Pasar al contenido principal

CONTROL AVANZADO Y APRENDIZAJE DE ROBOTS EN OPERACIONES DE TRANSPORTE

DPI2016-81002-R

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos I+D+i Retos de Investigación
Año convocatoria 2016
Plan Estatal 2013-2016 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 22
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Design and Manufacturing of an Ultra-Low-Cost Custom Torque Sensor for Robotics

  • Pérez-Ubeda, Rodrigo Alonso
  • Gutiérrez, S. C.|||0000-0002-9106-9120
  • Zotovic Stanisic, Ranko|||0000-0002-4746-2101
  • Perles, Angel|||0000-0002-5782-4284
[EN] This article describes a new, very low-cost torque sensor. It was designed to obtain a geometric shape suitable for very affordable manufacturing by machining. The torque sensor was developed under the principle of measurement by strain gauges. It has been designed in order to make manufacturing operations as simple as possible. Optimization was achieved through finite element analysis. Three test sensors for 1, 5, and 20 Nm were designed and machined. Calibration of the three sensors has been carried out obtaining excellent results. An analysis of the dimensional quality of the product and associated costs demonstrates that manufacturing is possible with very simple machining operations, standard tools, and economic equipment, The authors are grateful for the financial support of the Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Tesis doctoral (DoctoralThesis)

Propuesta de inclusión de esfuerzos en el control de un brazo robot para asegurar el cumplimiento de la rugosidad superficial durante operaciones de lijado en diferentes materiales

