Search

[EN] This work proposes a predictive control technique to be applied in fast processes using online optimization. Currently, advanced controllers are increasingly needed in industry at low levels of automation, which are associated with sampling times on the order of milliseconds or microseconds. The quadratic programming resulting from the control problem of a predictive control algorithm with constraints, such as Dynamic Matrix Control (DMC), which is one of the most used alternatives in industry, can be considered computationally expensive and becomes a limitation to embed and use the DMC in plants with fast sample rates. This paper proposes a solution to this problem based on the dual accelerated gradient projection method, which has shorter convergence times than other solutions in the literature based on predictive control strategies with online optimization. The proposed approach was tested in simulation to control a semiactive automotive suspension system, which has fast dynamics, showing that satisfactory results can be obtained with a sampling time of 5 ms. Moreover, the proposed control was implemented in a field-programmable gate array (FPGA) and the resulting quadratic programming problem was calculated in microseconds, which allows the use of the DMC to control very fast processes., [ES] Este trabajo propone una técnica de control predictivo para ser aplicada en procesos rápidos utilizando optimización en línea. Actualmente, en el sector industrial, los controladores avanzados son cada vez más necesarios en los bajos niveles de automatización, que están asociados con tiempos de muestreo del orden de milisegundos o microsegundos. La programación cuadrática resultante del problema de control de un algoritmo de control predictivo con restricciones, como por ejemplo el control por matriz dinámica (en inglés Dynamic Matrix Control - DMC), que es uno de los más usados en la industria, se puede considerar computacionalmente costosa y se convierte en una limitación para empotrar y usar el DMC en plantas con tasas de muestreo rápidas. Este artículo propone una solución a este problema basada en el método de proyección de gradiente acelerada dual, que tiene tiempos de convergencia menores que otras soluciones de la literatura basadas en estrategias de control predictivo con optimización en línea. Además, el control propuesto se implementó en una matriz de puertos programables (en inglés field-programmable gate array FPGA) y el problema de programación cuadrática resultante se calculó en microsegundos, lo que permite el uso del DMC en procesos muy rápidos., Los autores agradecen el apoyo financiero dado por CNPq – proyectos 304032/2019-0 y 315546/2021-2.

The COVID-19 pandemic is the most profound crisis of the twenty-first century. The SARS-CoV-2 virus was first registered in Brazil on March 2020, and its social and economic repercussions have been catastrophic. This paper investigates how to apply model predictive control (MPC) algorithms to plan appropriate social distancing policies that mitigate the pandemic effects. We consider MPC applications for the states of Bahia and Santa Catarina (Brazil), two regions of very different social and cultural demographics. We use Susceptible-Infected-Recovered-Deceased model to describe the pandemic dynamics in these two states, for which parameters are identified using a constrained optimization procedure. The control input to the process is a social isolation guideline passed to the population. Two MPC frameworks are developed and discussed: (a) a centralized approach, which coordinates a single predictive control policy for both states, and (b) a distributed strategy, for which a single MPC problem is solved for each state. We provide a series of simulation results in order to illustrate and compare the results obtained with both these MPC strategies. Discussions are drawn regarding the effectiveness of MPC to guide social distancing measures during pandemics and which approach (distributed, centralized) is more convenient, regarding different conditions.

Versión del autor This paper presents a new methodology to design multivariable Smith predictor for n×n processes with multiple time delays based on the centralized inverted decoupling structure. The controller elements are calculated in order to achieve good reference tracking and decoupling response. Independently of the system size, very simple general expressions for the controller elements are obtained. The realizability conditions are provided and the particular case of processes with all of its elements as first order plus time delay systems is discussed in more detail. A diagonal filter is added to the proposed control structure in order to improve the disturbance rejection without modifying the nominal set-point response and to obtain a stable output prediction in unstable plants. The effectiveness of the method is illustrated through different simulation examples in comparison with other works.

Refrigerant compressors are relatively complex products that require a wide variety of tests to be performed for both product development and quality control. Performance tests are an example of such tests and they are typically performed using test rigs with calorimeters. This paper presents the experimental results of the use of a multivariable dead-time compensating structure to control the conditions at the calorimeter inlet and outlet. Besides the control results themselves, the paper presents the identification of the multivariable model used for control purposes, as well as it discusses tuning and implementation of the proposed controller. The closed-loop experimental results are compared to the ones obtained by using PID controllers and they demonstrate the good performance of the implemented controller. As a result, both the test time and the result uncertainties have been reduced. Besides presenting an important contribution from the point of view of industrial application, this study presents experimental results of dead-time compensators, which are generally analyzed just towards simulation.