... OPERATIONAL PLANNING IN THE MANAGEMENT OF PROGRAMMED MAINTENANCES A MILP APPROACH Flavio Mane... more ... OPERATIONAL PLANNING IN THE MANAGEMENT OF PROGRAMMED MAINTENANCES A MILP APPROACH Flavio Manenti* and Maurizio Rovaglio CMIC dept. Giulio Natta, Politecnico di Milano Piazza Leonardo da Vinci, I-20133, MILANO ...
Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fi... more Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fine chemical industry is to produce a high quality product and operational optimization is the key-element to match it. This work investigates how batch and semibatch reactors can be optimized in order to increase the yield of a desired product. Optimization problem is formulated and applied to calculate the optimal operating parameters such as the reactor temperature and the feed flow rate. Comparison and considerations on the two reactor configurations are given.
This paper gives an overview of the numerical methods included in BzzMath library. The library is... more This paper gives an overview of the numerical methods included in BzzMath library. The library is written in C++, exploits the features of object-oriented programming, and is free for non-commercial applications at www.chem.polimi.it/homes/gbuzzi.
This paper introduces a freeware program for models analysis, which implements a robust method fo... more This paper introduces a freeware program for models analysis, which implements a robust method for parameter estimation of nonlinear models, able to detect outliers. A new class of criteria is proposed in order to achieve an optimal experimental design. Finally a strategy for model discrimination is presented. Since this is a critical point, the correct meaning of statistical tests and
ABSTRACT The paper illustrates the redefinition of the concept of biorefinery and its application... more ABSTRACT The paper illustrates the redefinition of the concept of biorefinery and its application onto the biogas field. The proposed integrated framework merges in a synergistic way the second and third generation biorefineries with the aim of reducing the CO2 emissions and maximizing the utilization of biomass potential. The specific focus of the paper is on the mathematical modelling of the main conversion steps of the lignocellulosic residuals into biosyngas by means of thermal treatment (pyrolysis, combustion, gasification) and, next, into biomethanol (heterogeneously catalysed synthesis). The models are implemented in user-friendly programs and adopted to estimate biomass conversion and process yield. A final study of sensitivity analysis is provided to open the way for the next process optimization based on detail models.
Modeling and simulation of molecular distillation to split of heavy petroleum cuts are proposed. ... more Modeling and simulation of molecular distillation to split of heavy petroleum cuts are proposed. A theoretical study of this problem is illustrated for a heavy petroleum cut divided into six pseudo-components. In the process simulation, a falling film is considered. The mathematical model comprises equations for the evaluation of the physicochemical properties, in order to characterize the distilled mixture. Heat and material balances on the liquid film are numerically solved by using a central finitedifference method. The simulation is carried out at the steady-state conditions, where the relevant process variables, such as film thickness, evaporation rate, film surface temperature, concentration profiles, and amount of distillate flow rate are computed. In molecular distillation, the concentration of the most volatile components shrinks in both axial and radial directions, especially due to the fast increase of the temperature in the falling film. As a consequence, the less volatile components of the liquid mixture undergoes a fast increase on the liquid interface, by clearly showing the potentialities of the molecular distillation process in separating heavy petroleum cuts. As the results show, the inlet variables such as the feed temperature and the feed flow rate largely influence the final composition of the condensate flow.
A generalized C++ class to solve nonlinear model predictive control (MPC) and dynamic optimizatio... more A generalized C++ class to solve nonlinear model predictive control (MPC) and dynamic optimization problems is proposed. Since optimal control problems involve (i) differential equations systems to foresee plants and/or process units dynamics and (ii) constrained optimization issues to meet process specs and requirements, BzzMath library is adopted as kernel to numerically face these tasks. The proposed class allows both FORTRAN and C++ users to easily solve MPC and dynamic optimization by defining their differential system and the desired objective function only, without taking care of any numerical problem that may occur in integrating differential systems, in searching for the minimum of a constrained/complex objective function, and in implementing a moving horizon methodology.
Increasing worldwide market competitiveness and reduced profit margins are pressing chemical and ... more Increasing worldwide market competitiveness and reduced profit margins are pressing chemical and process industries to move towards a predictive control approach, based on first-principles mathematical models, as well as plant dynamic optimization. In this perspective, the paper focuses on the development of a nonlinear model predictive control (NMPC) to manage the copolymerization process of methyl methacrylate (MMA) with vinyl acetate (VAc), consisting of a jacketed continuous stirred tank reactor, a separator, and a recycle loop. This system presents a highly complex behavior, thus making difficult the success of controllers based on linear models. A detailed differential and algebraic mathematical model consists of 53 equations and is implemented in Fortran 90/95 to simulate the plant and setup the NMPC. The numerical solution is performed by using IMSL library. NMPC is proved to be superior to a linear model predictive control approach and appears to hold a considerable promise for such a reactor system.