  • Pérez Ubeda, Rodrigo Alonso
[ES] El mecanizado con brazos robots ha sido estudiado aproximadamente desde los años 90, durante este tiempo se han llevado a cabo importantes avances y descubrimientos en cuanto a su campo de aplicación. En general, los robots manipuladores tienen muchos beneficios y ventajas al ser usados en operaciones de mecanizado, tales como, flexibilidad, gran área de trabajo y facilidad de programación, entre otras, frente a las Máquinas Herramientas de Control numérico (MHCN) que necesitan de una gran inversión para trabajar piezas muy grandes o incrementar sus grados de libertad. Como desventajas, frente a las MHCN, los brazos robóticos poseen menor rigidez, lo que combinado con las altas fuerzas producidas en los procesos de mecanizado hace que aparezcan errores de precisión, desviaciones en las trayectorias, vibraciones y, por consiguiente, una mala calidad en las piezas fabricadas. Entre los brazos robots, los brazos colaborativos están en auge debido a su programación intuitiva y a sus medidas de seguridad, que les permiten trabajar en el mismo espacio que los operadores sin que estos corran riesgos. Como desventaja añadida de los robots colaborativos se encuentra la mayor flexibilidad que estos tienen en sus articulaciones, debido a que incluyen reductores del tipo Harmonic drive. El uso de un control de fuerza en procesos de mecanizado con brazos robots permite controlar y corregir en tiempo real las desviaciones generadas por la flexibilidad en las articulaciones del robot. Utilizar este método de control es beneficioso en cualquier brazo robot; sin embargo, el control interno que incluyen los robots colaborativos presenta ventajas que permiten que el control de fuerza pueda ser aplicado de una manera más eficiente. En el presente trabajo se desarrolla una propuesta real para la inclusión del control de esfuerzos en el brazo robot, así como también, se evalúa y cuantifica la capacidad de los robots industriales y colaborativos en tareas de mecanizado. La propuesta plantea cómo mejorar la utilización de un control de fuerza por bucle interior/exterior aplicado en un brazo colaborativo cuando se desconocen los pares reales de los motores del robot, así como otros parámetros internos que los fabricantes no dan a conocer. Este bucle de control interior/exterior ha sido utilizado en aplicaciones de pulido y lijado sobre diferentes materiales. Los resultados indican que el robot colaborativo es factible para realizar tales operaciones de mecanizado. Sus mejores resultados se obtienen cuando se utiliza un bucle de control interno por velocidad y un bucle de control externo de fuerza con algoritmos, Proporcional-Integral-Derivativo o Proporcional más Pre-Alimentación de la Fuerza., [CA] El mecanitzat amb braços robots ha estat estudiat aproximadament des dels anys 90, durant aquest temps s'han dut a terme importants avanços i descobriments en el que fa al seu camp d'aplicació. En general, els robots manipuladors tenen molts beneficis i avantatges al ser usats en operacions de mecanitzat, com ara, flexibilitat, gran àrea de treball i facilitat de programació, entre d'altres, davant de Màquines Eines de Control Numèric (MECN) que necessiten d'una gran inversió per treballar peces molt grans o incrementar els seus graus de llibertat. Com a desavantatges, enfront de les MECN, els braços robòtics posseeixen menor rigidesa, el que combinat amb les altes forces produïdes en els processos de mecanitzat fa que apareguin errors de precisió, desviacions en les trajectòries, vibracions i, per tant, una mala qualitat en les peces fabricades. Entre els braços robots, els braços col·laboratius estan en auge a causa de la seva programació intuïtiva i a les seves mesures de seguretat, que els permeten treballar en el mateix espai que els operadors sense que aquests corrin riscos. Com desavantatge afegida als robots col·laboratius es troba la major flexibilitat que aquests tenen en les seves articulacions, a causa de que inclouen reductors del tipus Harmonic drive. L'ús d'un control de força en processos de mecanitzat amb braços robots permet controlar, i corregir, en temps real les desviacions generades per la flexibilitat en les articulacions del robot. Utilitzar aquest mètode de control és beneficiós en qualsevol braç robot, però, el control intern que inclouen els robots col·laboratius presenta avantatges que permeten que el control de força es puga aplicar d'una manera més eficient. En el present treball es desenvolupa una proposta real per a la inclusió del control d'esforços en el braç robot, així com s'avalua i quantifica la capacitat dels robots industrials i col·laboratius en tasques de mecanitzat. La proposta planteja com millorar la utilització d'un control de força per bucle interior/exterior aplicat en un braç col·laboratiu, quan es desconeixen els parells reals dels motors del robot, així com altres paràmetres interns que els fabricants no donen a conèixer. Aquest bucle de control interior/exterior ha estat utilitzat en aplicacions de polit sobre diferents materials. Els resultats indiquen que el robot col·laboratiu és factible de realitzar aquestes operacions de mecanitzat. Els seus millors resultats s'obtenen quan s'utilitza un bucle de control intern per velocitat i un bucle de control extern de força amb els algoritmes Proporcional-Integral-Derivatiu o Proporcional més Pre-alimentació de la Força., [EN] Machining with robot arms has been studied approximately since the 90s; during this time, important advances and discoveries have been made in its field of application. In general, manipulative robots have many benefits and advantages when they are used in machining operations, such as flexibility, large work area, and ease of programming, among others, compared to Numerical Control Machine Tools (NCMT) that need a great investment to work very large pieces or increase their degrees of freedom. As for disadvantages, compared to NCMT, robotic arms have lower rigidity, which, combined with the high forces produced in machining processes, causes precision errors, path deviations, vibrations, and, consequently, poor quality in the manufactured parts. Among robot arms, collaborative arms are on the rise due to their intuitive programming and safety measures, which allow them to work in the same space without risk for the operators. An added disadvantage of collaborative robots is their flexibility in their joints because they include Harmonic drive type reducers. The use of force control in machining processes with robot arms makes possible to control and correct, in real-time, the deviations generated by the flexibility in the robot's joints. The use of this control method is beneficial for any robot arm. However, the internal control included in collaborative robots has advantages that allow the force control to be applied more efficiently. In this work, a real proposal is developed to include effort control in the robot arm. The capacity of industrial and collaborative robots in machining tasks is evaluated and quantified. The proposal recommends how to improve the use of an inner/outer force control loop applied in a collaborative arm, when the real torques of the robot's motors are unknown and other internal parameters that manufacturers do not disclose. This inner/outer control loop has been used in polishing and sanding applications on different materials. The results indicate that the collaborative robot is feasible to perform such machining operations. Best results are obtained using an internal velocity control loop and external force control loop with Proportional-Integral-Derivative or Proportional plus Feed Forward., The authors are grateful for the financial support of the Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE). This work was funded by the CONICYT PFCHA/DOCTORADO BECAS CHILE/2017 – 72180157.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Tesis doctoral (DoctoralThesis)

