimise the biogas process. The predictive performance of a software sensor determining alkalinity that was designed using fuzzy logic and subtractive clustering and was validated against multiple linear regression models that were developed (Partner N° 2, Rothamsted Research 2010) for the same purpose. More accurate alkalinity predictions were achieved by utilizing a fuzzy software sensor designed with less amount of data compared to a multiple linear regression model whose design was based on a larger database. Those models were utilised to control the organic loading rate of a twostage, semi-continuously fed stirred reactor system. Three 5l reactors without support media and three 5l reactors with different support media (burst cell reticulated polyurethane foam coarse, burst cell reticulated polyurethane foam medium and sponge) were operated with cow slurry for a period of seven weeks and twenty weeks respectively. Reactors with support media were proven to be more stable than the reactors without support media but did not exhibit higher gas productivity. Biomass support media were found to influence digester recovery positively by reducing the recovery period. Optimum process parameter ranges were identified for reactors with and without support media. Increased biogas production was found to occur when the loading rates were 3-3.5g VS/l/d and 4-5g VS/l/d respectively. Optimum pH ranges were identified between 7.1-7.3 and 6.9-7.2 for reactors with and without support media respectively, whereas all reactors became unstable at ph<6.9. Alkalinity levels for system stability appeared to be above 3500 mg/l of HCO3 - for reactors without media and 3480 mg/l of HCO3 - for reactors with support media. Biogas production was maximized when alkalinity was 3 between 3500-4500 mg/l of HCO3 - for reactors without support media and 3480- 4300 mg/l of HCO3 - for reactors with support media. Two fuzzy logic models predicting alkalinity based on the operation of the three 5l reactors with support media were developed (FIS I, FIS II). The FIS II design was based on a larger database than FIS I. FIS II performance when applied to the reactor where sponge was used as the support media was characterized by quite good MAE and bias values of 466.53 mg/l of HCO3- and an acceptable value for R2= 0.498. The NMSE was close to 0 with a value of 0.03 and a slightly higher FB= 0.154 than desired. The fuzzy system robustness was tested by adding NaHCO3 to the reactor with the burst cell reticulated polyurethane foam medium and by diluting the reactor where sponge was used as the support media with water. FIS I and FIS II were able to follow the system output closely in the first case, but not in the second. FIS II functionality as an alkalinity predictor was tested through the application on a 28l cylindrical reactor with sponge as the biomass support media treating cow manure. If data that was recorded when severe temperature fluctuations occurred (that highly impact digester performance), are excluded, FIS II performance can be characterized as good by having R2= 0.54 and MAE=Bias= 587 mg/l of HCO3-. Predicted alkalinity values followed observed alkalinity values closely during the days that followed NaHCO3 addition and water dilution. In a second experiment a rulebased and a Mamdani fuzzy logic controller were developed to regulate the organic loading rate based on alkalinity predictions from FIS II. They were tested through the operation of five 6.5l reactors with biomass support media treating cellulose. The performance indices of MAE=763.57 mg/l of HCO3-, Bias= 398.39 mg/l of HCO3-, R2= 0.38 and IA= 0.73 indicate a pretty good correlation between predicted and observed values. However, although both controllers managed to keep alkalinity within the desired levels suggested for stability (>3480 mg/l of HCO3-), the reactors did not reach a stable state suggesting that different loading rates should be applied for biogas systems treating cellulose.

No description available.

