Showing papers in "Revista Iberoamericana De Automatica E Informatica Industrial in 2013"

Algoritmos Evolutivos y su empleo en el ajuste de controladores del tipo PID: Estado Actual y Perspectivas

Abstract: PID controllers are a reliable, robust, practical and easy to implement control solution for industrial processes. They provide the first control layer for a vast majority of industrial applications. Owing to this, several researches invest time and resources to improve their performance. The research lines in this field scope with new tuning methods, new types of structures and integral design methods. For tuning methods, improvements could be fulfilled stating an optimization problem, which could be non-linear, non-convex and highly constrained. In such instances, evolutionary algorithms have shown a good performance and have been used in various proposals related with PID controllers tuning. This work shows a review of these proposals and the benefits obtained in each case. Some trends and possible future research lines are also identified.

Control Multivariable por Desacoplo

Abstract: The interaction between variables is inherent in multivariable processes and this fact may complicate their operation and control system design. Under the paradigm of decoupling control, several methodologies that traditionally have been addressed to cancel or reduce the interactions are gathered. Recently, this approach has been reoriented by several researchers with the aim to solve such a complex problem as the multivariable control. Parts of the material in this work are well known in the process control field; however, most of them are results obtained by the authors after several years of research giving priority to the problem generalization and practical issues like easiness of implementation and utilization of PID controllers as elementary blocks. This combination of interests makes difficult to obtain perfect decoupling in all cases; although it is possible to achieve an important interaction reduction at the basic level of the control pyramid in such a way that other control systems at higher hierarchical levels benefit of this fact. This article summarizes the main aspects of decoupling control and presents its application to two illustrative examples: an experimental quadruple tank process and a 4×4 model of a heat, ventilation and air conditioning system.

Los Sistemas de Suspensión Activa y Semiactiva: Una Revisión

Abstract: This paper reviews the state of the art in modeling and control of active and semi-active suspension systems. Distinctive characteristics are established for the major types of suspension systems: passive, active, and semi-active. Regarding modeling and simulation, different approaches, tools and applications are discussed in the context of vehicle dynamics. Besides, the quarter car model, which is widely used in research, is developed with differential equations, transfer functions, and state-space equations, as well as solutions for simulation in Simulink and SimMechanics. As for control of active and semi-active systems, the major strategies for vehicle suspension are reviewed. Furthermore, the paper outlines suspension control in other engineering applications.

Emisión Acústica y Redes Neuronales para Modelado y Caracterización del Proceso de Soldadura por Fricción Agitación

Abstract: This paper presents an analysis of the correlation between acoustic emission (AE) signals and the main parameters of friction stir welding (FSW) process, based on artificial neural networks (ANN). The AE signals have been acquired by the data acquisition instrument NI USB-9234, applied during the welding process carried out on plates of 3mm thick of aluminium AA1050 alloy. Statistical and temporal parameters of discomposed EA signals using Wavelet Transform (WT) have been used as input of the ANN, while the outputs of model include the welding parameters: tool rotation speed and travel speed, as well as the tool profile. A multilayer feed-forward ANN has been selected and trained, using different algorithms and network architectures. The parameters provided by the ANN constitute the model and the characterization of the FSW process; finally an analysis of the comparison between the measured and the calculated data is presented, validating the results. The model obtained can be used to develop the automatic control of the parameters of the FSW process, based on vibro-acoustic signals, which constitutes the following step in this research line.

Identificación y Control Wavenet de un Motor de CA

Abstract: This paper presents a control scheme to tune online the proportional, integral and derivative gains of a discrete PID controller, through the identification and control of a SISO stable and minimum phase dynamic system. This is accomplished using a radial basis network neural with daughter Morlet wavelets activation functions in cascaded with an infinite impulse response (IIR) filter. This scheme is applied in real time to control the speed of an AC three-phase induction motor supplied with a three-phase inverter. So in this way we show how the identification and control scheme can be implemented in this type of plants that are widely used in industry, without the need of mathematical model parameters of the induction motor. We present numerical simulation and experimental results.