Value Function Estimation in Optimal Control via Takagi-Sugeno Models and Linear Programming

  • Díaz Iza, Henry Paúl
[ES] La presente Tesis emplea técnicas de programación dinámica y aprendizaje por refuerzo para el control de sistemas no lineales en espacios discretos y continuos. Inicialmente se realiza una revisión de los conceptos básicos de programación dinámica y aprendizaje por refuerzo para sistemas con un número finito de estados. Se analiza la extensión de estas técnicas mediante el uso de funciones de aproximación que permiten ampliar su aplicabilidad a sistemas con un gran número de estados o sistemas continuos. Las contribuciones de la Tesis son: -Se presenta una metodología que combina identificación y ajuste de la función Q, que incluye la identificación de un modelo Takagi-Sugeno, el cálculo de controladores subóptimos a partir de desigualdades matriciales lineales y el consiguiente ajuste basado en datos de la función Q a través de una optimización monotónica. -Se propone una metodología para el aprendizaje de controladores utilizando programación dinámica aproximada a través de programación lineal. La metodología hace que ADP-LP funcione en aplicaciones prácticas de control con estados y acciones continuos. La metodología propuesta estima una cota inferior y superior de la función de valor óptima a través de aproximadores funcionales. Se establecen pautas para los datos y la regularización de regresores con el fin de obtener resultados satisfactorios evitando soluciones no acotadas o mal condicionadas. -Se plantea una metodología bajo el enfoque de programación lineal aplicada a programación dinámica aproximada para obtener una mejor aproximación de la función de valor óptima en una determinada región del espacio de estados. La metodología propone aprender gradualmente una política utilizando datos disponibles sólo en la región de exploración. La exploración incrementa progresivamente la región de aprendizaje hasta obtener una política convergida., [CA] La present Tesi empra tècniques de programació dinàmica i aprenentatge per reforç per al control de sistemes no lineals en espais discrets i continus. Inicialment es realitza una revisió dels conceptes bàsics de programació dinàmica i aprenentatge per reforç per a sistemes amb un nombre finit d'estats. S'analitza l'extensió d'aquestes tècniques mitjançant l'ús de funcions d'aproximació que permeten ampliar la seua aplicabilitat a sistemes amb un gran nombre d'estats o sistemes continus. Les contribucions de la Tesi són: -Es presenta una metodologia que combina identificació i ajust de la funció Q, que inclou la identificació d'un model Takagi-Sugeno, el càlcul de controladors subòptims a partir de desigualtats matricials lineals i el consegüent ajust basat en dades de la funció Q a través d'una optimització monotónica. -Es proposa una metodologia per a l'aprenentatge de controladors utilitzant programació dinàmica aproximada a través de programació lineal. La metodologia fa que ADP-LP funcione en aplicacions pràctiques de control amb estats i accions continus. La metodologia proposada estima una cota inferior i superior de la funció de valor òptima a través de aproximadores funcionals. S'estableixen pautes per a les dades i la regularització de regresores amb la finalitat d'obtenir resultats satisfactoris evitant solucions no fitades o mal condicionades. -Es planteja una metodologia sota l'enfocament de programació lineal aplicada a programació dinàmica aproximada per a obtenir una millor aproximació de la funció de valor òptima en una determinada regió de l'espai d'estats. La metodologia proposa aprendre gradualment una política utilitzant dades disponibles només a la regió d'exploració. L'exploració incrementa progressivament la regió d'aprenentatge fins a obtenir una política convergida., [EN] The present Thesis employs dynamic programming and reinforcement learning techniques in order to obtain optimal policies for controlling nonlinear systems with discrete and continuous states and actions. Initially, a review of the basic concepts of dynamic programming and reinforcement learning is carried out for systems with a finite number of states. After that, the extension of these techniques to systems with a large number of states or continuous state systems is analysed using approximation functions. The contributions of the Thesis are: -A combined identification/Q-function fitting methodology, which involves identification of a Takagi-Sugeno model, computation of (sub)optimal controllers from Linear Matrix Inequalities, and the subsequent data-based fitting of Q-function via monotonic optimisation. -A methodology for learning controllers using approximate dynamic programming via linear programming is presented. The methodology makes that ADP-LP approach can work in practical control applications with continuous state and input spaces. The proposed methodology estimates a lower bound and upper bound of the optimal value function through functional approximators. Guidelines are provided for data and regressor regularisation in order to obtain satisfactory results avoiding unbounded or ill-conditioned solutions. -A methodology of approximate dynamic programming via linear programming in order to obtain a better approximation of the optimal value function in a specific region of state space. The methodology proposes to gradually learn a policy using data available only in the exploration region. The exploration progressively increases the learning region until a converged policy is obtained., This work was supported by the National Department of Higher Education, Science, Technology and Innovation of Ecuador (SENESCYT), and the Spanish ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER,UE). The author also received the grant for a predoctoral stay, Programa de Becas Iberoamérica- Santander Investigación 2018, of the Santander Bank.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Subspace-Based Takagi-Sugeno Modeling for Improved LMI Performance