Sensores virtuais empregando modelos de inferência da composição(responsável pela qualidade) dos produtos de uma coluna de destilação correspondem a medidores implementados em software, capazes de estimar, em tempo real, a composição dos produtos da mesma, a partir de informações do tipo temperaturas e pressões em diversos pontos da coluna e vazões de entrada, de saída e de reciclo. O objetivo deste trabalho é obter esse tipo de sensor para uma coluna de destilação, capaz de estimar instantaneamente a composição dos produtos no topo de uma coluna de destilação multicomponente com condensador parcial, empregando redes neurais artificiais. Foi desenvolvido um simulador dinâmico baseado em modelo não-linear da coluna para aquisição de dados. Neste projeto foi incluído um estudo sobre a influência do treinamento parcial no desempenho do sensor virtual. A idéia é estudar o desempenho para o caso de um sensor virtual treinado de antemão, com dados coletados a partir de um simulador da coluna. Este procedimento disponibiliza um sensor operacional, treinado através de um conjunto de dados simulados ou através de um pequeno conjunto de pontos e retreinado, quando dados reais ou um conjunto maior de dados estiver disponível. Outra contribuição importante é o estudo realizado sobre os principais erros que podem ocorrer neste tipo de sensores, que são raramente tratados em publicações científicas. É também proposta uma metodologia para detecção e correção destes erros que foram encontrados e que afetam o comportamento do sensor, alterando sua precisão e capacidade de ser utilizado em um controle inferencial da planta. / Soft sensors for composition inference models (that are responsible for the quality) of distillation column products, correspond to virtual instruments implemented in software. This software is able to estimate, in real time, the composition of the output products of the column, based on information such as temperature and pressure on several points of the column and on input, output and recycle flow. The purpose of this work is to obtain a soft sensor that estimates the instantaneous composition of the product at the top of a multicomponent distillation column with a partial condenser, employing artificial neural networks. The chosen architecture was the feedforward neural network with three layers. It was chosen based on many tested options. It was developed a dynamical simulator of this column for data acquisition based on a non-linear model. In this study, it was included an investigation about the influence of partial training in the performance of the soft sensor. The goal is to study the results achieved in the case of a soft sensor trained beforehand, with data acquired from the simulator of this column. This procedure makes possible to have an operational soft sensor, trained based on a simulated data set or on a small amount of points and then retrained when a real or larger data set is available. Another important contribution is the study performed about the main errors that may appear in this kind of sensor. These errors are rarely mentioned in scientific papers. It also aims at implementing techniques to enable detection and correction of those errors that the soft sensor may present, and that affect the performance of the soft sensor, changing its precision and making it inadequate for inferential control.

Sensores virtuais empregando modelos de inferência da composição(responsável pela qualidade) dos produtos de uma coluna de destilação correspondem a medidores implementados em software, capazes de estimar, em tempo real, a composição dos produtos da mesma, a partir de informações do tipo temperaturas e pressões em diversos pontos da coluna e vazões de entrada, de saída e de reciclo. O objetivo deste trabalho é obter esse tipo de sensor para uma coluna de destilação, capaz de estimar instantaneamente a composição dos produtos no topo de uma coluna de destilação multicomponente com condensador parcial, empregando redes neurais artificiais. Foi desenvolvido um simulador dinâmico baseado em modelo não-linear da coluna para aquisição de dados. Neste projeto foi incluído um estudo sobre a influência do treinamento parcial no desempenho do sensor virtual. A idéia é estudar o desempenho para o caso de um sensor virtual treinado de antemão, com dados coletados a partir de um simulador da coluna. Este procedimento disponibiliza um sensor operacional, treinado através de um conjunto de dados simulados ou através de um pequeno conjunto de pontos e retreinado, quando dados reais ou um conjunto maior de dados estiver disponível. Outra contribuição importante é o estudo realizado sobre os principais erros que podem ocorrer neste tipo de sensores, que são raramente tratados em publicações científicas. É também proposta uma metodologia para detecção e correção destes erros que foram encontrados e que afetam o comportamento do sensor, alterando sua precisão e capacidade de ser utilizado em um controle inferencial da planta. / Soft sensors for composition inference models (that are responsible for the quality) of distillation column products, correspond to virtual instruments implemented in software. This software is able to estimate, in real time, the composition of the output products of the column, based on information such as temperature and pressure on several points of the column and on input, output and recycle flow. The purpose of this work is to obtain a soft sensor that estimates the instantaneous composition of the product at the top of a multicomponent distillation column with a partial condenser, employing artificial neural networks. The chosen architecture was the feedforward neural network with three layers. It was chosen based on many tested options. It was developed a dynamical simulator of this column for data acquisition based on a non-linear model. In this study, it was included an investigation about the influence of partial training in the performance of the soft sensor. The goal is to study the results achieved in the case of a soft sensor trained beforehand, with data acquired from the simulator of this column. This procedure makes possible to have an operational soft sensor, trained based on a simulated data set or on a small amount of points and then retrained when a real or larger data set is available. Another important contribution is the study performed about the main errors that may appear in this kind of sensor. These errors are rarely mentioned in scientific papers. It also aims at implementing techniques to enable detection and correction of those errors that the soft sensor may present, and that affect the performance of the soft sensor, changing its precision and making it inadequate for inferential control.