A new strategy to improve the parameters estimation is formulated analytically and validate on a ... more A new strategy to improve the parameters estimation is formulated analytically and validate on a specific test case. Benefits coming from the novel strategy are then discussed and compared to the existing approaches. The proposed strategy deals with the problem of the optimal design of experiments for nonlinear models and it is based on the detection of possible correlations among the new planned experimental points and the existing data points on the overall experimental domain.
This research activity is mainly aimed at showing potentialities in coupling objectoriented progr... more This research activity is mainly aimed at showing potentialities in coupling objectoriented programming with parallel computing. Wide margins of benefits could be obtained in algorithm efficiency and robustness with a relative small programming effort. The case of unconstrained multi-dimensional optimization is proposed as quantitative example.
ABSTRACT The chemical engineer has nowadays a wide choice of tools, numerical libraries, and prog... more ABSTRACT The chemical engineer has nowadays a wide choice of tools, numerical libraries, and programming languages to perform computations. Actually, it is possible to use several well established commercial packages, implement dedicated solvers into specific programming languages, or use existing numerical libraries. Also, it is possible to combine these possibilities to get either superior performances or more robustness, according to the problem features through the so-called mixed-language approach, which is increasingly spreading in the scientific communities. Since there is no full clarity on their benefits in handling numerical problems and their performances have not been yet compared in the literature, this paper is aimed at analyzing efficiency and robustness of some of the most widespread methodologies adopted for numerical computations: the conventional methods, the implementation of numerical libraries, the mixed-language, and the commercial tools. Specifically, the common case of differential systems is selected as comparison field.
ABSTRACT If there exists a gradient pointing outwards a save-operation region, then there exists ... more ABSTRACT If there exists a gradient pointing outwards a save-operation region, then there exists a possible path for the plant to cross out of the save-operation region. A method is introduced that allows the identification of such surface pieces in the phase space of the state variable for the analysed equipment. The idea is based on splitting the phase space into subspaces separated by the zero-dynamic component surface. The computation requires only root solving of the right-hand side of the dynamic equations, one at the time.
Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fi... more Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fine chemical industry is to produce a high quality product and operational optimization is the key-element to match it. This work investigates how batch and semibatch reactors can be optimized in order to increase the yield of a desired product. Optimization problem is formulated and applied to calculate the optimal operating parameters such as the reactor temperature and the feed flow rate. Comparison and considerations on the two reactor configurations are given.
... OPERATIONAL PLANNING IN THE MANAGEMENT OF PROGRAMMED MAINTENANCES A MILP APPROACH Flavio Mane... more ... OPERATIONAL PLANNING IN THE MANAGEMENT OF PROGRAMMED MAINTENANCES A MILP APPROACH Flavio Manenti* and Maurizio Rovaglio CMIC dept. Giulio Natta, Politecnico di Milano Piazza Leonardo da Vinci, I-20133, MILANO ...
Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fi... more Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fine chemical industry is to produce a high quality product and operational optimization is the key-element to match it. This work investigates how batch and semibatch reactors can be optimized in order to increase the yield of a desired product. Optimization problem is formulated and applied to calculate the optimal operating parameters such as the reactor temperature and the feed flow rate. Comparison and considerations on the two reactor configurations are given.
This paper gives an overview of the numerical methods included in BzzMath library. The library is... more This paper gives an overview of the numerical methods included in BzzMath library. The library is written in C++, exploits the features of object-oriented programming, and is free for non-commercial applications at www.chem.polimi.it/homes/gbuzzi.
This paper introduces a freeware program for models analysis, which implements a robust method fo... more This paper introduces a freeware program for models analysis, which implements a robust method for parameter estimation of nonlinear models, able to detect outliers. A new class of criteria is proposed in order to achieve an optimal experimental design. Finally a strategy for model discrimination is presented. Since this is a critical point, the correct meaning of statistical tests and
ABSTRACT The paper illustrates the redefinition of the concept of biorefinery and its application... more ABSTRACT The paper illustrates the redefinition of the concept of biorefinery and its application onto the biogas field. The proposed integrated framework merges in a synergistic way the second and third generation biorefineries with the aim of reducing the CO2 emissions and maximizing the utilization of biomass potential. The specific focus of the paper is on the mathematical modelling of the main conversion steps of the lignocellulosic residuals into biosyngas by means of thermal treatment (pyrolysis, combustion, gasification) and, next, into biomethanol (heterogeneously catalysed synthesis). The models are implemented in user-friendly programs and adopted to estimate biomass conversion and process yield. A final study of sensitivity analysis is provided to open the way for the next process optimization based on detail models.