  • Robles-Ruiz, Ruben
  • Bernal Reza, Miguel Ángel
  • Gonzalez-German, Ivan Temoatzin
  • Sala, Antonio|||0000-0002-5691-8772
[EN] Given a nonlinear system, the sector-nonlinearity methodology provides a systematic way of transforming it in an equivalent Takagi-Sugeno (T-S) model. However, such transformation is not unique: conservatism of shape-independent performance conditions in the form of linear matrix inequalities results in some models yielding better results than others. This paper provides some guidelines on choosing a sector-nonlinearity T-S model, with provable optimality (in a particular sense) in the case of quadratic nonlinearities. The approach is based on Hessian and restrictions of a function onto a subspace., This work was supported by the following institutions: Project Ciencia Basica SEP-CONACYT CB-168406, Project DPI2016-81002, (Spanish government, MINECO), Grant PROMETEOII/2013/004 (Generalitat Valenciana) and, the Scholarship GRISOLIA/2014/006.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A linear programming methodology for approximate dynamic programming

  • Diaz, Henry
  • Sala, Antonio|||0000-0002-5691-8772
  • Leopoldo Armesto|||0000-0003-0979-4428
[EN] The linear programming (LP) approach to solve the Bellman equation in dynamic programming is a well-known option for finite state and input spaces to obtain an exact solution. However, with function approximation or continuous state spaces, refinements are necessary. This paper presents a methodology to make approximate dynamic programming via LP work in practical control applications with continuous state and input spaces. There are some guidelines on data and regressor choices needed to obtain meaningful and well-conditioned value function estimates. The work discusses the introduction of terminal ingredients and computation of lower and upper bounds of the value function. An experimental inverted-pendulum application will be used to illustrate the proposal and carry out a suitable comparative analysis with alternative options in the literature., The authors are grateful for the financial support of the Spanish Ministry of Economy and the European Union, grant DPI2016-81002-R (AEI/FEDER, UE), and the PhD grant from the Government of Ecuador (SENESCYT).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Shape-independent model predictive control for Takagi-Sugeno fuzzy systems

  • Ariño-Latorre, Carlos Vicente
  • Querol-Ferrer, Andrés
  • Sala, Antonio|||0000-0002-5691-8772
[EN] Predictive control of TS fuzzy systems has been addressed in prior literature with some simplifying assumptions or heuristic approaches. This paper presents a rigorous formulation of the model predictive control of TS systems, so that results are valid for any membership value (shape-independent) with a suitable account of causality (control can depend on current and past memberships and state). As in most fuzzy control results, a family of progressively better controllers can be obtained by increasing Polya-related complexity parameters, generalising over prior proposals. (C) 2017 Elsevier Ltd. All rights reserved., The authors are grateful to the financial support of Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE), and grant P11B2015-36 (Universitat Jaume I).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Piecewise-Takagi-Sugeno asymptotically exact estimation of the domain of attraction of nonlinear systems