Estudio comparativo de técnicas de reducción de ruido en señales industriales mediante Transformada Wavelet Discreta y selección adaptativa del umbral

Abstract: Noise reduction techniques are widely used for audio recording, image editing, and industrial signal processing. The idea is to reconstruct the original data from the noise-corrupted signal suppressing, all or almost all, the distortion caused by the inherent noise of the physical processes. In the present paper, we perform a comparative review of several noise reduction techniques based on adaptive threshold selection. These techniques have been extensively used for image processing. However, we aimed at evaluating their performance for industrial signal noise reduction. In particular, we analyze the behaviour of the Bayes Shrink, Normal Shrink, Modified Shrink, and Neight Shrink methods for the reduction of the Gaussian noise in industrial signals. To that aim, we perform experiments on a set of pattern signals proposed by Donoho and other representative measurements obtained from real processes in Cuban1s Nickel plants. Our results indicate that, for this kind of data, the Neigh Shrink algorithm outperforms.

Identificación y control en cascada mediante inversión de no linealidades del cuatrirrotor para el Concurso de Ingeniería de Control CEA IFAC 2012

Abstract: This paper describes and analyzes the processes carried out to obtain the model identification and the controller design, as well as its posterior implementation, for the quadrotor Parrot AR- Drone. The results are intended to control the autonomous tracking of that quadrotor on a path in the XY plane, flexible defined by the user. The identification phase is characterized by the division of the model into a linear and a non-linear part, resulting in a Hammerstein's model. The control is characterized by the use of a cascade control with an inner speed control loop and an outer position control loop. As characteristic features, it is highlighted that: the speed controller includes a static linearization of the model, and the position loop includes a nonlinear block to reduce the tracking error of the trajectory but using a simple position controller logic. The context of this study is the Control Engineering Competition 2012, “Autonomous trajectory tracking control of a quadrotor vehicle” organized by the Spanish Committee for Automation, which took place at the Automation Symposium in Vigo, last September 2012. The work presented was awarded in the first and second phase of the competition with the first and third prize, respectively.

Identificación de sistemas en lazo cerrado basada en una estrategia híbrida AGA-Simplex

Abstract: Closed-loop identification of continuous systems, which can be considered as a nonlinear optimization problem, may result in a difficult solution problem when conventional methods are used. In this paper it is presented a hybrid strategy based on an Adaptive Genetic Algorithm and the Simplex method, that results in a satisfactory solution for this problem. The proposal is compared with other techniques reported in the literature. Three examples show the performance of the method: identification of high order dynamics; identification of unstable second order dynamics in open-loop; and parameter estimation in power generation systems. Simulation results show that the proposed is a robust method for close-loop system identification.

Control predictivo para seguimiento de sistemas no lineales. Aplicación a una planta piloto

Abstract: This paper deals with the tracking problem for constrained nonlinear systems using a model predictive control (MPC) law. MPC provides a control law suitable for regulating constrained linear and nonlinear systems to a given target steady state. However, when the target operating point changes, the feasibility of the controller may be lost and the controller fails to track the reference. Recently, a novel MPC for tracking constrained nonlinear systems has been presented (Ferramosca et al., 2009a). This is capable to steer the system to any reference, even in the case of changing references. In this paper, this controller is used for the real-time control of a chemical pilot plant. The obtained experimental results demonstrate that the MPC for tracking is suitable for the control of plants with nonlinear dynamics since it ensures stability and offset-free convergence in case of large changes in the reference even using short prediction horizons. Besides, in case of unreachable set points, the controller steers the system to the closest reachable equilibrium point.

Control no lineal de un aerodeslizador no holonómico con acciones de control limitadas

Abstract: This paper deals with the interesting problem of trajectory tracking of an underactuated hovercraf. This system is difficult to control because it exhibits second order nonholonomic restrictions on its movement. In order to solve the control problem the dynamics of the system is discomposed in two subsystems (one for position and another for orientation) and a nonlinear control law based on two nonlinear cascaded controllers is designed. On one hand the outer, loop computes the forces that are necessary to track the spatial trajectory Fx and Fy in the inertial frame. Based on this forces, the reference for the orientation ψc is computed and it is feedback to the inner loop. On the other hand, the inner loop tries to converge the orientation to the reference ψc given by the outer loop. Both controllers are designed taking into account saturation of the actuators. In fact the main result of this paper shows that if the reference trajectory is given properly it is possible to avoid saturation of the actuators. With adequate selection, global asymptotic stability of the tracking error under the proposed control law is theoretically demonstrated. In addition simulations are made in order to analyze the performance of the control law and the practical meaning of the theoretical results.