Modeling and simulation of molecular distillation to split of heavy petroleum cuts are proposed. ... more Modeling and simulation of molecular distillation to split of heavy petroleum cuts are proposed. A theoretical study of this problem is illustrated for a heavy petroleum cut divided into six pseudo-components. In the process simulation, a falling film is considered. The mathematical model comprises equations for the evaluation of the physicochemical properties, in order to characterize the distilled mixture. Heat and material balances on the liquid film are numerically solved by using a central finitedifference method. The simulation is carried out at the steady-state conditions, where the relevant process variables, such as film thickness, evaporation rate, film surface temperature, concentration profiles, and amount of distillate flow rate are computed. In molecular distillation, the concentration of the most volatile components shrinks in both axial and radial directions, especially due to the fast increase of the temperature in the falling film. As a consequence, the less volatile components of the liquid mixture undergoes a fast increase on the liquid interface, by clearly showing the potentialities of the molecular distillation process in separating heavy petroleum cuts. As the results show, the inlet variables such as the feed temperature and the feed flow rate largely influence the final composition of the condensate flow.
A generalized C++ class to solve nonlinear model predictive control (MPC) and dynamic optimizatio... more A generalized C++ class to solve nonlinear model predictive control (MPC) and dynamic optimization problems is proposed. Since optimal control problems involve (i) differential equations systems to foresee plants and/or process units dynamics and (ii) constrained optimization issues to meet process specs and requirements, BzzMath library is adopted as kernel to numerically face these tasks. The proposed class allows both FORTRAN and C++ users to easily solve MPC and dynamic optimization by defining their differential system and the desired objective function only, without taking care of any numerical problem that may occur in integrating differential systems, in searching for the minimum of a constrained/complex objective function, and in implementing a moving horizon methodology.
Increasing worldwide market competitiveness and reduced profit margins are pressing chemical and ... more Increasing worldwide market competitiveness and reduced profit margins are pressing chemical and process industries to move towards a predictive control approach, based on first-principles mathematical models, as well as plant dynamic optimization. In this perspective, the paper focuses on the development of a nonlinear model predictive control (NMPC) to manage the copolymerization process of methyl methacrylate (MMA) with vinyl acetate (VAc), consisting of a jacketed continuous stirred tank reactor, a separator, and a recycle loop. This system presents a highly complex behavior, thus making difficult the success of controllers based on linear models. A detailed differential and algebraic mathematical model consists of 53 equations and is implemented in Fortran 90/95 to simulate the plant and setup the NMPC. The numerical solution is performed by using IMSL library. NMPC is proved to be superior to a linear model predictive control approach and appears to hold a considerable promise for such a reactor system.
A new strategy to improve the parameters estimation is formulated analytically and validate on a ... more A new strategy to improve the parameters estimation is formulated analytically and validate on a specific test case. Benefits coming from the novel strategy are then discussed and compared to the existing approaches. The proposed strategy deals with the problem of the optimal design of experiments for nonlinear models and it is based on the detection of possible correlations among the new planned experimental points and the existing data points on the overall experimental domain.
This research activity is mainly aimed at showing potentialities in coupling objectoriented progr... more This research activity is mainly aimed at showing potentialities in coupling objectoriented programming with parallel computing. Wide margins of benefits could be obtained in algorithm efficiency and robustness with a relative small programming effort. The case of unconstrained multi-dimensional optimization is proposed as quantitative example.
ABSTRACT The chemical engineer has nowadays a wide choice of tools, numerical libraries, and prog... more ABSTRACT The chemical engineer has nowadays a wide choice of tools, numerical libraries, and programming languages to perform computations. Actually, it is possible to use several well established commercial packages, implement dedicated solvers into specific programming languages, or use existing numerical libraries. Also, it is possible to combine these possibilities to get either superior performances or more robustness, according to the problem features through the so-called mixed-language approach, which is increasingly spreading in the scientific communities. Since there is no full clarity on their benefits in handling numerical problems and their performances have not been yet compared in the literature, this paper is aimed at analyzing efficiency and robustness of some of the most widespread methodologies adopted for numerical computations: the conventional methods, the implementation of numerical libraries, the mixed-language, and the commercial tools. Specifically, the common case of differential systems is selected as comparison field.
ABSTRACT If there exists a gradient pointing outwards a save-operation region, then there exists ... more ABSTRACT If there exists a gradient pointing outwards a save-operation region, then there exists a possible path for the plant to cross out of the save-operation region. A method is introduced that allows the identification of such surface pieces in the phase space of the state variable for the analysed equipment. The idea is based on splitting the phase space into subspaces separated by the zero-dynamic component surface. The computation requires only root solving of the right-hand side of the dynamic equations, one at the time.
Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fi... more Batch and semi-batch reactors are widely used for fine chemical productions. The target in the fine chemical industry is to produce a high quality product and operational optimization is the key-element to match it. This work investigates how batch and semibatch reactors can be optimized in order to increase the yield of a desired product. Optimization problem is formulated and applied to calculate the optimal operating parameters such as the reactor temperature and the feed flow rate. Comparison and considerations on the two reactor configurations are given.
Uploads
Papers by Flavio Manenti