  • Gonzalez-German, Ivan Temoatzin
  • Bernal Reza, Miguel Ángel
  • Robles-Ruiz, Ruben
  • Sala, Antonio|||0000-0002-5691-8772
[EN] This report generalises recent results on stability analysis and estimation of the domain of attraction of nonlinear systems via exact piecewise affine Takagi Sugeno models. Algorithms in the form of linear matrix inequalities are proposed that produce progressively better estimates which are proved to asymptotically render the actual domain of attraction; regions already proven to belong to such domain of attraction can be removed and the estimate can contain significant portions of the modelling region boundary; in this way, level-set approaches in prior literature can be significantly improved. Illustrative examples and comparisons are provided. (C) 2016 The Franklin Institute. Published by Elsevier Ltd. All rights reserved., The authors gratefully acknowledge the support of the following institutions: Project Ciencia Basica SEP-CONACYT CB-168406, the CONACyT/COECYT Sonora scholarship 383252, project DPI2016-81002-R (Spanish government, MINECO), and the scholarship GRISOLIA/2014/006 from Generalitat Valenciana (regional government).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A generalised integral polynomial Lyapunov function for nonlinear systems

  • Gonzalez-German, Ivan Temoatzin
  • Bernal Reza, Miguel Ángel
  • Sala, Antonio|||0000-0002-5691-8772
[EN] This work generalises the line-integral Lyapunov function in (Rhee and Won, 2006) for stability analysis of continuous-time nonlinear models expressed as fuzzy systems. The referred result applied only to Takagi¿Sugeno representations, and required memberships to be a tensor-product of functions of a single state; these are generalised here so that membership arguments can be arbitrary polynomials of the state variables; in this way, systems for which earlier results cannot be applied are now covered. Both the modelling and the integral terms appearing in the Lyapunov functions are generalised to a fuzzy polynomial case. Illustrative examples show the advantage of the proposed method against previous literature, even in the TS case., The authors gratefully to the financial support of Spanish ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE), the CONACyT/COECyT Sonora scholarship 383252, and Project ITSON-PROFAPI-CA 2017-0088.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Tesis doctoral (DoctoralThesis)