Diseño de AUV.Arquitectura de hardware y software

Abstract: El presente documento discute la estrategia bajo la que fueron concebidas la arquitectura de hardware y software para el prototipo de vehiculo autonomo: HRC-AUV , asi como la seleccion de los elementos fundamentales que las componen. El diseno obtenido pondera la sencillez y el desarrollo en condiciones de bajo costo, factores utiles a investigadores que comienzan su actividad en este campo. El trabajo resume las prestaciones que brindan dichas estructuras y las pruebas preliminares de operatividad a que han sido sometidas para demostrar la validez de su empleo en la explotacion de un AUV . De igual forma se presentan los modelos dinamicos linealizados de la planta, utilizados en la sintonia de los lazos de control. La respuesta de dichos lazos y en general del HRC-AUV navegando en el oceano, es presentada a traves de los resultados obtenidos en varias pruebas experimentales.

Calibración de cámaras de tiempo de vuelo: Ajuste adaptativo del tiempo de integración y análisis de la frecuencia de modulación

Abstract: Este trabajo ha sido co-financiado por el Gobierno regional de la Generalitat Valenciana, Universidad de Alicante y CICYT traves de los proyectos GV2012/102, GRE10-16 y DPI2012-32390.

Control Basado en Eventos de Sistemas de Primer Orden Con Retardo

Abstract: PID control theory based on periodic managing of samples has become a well-known area in automatic control. Asynchronous sampling inherent to event-based scheduling causes non-linear dynamics. Under this situation, complex problems arise that must be studied. Issues such as controller tuning, conditions for global stability and the absence of limit cycles are topics that are far from being fully solved yet. In this work, a new event-based scheme of sampling and control for FOPTD processes is presented. The scheme is based on the Smith Predictor structure for delay compensation and the SSOD (Symmetric-Send-On- Delta) scheme for events generation. By means of the proposed scheme, a stability analysis is addressed and a simple tuning methodology with effective interpretation is proposed.

Arquitectura de integración basada en Servicios Web para sistemas heterogéneos y distribuidos: aplicación a robots móviles interactivos.

Abstract: Development of complex robotics systems usually requires tools and software frameworks such as the WillowGarage Robot Operating System (ROS) to ease code integration and reuse tasks. Nevertheless, research in robotics is increasingly becoming multidisciplinary, involving areas of voice technology, artificial intelligence or web systems, which can make researchers highly reluctant to use such robotic-specific systems. Even existing constraints, such as using different Operating Systems, can make this task impossible. This paper presents a novel architecture completely based on standards such as Web Services, that allows a totally decoupled development of agents with minimal intrusiveness and high simplicity by using a dynamic broker to interconnect agents at runtime. This architecture, which does not replace other systems but complements them. It has been used successfully inside a complex multi-robot demonstrator in a fully automated system deployment of an interactive tour guide robot in a new environment.

Beneficios que aporta la metodología MDE a los entornos de desarrollo de sistemas de tiempo real

Abstract: This paper analyses the benefits provided by the Model-Driven Engineering (MDE) methodology when it is used as the base support for real-time systems design environments. The application of MDE eases the environment usage by the designer, because depending on the chosen paradigm and on the stage of the development process, the environment offers a specialized view of the system, presenting in a precise and coherent fashion the information under decision. On the other hand, an MDE infrastructure facilitates the development of tools and their integration within the environment, as, through the model management and transformation mechanisms this infrastructure is equipped with, every tool receives as input the right information it uses, structured as it needs, and likewise, each tool can generate its results in its natural format, relying in that the environment is able to manage them. Finally, the MDE methodology allows considering the real-time design environment as a specialized view meant to be integrated in a more general framework that encompasses the remaining system design stages.