The recovery and recycling of the spent chemicals from the kraft pulping process are economically and environmentally essential in an integrated kraft pulp and paper mill. The recovery process can be optimised by firing high-solids black liquor in the recovery boiler. Unfortunately, due to a corresponding increase in the liquor viscosity, in many mills, black liquor is fired at reduced solids concentration to avoid possible rheological problems. Online measurement, monitoring and control of the liquor viscosity are deemed essential for the recovery boiler optimization. However, in most mills, including those in New Zealand, black liquor viscosity is not routinely measured. Four batches of black liquors having solids concentrations ranging between 47 % and 70 % and different residual alkali (RA) contents were obtained from Carter Holt Harvey Pulp and Paper (CHHP&P), Kinleith mill, New Zealand. Weak black liquor samples were obtained by diluting the concentrated samples with deionised water. The viscosities of the samples at solids concentrations ranging from 0 to 70 % were measured using open-cup rotational viscometers at temperatures ranging from 0 to 115 oC and shear rates between 10 and 2000 s-1. The effect of post-pulping process, liquor heat treatment (LHT) on the liquors’ viscosities was investigated in an autoclave at a temperature >=180 oC for at least 15 mins. The samples exhibit both Newtonian and non-Newtonian behaviours depending on temperature and solids concentration; the onsets of these behaviours are liquor-dependent. In conformity with the literature data, at high solids concentrations (> 50 %) and low temperatures, they exhibit shear-thinning behaviour with or without thixotropy but the shear-thinning/thixotropic characteristics disappear at high temperatures (>= 80 oC). Generally, when the apparent viscosities of the liquors are <= ~1000 cP, the liquors show a Newtonian or a near-Newtonian behaviour. These findings demonstrate that New Zealand black liquors can be safely treated as Newtonian fluids under industrial conditions. Further observations show that at low solids concentrations (< 50 %), viscosity is fairly independent of the RA content; however at solids concentrations > 50 %, viscosity decreases with increasing RA content of the liquor. This shows that the RA content of black liquor can be manipulated to control the viscosity of high-solids black liquors. The LHT process had negligible effect on the low-solids liquor viscosity but led to a significant and permanent reduction of the high-solids liquor viscosity by a factor of at least 6. Therefore, the incorporation of a LHT process into an existing kraft recovery process can help to obtain the benefits of high-solids liquor firing without a concern for the attending rheological problems. A variety of the existing and proposed viscosity models using the traditional regression modelling tools and an artificial neural network (ANN) paradigm were obtained under different constraints. Hitherto, the existing models rely on the traditional regression tools and they were mostly applicable to limited ranges of process conditions. On the one hand, composition-dependent models were obtained as a direct function of solids concentration and temperature, or solids concentration, temperature and shear rate; the relationships between these variables and the liquor viscosity are straight forward. The ANN-based models developed in this work were found to be superior to the traditional models in terms of accuracy, generalization capability and their applicability to a wide range of process conditions. If the parameters of the resulting ANN models can be successfully correlated with the liquor composition, the models would be suitable for online application. Unfortunately, black liquor viscosity depends on its composition in a complex manner; the direct correlation of its model parameters with the liquor composition is not yet a straight forward issue. On the other hand, for the first time in the Australasia, the limitations of the composition-dependent models were addressed using centrifugal pump performance parameters, which are easy to measure online. A variety of centrifugal pump-based models were developed based on the estimated data obtained via the Hydraulic Institute viscosity correction method. This is opposed to the traditional approaches, which depend largely on actual experimental data that could be difficult and expensive to obtain. The resulting age-independent centrifugal pump-based model was implemented online as a black liquor viscosity soft sensor at the number 5 recovery boiler at the CHHP&P, Kinleith mill, New Zealand where its performance was evaluated. The results confirm its ability to effectively account for variations in the liquor composition. Furthermore, it was able to give robust viscosity estimates in the presence of the changing pump’s operating point. Therefore, it is concluded that this study opens a new and an effective way for kraft black liquor viscosity sensor development.