Generación de maniobras suaves en el espacio 3D

  • Vanegas Zabala, Gloria Isabel
[ES] El desarrollo tecnológico en la creación de trayectorias que permitan navegación libre de colisiones de Vehículos Autónomos (AVs) ha sido un objetivo constante de estudio debido a su fuerte interés científico y tecnológico en las últimas tres décadas. Las diferentes clases de AVs, ya sean Vehículos Aéreos no Tripulados (UAVs), Vehículos Terrestres no Tripulados (UGVs) o Vehículos Submarinos no Tripulados (UUVs), fomentan el desarrollo e implementación de trayectorias en el espacio tridimensional (3D). Un grupo especial de tecnología UAV está caracterizado por su ala fija, lo cual destaca características particulares en los AVs, debido a las restricciones no-holonómicas (un sistema que se describe mediante un conjunto de parámetros sujetos a restricciones diferenciales que no permiten que un vehículo se mueva de forma instantánea en cualquier dirección). En este sentido, las trayectorias navegables para estos UAVs no deben ser construidas como un conjunto de líneas rectas y círculos como en la gran mayoría de planificadores basados en primitivas, ya que no se garantiza una continuidad en su curvatura. Por lo tanto, las trayectorias construidas para esta rama tecnológica deben ser resueltas considerando las diferentes restricciones de maniobrabilidad del UAV, además de criterios de continuidad de curvas (el problema de continuidad se refiere principalmente a la continuidad geométrica, en términos de continuidad tangencial o de curvatura), suavidad en las curvas (una curva es suave si sus derivadas son continuas en el intervalo definido) y la seguridad en el vuelo (el control de seguridad garantiza que una trayectoria suave esté suficientemente lejos de los obstáculos). Finalmente, la cinemática del movimiento de los vehículos es otro factor que debe ser considerado mientras se suavizan las trayectorias. El presente trabajo está enfocado en la creación de trayectorias navegables en el espacio 3D, para UAVs de características no-holonómicas. La principal dificultad al solventar este problema se debe a la movilidad de esta clase de UAVs, pues se ven obligados a avanzar sin la posibilidad de detenerse a través de trayectorias 3D, realizando curvas con curvaturas limitadas (una máxima capacidad de giro a una velocidad definida). En consecuencia, se han desarrollado las herramientas necesarias para proporcionar una completa caracterización de trayectorias óptimas (con un radio de giro limitado) para UAVs que se mueven en el espacio 3D a una velocidad constante. Esta tesis se centra en la generación de caminos con trayectorias navegables en el espacio Euclídeo 3D, que contenga curvas con curvatura continua, considerando de esta manera las restricciones cinemáticas de los UAVs. Por tal motivo el objetivo principal es el desarrollo de la matemática necesaria para definir curvas clotoides en el espacio tridimensional, de modo que puedan ser utilizadas como primitivas en la generación de trayectorias. Finalmente, culminado el desarrollo de esta herramienta básica, y en función de los obstáculos del entorno, se puede completar una planificación y replanificación activa de movimientos. Para complementar la investigación, la verificación de las herramientas de planificación de trayectorias y del sistema, se han realizado simulaciones con la ayuda del entorno de desarrollo integrado (IDE) Matlab. De la misma forma, se ha preparado una plataforma de simulación de vuelo, tomando las virtudes del simulador de vuelo FlightGear 2018, y el modelo dinámico del avión de ala fija con restricciones no-holonómicas (Kadett 2400 ). En cuanto a la generación de trayectorias 3D, se han desarrollado simulaciones off-line, donde las acciones de control que debe ejecutar el avión para que siga la trayectoria calculada son definidas por: acceleración, brusquedad de curvatura y brusquedad de torsión. Por último, el enfoque de revisión bibliográfica presente en este documento se ha centrado en trabajos realizados que buscan cumplir con las tareas de planificación., [CA] El desenvolupament tecnològic en la creació de trajectòries que permeten navegació lliure de col·lisions de Vehicles Autònoms (AVs) ha estat un objectiu constant d'estudi a causa del seu fort interés científic i tecnològic en les últimes tres dècades. Les diferents classes d'AVs, ja siguen Vehicles Aeris no Tripulats (UAVs), Vehicles Terrestres no Tripulats (UGVs) o Vehicles Submarins no Tripulats (UUVs), fomenten el desenvolupament i la implementació de trajectòries a l'espai tridimensional (3D). Un grup especial de tecnologia UAV està caracteritzat per la seua ala fixa, cosa que destaca característiques particulars en els AVs, a causa de les restriccions no-holonòmiques (un sistema que es descriu mitjançant un conjunt de paràmetres subjectes a restriccions diferencials que no permeten que un vehicle es menege de forma instantània en qualsevol direcció). En aquest sentit, les trajectòries navegables per a aquests UAVs no han de ser construïdes com un conjunt de línies rectes i cercles com a la gran majoria de planificadors basats en primitives, ja que no es garanteix una continuïtat en la seua curvatura. Per tant, les trajectòries construïdes per a aquesta branca tecnològica han de ser resoltes considerant les diferents restriccions de maniobrabilitat de l'UAV, a més de criteris de continuïtat de corbes (el problema de continuïtat es refereix principalment a la continuïtat geomètrica, en termes de continuïtat tangencial o de curvatura), suavitat a les corbes (una corba és suau si les seves derivades són contínues en l'interval definit) i la seguretat en el vol (el control de seguretat garanteix que una trajectòria suau estiga prou lluny dels obstacles). Finalment, la cinemàtica del moviment dels vehicles és un altre factor que cal considerar mentre se suavitzen les trajectòries. Aquest treball està enfocat a la creació de trajectòries navegables a l'espai 3D, per a UAVs de característiques no-holonòmiques. La principal dificultat en solucionar aquest problema es deu a la mobilitat d'aquesta classe de UAVs, ja que es veuen obligats a avançar sense la possibilitat d'aturarse a través de trajectòries 3D, fent corbes amb curvatures limitades (una màxima capacitat de gir a una velocitat definida). En conseqüència, s'han desenvolupat les ferramentes necessàries per proporcionar una completa caracterització de trajectòries òptimes (amb un radi de gir limitat) per a UAVs que es mouen al pla 3D a una velocitat constant. Aquesta tesi se centra en la generació de camins amb trajectòries navegables a l'espai Euclidià 3D, que continguen corbes amb curvatura contínua, considerant així les restriccions cinemàtiques dels UAVs. Per aquest motiu, l'objectiu principal és el desenvolupament de la matemàtica necessària per definir corbes clotoides a l'espai tridimensional, de manera que puguen ser utilitzades com a primitives en la generació de trajectòries. Finalment, culminat el desenvolupament d'aquesta ferramenta bàsica, i en funció dels obstacles de l'entorn, es pot completar una planificació i una replanificació activa de moviments. Per complementar la investigació, la verificació de les ferramentes de planificació de trajectòries i del sistema, s'han fet simulacions amb l'ajuda de l'entorn de desenvolupament integrat (IDE) Matlab. De la mateixa manera, s'ha preparat una plataforma de simulació de vol, prenent les virtuts del simulador de vol FlightGear 2018 i el model dinàmic de l'avió d'ala fixa amb restriccions no-holonòmiques (Kadett 2400). Pel que fa a la generació de trajectòries 3D, s'han desenvolupat simulacions off-line, on les accions de control que ha d'executar l'avió perquè seguisca la trajectòria calculada són definides per: acceleració, brusquedat de curvatura i brusquedat de torsió. Finalment, l'enfocament de revisió bibliogràfica present en aquest document s'ha centrat en treballs realitzats que busquen complir les tasques de planificació de trajectòria, planificació de moviment i construcció de corbes suaus per a AVs., [EN] The technological development in the creation of trajectories that allow collision-free navigation of Autonomous Vehicles (AVs) has been a continuous target of study due to its strong scientific and technological interest in the last three decades. Different classes of AVs, whether, Unmanned Aerial Vehicles (UAVs), Unmanned Ground Vehicles (UGVs) or Unmanned Underwater Vehicles (UUVs), encourage the development and implementation of paths in three-dimensional (3D) space. A special group of UAV technology is characterized by its fixed wing, which emphasizes particular characteristics in UAVs, due to non-holonomic constraints (a system that is described by a set of parameters subject to differential constraints that do not allow a vehicle to move instantaneously in any direction). In this sense, navigable paths for these UAVs should not be built as a set of straight lines and circles as in the vast majority of primitive-based planners, since no continuity in their curvature is guaranteed. Therefore, the paths built for this technology branch must be solved considering the different maneuverability constraints of the UAV, in addition to curve continuity criteria (the continuity problem refers mainly to geometric continuity, in terms of tangential or curvature continuity), curve smoothness (a curve is smooth if its derivatives are continuous in the defined interval) and flight safety (safety control ensures that a smooth path is sufficiently far away from obstacles). Finally, the kinematics of vehicle motion is another factor to be considered while smoothing paths. This thesis work is focused on the creation of navigable paths in 3D space for UAVs with non-holonomic characteristics. The main difficulty in solving this problem is due to the mobility of this kind of UAVs, since they are forced to move without the possibility of stopping through 3D paths, performing curves with limited curvatures (a maximum turning capacity at a defined speed). Consequently, the needed tools have been developed to provide a complete characterization of optimal paths (with a limited turning radius) for UAVs moving in the 3D plane at a constant velocity. This thesis focuses on the generation of paths with navigable trajectories in 3D Euclidean space, containing curves with continuous curvature, thus considering the kinematic constraints of UAVs. Therefore, the main aim is the development of the necessary mathematics to define clothoid curves in the three-dimensional space, so that they can be used as primitives in the generation of paths. Finally, once the development of this basic tool has been completed, and depending on the obstacles in the environment, an active planning and replanning of movements can be completed. To complement the research, the verification of the path planning tools and the system, simulations have been performed with the help of the integrated development environment (IDE) Matlab. In the same way, a flight simulation platform has been prepared, taking the virtues of the FlightGear 2018 flight simulator, and the dynamic model of the fixed-wing aircraft with non-holonomic constraints (Kadett 2400 ). Regarding the generation of 3D paths, off-line simulations have been developed, where the control actions to be executed by the aircraft to follow the calculated path are defined by: acceleration, curvature sharpness and torsion sharpness. Finally, the literature review approach presented in this document has focused on works that address the tasks of path planning, motion planning and construction of smooth curves for AVs. Special care has been taken in the methodologies used, the variety of techniques, in addition to the advantages and disadvantages presented throughout the literature review., The authors are grateful to the financial support of Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE). This work was also supported by the postdoctoral fellowship “APOSTD/2017/055” and the local administration “GV/2017/029” (Generalitat Valenciana, Conse- lleria d’Educació) Valencia - Spain.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Asymptotically exact stabilisation for constrained discrete Takagi-Sugeno systems via set-invariance