Un Algoritmo basado en Grafos para la Detección Automática de la Luz Arterial en Imágenes Ultrasonográficas

Abstract: Resumen Las enfermedades cardiovasculares se han convertido en la primera causa de fallecimientos en Mexico y en el mundo. La medicion de la intima-media carotidea y del diametro de la luz humeral en imagenes de ultrasonido, son la base de dos de las pruebas destacadas para el diagnostico temprano de este tipo de enfermedades. En este trabajo se presenta una metodologia completamente automatica para la deteccion de la luz arterial en ultrasonografias, necesaria en las pruebas citadas; basada en grafos y deteccion de bordes. El grafo se crea con los puntos intermedios entre los bordes y, su arbol de expansion minima, permite segmentar la luz arterial correcta, aplicando sobre las rutas internas del grafo un criterio de seleccion de longitud y otro de obscuridad. En 294 imagenes, se obtuvo un error promedio en la deteccion de la interfaz entre el lumen de la arteria humeral y la capa intima de su pared cercana de 13.9 μm, con una desviacion estandar de 12.3 μm; y, para la misma interfaz en la pared arterial lejana, de 15.4 μm, con una desviacion estandar de 15.0 μm. La comparacion del desempeno de nuestro algoritmo se realizo contra los resultados presentados en la literatura reciente para tecnicas tanto automaticas como semi-automaticas en esta area, a las cuales supera en precision.

Arquitectura multi-controlador con transferencia sin salto para procesos con conmutación de modos

Abstract: Switched mode systems are defined as those represented by a finite set of linear subsystems and a set of logic rules orchestrating the switching among them. A set of dynamic subsystems together with a logical system that orchestrates the switching among them could be used in order to describe it. An alternative to control them is a supervisory multi-mode controller architecture that allows to meeting multiple control objectives. But switching can consequently derived in bumps or undesirable transient responses. This type of problem is commonly encountered in several industrial application fields, where the Programmable Logic Controller (PLCs) is the most used technology. This paper describes a methodology for the design and the implementation of Bumpless Transfer (BT) mechanisms within multi-rate control architecture for switched mode systems, to be implemented in PLC following the IEC 61131-3 standard. The BT strategy is based on to make controllers which are candidate to switch track to active controllers. By detecting the system state and the feasible next states, the supervisor decides the way in which the different controllers of the architecture execute. The steps to design the complete architecture are presented and experimental results that validate the architecture are shown.

Resultados del Concurso de Ingeniería de Control 2012 y Convocatoria 2013

Abstract: Este trabajo ha sido realizado parcialmente gracias al apoyo del Ministerio de Ciencia e Innovacion (Gobierno de Espana) mediante la Acciones Complementarias DPI2010-12026-E y DPI2011-15857-E, y la Universitat Politecnica de Valencia a ` traves de la beca UPV-FPI-2010/19.

Estabilización del Péndulo Invertido Sobre Dos Ruedas mediante el método de Lyapunov

Abstract: In this paper, a nonlinear controller is presented for the stabilization of the two wheels inverted pendulum. The control strategy is based on partial feedback linealization, in first stage and then a suitable function Lyapunov in conjunction with LaSalle's invariance principle is formed to obtain a stabilizing feedback controller. The obtained closed-loop system is locally asymptotically stable around its unstable equilibrium point, with a computable domain of attraction.

Localización personal en entornos interiores con tecnología RFID

Abstract: Este trabajo presenta un sistema de posicionamiento local (LPS) para personas en entornos interiores basado en la combinacion de tecnologia RFID activa y una metodologia bayesiana de estimacion de la posicion a partir de la fuerza de las senales de RF recibidas. La complejidad inherente a la propagacion de las ondas de RF en entornos interiores causa grandes fluctuaciones en el nivel de la fuerza de la senal, por lo que las tecnicas bayesianas, de naturaleza estadistica, tienen ventajas significativas frente a metodos de posicionamiento mas comunes, como multilateracion, minimizacion cuadratica o localizacion por fingerprinting. En la validacion experimental del sistema RFID-LPS se consigue un error de posicionamiento medio de 2.10 m (mediana de 1.84 m y 3.89 m en el 90% de los casos), en un area abarcada de 475 m2 con 29 tags RFID, y con velocidades de desplazamiento de hasta 0.5 m/s, prestaciones iguales o superiores a otros sistemas del estado del arte. Aunque existen precedentes en Robotica movil, la combinacion de metodos bayesianos y tecnologia RFID activa usada en este trabajo es original en el marco de los sistemas de localizacion de personas, cuyos desplazamientos son generalmente mas impredecibles que los de los robots. Otros aspectos novedosos investigados son la posibilidad de alcanzar una estimacion conjunta de posicion y orientacion de un usuario con dos metodos distintos (uso de antenas directivas y aprovechamiento de la atenuacion de la senal de RF por el cuerpo humano), la escalabilidad del sistema RFID-LPS, y la estimacion de la posicion por tecnicas bayesianas en sistemas simples que pueden detectar los marcadores RFID, pero no medir su fuerza de senal.