Submitted by Automação e Estatística (sst@bczm.ufrn.br) on 2018-07-03T14:17:15Z No. of bitstreams: 1 JeanMarioMoreiraDeLima_DISSERT.pdf: 2889834 bytes, checksum: f4a43ed1a203489a4ac1293aa8e04bf0 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2018-07-11T11:19:44Z (GMT) No. of bitstreams: 1 JeanMarioMoreiraDeLima_DISSERT.pdf: 2889834 bytes, checksum: f4a43ed1a203489a4ac1293aa8e04bf0 (MD5) / Made available in DSpace on 2018-07-11T11:19:44Z (GMT). No. of bitstreams: 1 JeanMarioMoreiraDeLima_DISSERT.pdf: 2889834 bytes, checksum: f4a43ed1a203489a4ac1293aa8e04bf0 (MD5) Previous issue date: 2018-05-28 / Atualmente, diante de um mercado cada vez mais competitivo, produzir de forma eficiente é essencial para se obter um balanço econômico positivo. Reduzir custos, realizar processos otimizados e ofertar produtos cada vez melhores são fatores que influenciam diretamente na economia de qualquer indústria. Diante disso, técnicas que podem melhorar e/ou otimizar processos, como o monitoramento da qualidade de produto ou controle avançado tornam-se fundamentais para a indústria como um todo. No caso de Unidades de Processamento de Gás Natural (UPGNs), o monitoramento da qualidade do produto produzido é intrínseco a uma produção satisfatória, e esse controle da qualidade faz-se, como na maioria dos processos químicos, através da composição química dos produtos. Entretanto, mesmo quando cromatógrafos a gás são utilizados para análise química dos componentes, tem-se lentidão no processo analítico e longos intervalos de medição são observados. Isso impede que técnicas de monitoramento em tempo real, ou de controle, sejam estabelecidas para obtenção de melhor rendimento do processo. Dentre esses produtos, em termos econômicos, o principal é o GLP (Gás Liquefeito de Petróleo), composto por propano, butano e contaminantes como etano e pentano. Neste trabalho é proposto um sistema chamado de soft sensor que faça a inferência, baseada em redes neurais, da composição do GLP, isto é, dos seus principais componentes. Dessa forma, o monitoramento, em tempo real, da qualidade do GLP produzido torna-se possível, uma vez que a medição de sua composição não será feita através do lento processo analítico. Assim, melhora-se a qualidade do processo, do produto e, consequentemente, a lucratividade. Para o desenvolvimento deste trabalho, utilizou-se uma UPGN simulada no software HYSYS, formada por uma coluna deetanizadora em série com uma coluna debutanizadora. Na instrumentação da planta, têm-se alguns controladores PIDs. O sensor virtual tem como entradas algumas das variáveis de processo desses controladores. Neste tabalho, é proposto também um sistema de correção do erro, em tempo real, do soft sensor, tendo com base a leitura da composição do GLP feita por cromatógrafos presentes no processo. Os resultados se mostraram promissores, atestando o funcionamento adequado do soft sensor. / In face of an increasingly competitive market, producing efficiently and effectively is essential for a positive economic balance. Reducing costs, optimizing processes and offering even better products are factors that directly influence the economy of any industry. In this view, techniques that can improve and / or guarantee optimization of processes, such as the monitoring of product quality or advanced and intelligent control become fundamental for the industry as a whole. In case of Natural Gas Processing Units (NGPUs), monitoring the quality of the product is intrinsic to a satisfactory production, and quality control has been done such as in most chemical processes, through the chemical composition of the products. However, even when chromatographs are used for chemical analysis of the components, the analytical process is slow and long measurement intervals are observed. This hampers real-time product monitoring or control techniques from being established to obtain better process performance.Among these products, the most important, economically speaking, is LPG (Liquefied Petroleum Gas) composed of propane, butane and contaminats such as ethane and pentane. In this work, a system called soft sensor that makes the inference of the main components of GLP based on artificial neural network is proposed. Then, the real-time monitoring of the quality of the produced LPG becomes possible, since the measurement of the composition of the LPG will not be obtained through the slow analytical process. Thus, the quality of the process, the LPG itself, consequently, its profitability are improved.In the development of this work, a simulated GNPU has been used in HYSYS, consisting of a deethanizing column in series with a debutanizer column. In the instrumentation of the plant, there are some PID controllers. The virtual sensor is based on process variables of these controllers. In this work, a real-time error correction module of the softsensor is also proposed, based on the measuruments of the LPG composition made by the chromatographs present in the process. The results are promising, attesting the adequate behavior of the soft sensor.