  • Ariño-Latorre, Carlos Vicente
  • Pérez Soler, Emilio
  • Bedate Boluda, Fernando
  • Querol-Ferrer, Andrés
  • Sala, Antonio|||0000-0002-5691-8772
[EN] Given a Takagi-Sugeno (TS) system, this paper proposes a novel methodology to obtain the state feedback controller guaranteeing, asymptotically as a Polya-related complexity parameter grows, the largest (membership-shape independent) possible domain-of-attraction with contraction-rate performance lambda, based on polyhedral lambda-contractive sets from constrained linear systems literature. The resulting controller is valid for any realisation of the memberships, as usual in most TS literature. For a finite complexity parameter, an inner estimate of such largest set is obtained; the frontier of such approximation can be understood as the level set of a polyhedral control-Lyapunov function. Convergence of a proposed iterative algorithm is asymptotically necessary and sufficient for TS system stabilisation: for a high-enough value of the complexity parameter, any conceivable shape-independent Lyapunov controller design procedure will yield a proven domain of attraction smaller or equal to the algorithm's output. (C) 2016 Elsevier B.V. All rights reserved., This work has been supported by grants DPI2015-70433- P and DPI2016-81002-R, from Spanish Government (MINECO) and grant PROMETEOII/2013/004 from Generalitat Valenciana.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Asynchronous Sensor Fusion of GPS, IMU and CAN-Based Odometry for Heavy-Duty Vehicles