Implementación basada en el middleware OROCOS de controladores dinámicos pasivos para un robot paralelo

Abstract: Automatic control of robotic systems nowadays deals more and more with the implementation of different tasks to be achieved by the robot with distinct complexity degree, periodic or aperiodic, with local execution or distributed along a communication network and needing to deal with different kinds of hardware. Deal with the controller implementation for a new robotic platform involved to develop suitable software again for the new hardware or, in the best case, to adapt the existing one. In the recent years it has been tending to a component-based programing that allows to develop reusable software and to use a middleware that make it possible to abstract the developed software from the used hardware and from the available communication protocols. At this paper one of these middleware has been used, specially oriented to robotics as is OROCOS. Using the Orocos Toolchain library the real-time components needed for the implementation of several dynamic controllers for a parallel manipulator have been developed. At this paper the robot and the designed controllers are described and the results over the actual robot are shown.

Análisis del deslizamiento en el punto de apoyo de un robot bípedo de 5-gdl

Abstract: Most of the works which deal with biped robots, consider certain assumptions to evaluate their strategies in order to get a specific objective. One of these assumptions considers a non slip condition, which implies that the contact point with the walking surface does not move along the walking axis. Although, in realistic terms, this movement could be negligible, it depends on the friction characteristics between the points of contact. This work is focused on the slip dynamic analysis of the support contact point and leads to analytical expressions to determine conditions of the slip phenomenon. This is done by considering an additional degree of freedom at the end of the supporting leg in the walking cycle.

Mejora del Proceso de Correspondencia en Imágenes Estereoscópicas Mediante Filtrado Homomórfico y Agrupaciones de Disparidad

Abstract: Resumen En este trabajo se aborda la problematica del proceso de correspondencia de imagenes estereo procedentes de terrenos reales obtenidas con el sistema estereoscopico Videre STH-DCSG 9 mm. Se utilizan una serie de algoritmos de correspondencia, que forman parte del proceso estereoscopico global, cuya finalidad es realizar una reconstruccion 3D para navegacion de robots autonomos en entornos naturales y no estructurados. En primer lugar, se realiza un estudio acerca del efecto que tiene la aplicacion del filtrado homomorfico sobre las imagenes de entrada como paso previo a la correspondencia. Mediante dicho filtrado se consigue mejorar considerablemente el mapa de disparidad, logrando un mayor numero de correspondencias verdaderas en relacion a los procesos de correspondencia sin filtrado. Despues, se realiza un filtrado del mapa de disparidad dirigido por clusters y basado en el principio de continuidad espacial con el que se consigue la eliminacion de falsos positivos o correspondencias erroneas. Ambos filtrados constituyen la principal aportacion del trabajo.

Atributos Relevantes para el Diagnóstico Automático de Eventos de Tensión en Redes de Distribución de Energía Eléctrica

Abstract: This paper focuses on diagnosis of voltage events collected in power distribution networks. Fault networks, induction motor starting, transformer energization and capacitor bank switching cause voltage events. A characterization of voltage events using attributes directly extracted from the voltage and current waveforms is done in this paper. The used attributes are highly related with the event root-cause as well as the relative location of the event source with respect to the measurement point. The relevance of each attribute has been assessed applying a statistical analysis of variance (MANOVA). The most relevant attributes have been used as input to rule-extraction algorithms in order to extract classification rules. The results were obtained using 484 real-world and 38 synthetic voltage events.

Coordinación de sistemas con diferentes dinámicas utilizando conceptos de invarianza geométrica y modos deslizantes

Abstract: Among the possible meanings of the word, this paper will discuss coordinated action to refer to the act on references of systems to achieve some collective behavior desired but considering the constraints and capabilities of each system. To this end, we develop a new methodology based on conditioning techniques reference using geometric invariance sets and sliding mode control. From a general framework, we propose two approaches: a global centralized one and a local through interactions between different systems. The approach addresses the problem of coordinating dynamical systems with possibly different dynamics (e.g. linear and nonlinear, different orders, constraints, etc.).The main idea is to shape the systems local feasible references in order to keep them coordinated.To show the applicability of the approach, the problem of coordinating different dynamical systems with control saturations is addressed as a particular case with both global and local configuration.

Control Neuronal Combinado para Generar Espectros de Oleajes

Abstract: A novel method is presented to control the generation of wave energy spectrum, useful in research and development laboratories of Maritime Hydraulic. The final control element is a linear electric motor connected to a servo-control which is monitored by a computer. The automatic control algorithm is performed with a combined neural scheme. It consists of an artificial neural network “feed-forward” and a proportional integral controller. The computer system includes self-learning based on an online training of the neural network. It makes possible to adapt to changes in the parameters of the “controlled process” and disturbances that impact the studied hydraulic work.

Algoritmo genético permutacional para el despliegue y la planificación de sistemas de tiempo real distribuidos

Abstract: The deployment and scheduling of tasks and messages in distributed real-time systems are NP-hard problems, so there are no optimal methods to solve them in polynomial time. Consequently, these problems are suitable to be approached with generic search and optimisation algorithms. In this paper we propose a multi-objective genetic algorithm based on a permutational solution encoding for the deployment and scheduling of distributed real-time systems. Besides deploying and scheduling tasks and messages, the algorithm can minimize the number of the used computers, the utilization of computing and networking resources and the average worst-case response times of the applications. The experiments show that this genetic algorithm can successfully synthesize complex distributed real-time systems in reasonable times.

Estimación de Estado y Control de un Gasificador de Carbón en Lecho Fluidizado Presurizado

Abstract: Fluidization number (N f ) is an index of fluidization quality, therefore it is a critical variable. However, this variable has not a direct measurement. In this work a structure of estimation state and control of N f is propossed for a coal gasifier in a pressurized fluidized bed. The proposed estimation structure is based on process model, complementary equations, and the combination of classic estimators in order to obtain the N f based on the information of other directly mesaured variables. Based on designed estimator, control loops for N f and hydrogen concentration are proposed. The full structure is analized with changes in inputs and set points, finding totally corrects results.

Herramienta Web Ligera para La Programación en C-Concurrente

Abstract: Tools for computer-aided teaching and learning provide multiple benefits from the point of view of teaching because it allows emphasizing or illustrating certain issues that are sometimes difficult to emphasize without such type of support. This is exactly the case for the tools to detect if there is any type of problem in a concurrent-C program. These tools provide interfaces that can complement the information given by a compiler with additional information about different types of race conditions and memory leaks that appear in the code. This article aims to address how to integrate a core validation tools for concurrent-C as a web application, allowing you to be accessible through the Internet. This tool has been evaluated in an existing programming course, which has shown to be able to provide additional information useful to the learner and the teacher. There have also been a number of performance measures to establish operational limits designed tool within a course that teaches concurrent-C programming.

Mejoras de las Heurísticas Greedy empleadas en el secuenciamiento de los Sistemas de Producción Multi-modelo y Multi-nivel

Abstract: The Car Sequencing Problem consists in maintaining a certain order in the vehicles as they pass through the assembly line. Sequences have to be built according to each vehicle's options, each one requiring different resources and production time, with the objective of avoiding to exceed the maximum human and facility potential. In this paper, we use a Greedy heuristic, the Goal-chasing method developed by Toyota, to solve the Car Sequencing Problem. The concept of different weights for each option is introduced to improve the ordering of the sequence. Nelder-Mead method of nonlinear optimization is applied to obtain the weights of options that minimize the cost of the resulting sequence. The results obtained for a initial sequence are expanded to a set of 30 representative sequences of the automotive industry and they are contrasted with some of the classical benchmarks from the literature. Finally, real data is used to validate the proposal.