  • Girbes-Juan, Vicent
  • Hernandez-Ferrandiz, Daniel
  • Leopoldo Armesto|||0000-0003-0979-4428
  • Dols Ruiz, Juan Francisco|||0000-0003-1815-1360
  • Sala, Antonio|||0000-0002-5691-8772
[EN] In heavy-duty vehicles, multiple signals are available to estimate the vehicle's kinematics, such as Inertial Measurement Unit (IMU), Global Positioning System (GPS) and linear and angular speed readings from wheel tachometers on the internal Controller Area Network (CAN). These signals have different noise variance, bandwidth and sampling rate (being the latter, possibly, irregular). In this paper we present a non-linear sensor fusion algorithm allowing asynchronous sampling and non-causal smoothing. It is applied to achieve accuracy improvements when incorporating odometry measurements from CAN bus to standard GPS+IMU kinematic estimation, as well as the robustness against missing data. Our results show that this asynchronous multi-sensor (GPS+IMU+CAN-based odometry) fusion is advantageous in low-speed manoeuvres, improving accuracy and robustness to missing data, thanks to non-causal filtering. The proposed algorithm is based on Extended Kalman Filter and Smoother, with exponential discretization of continuous-time stochastic differential equations, in order to process measurements at arbitrary time instants; it can provide data to subsequent processing steps at arbitrary time instants, not necessarily coincident with the original measurement ones. Given the extra information available in the smoothing case, its estimation performance is less sensitive to the noise-variance parameter setting, compared to causal filtering. Working Matlab code is provided at the end of this work., This research was supported in part by the Agencia Espanola de Investigacion (European Union) under Grants PID2020-116585GB-I00 and PID2020-118071GB-I00, and in part by the Generalitat Valenciana under Grant GV/2021/074. The review of this article was coordinated by Dr. Sohel Anwar. (Corresponding author: Vicent Girbes-Juan.)
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Stability analysis of LPV systems: Scenario approach

  • Sala, Antonio|||0000-0002-5691-8772
[EN] This paper discusses a `scenario¿ approach to prove decay-rate stability of discrete-time polytopic linear parameter-varying systems, dealing with sets of sequences of vertex models of different length. When all sequences have the same length, parameter-trajectory dependent results in earlier literature are obtained as particular cases. The approach in this paper discusses `classical¿ stability, without the need of probabilistic ingredients present in other scenario-based ideas in literature. A numerical example shows that the proposal achieves a sensible tradeoff between proven performance and computing requirements., The author is grateful to Ministerio de Economia (Spain) and European Union, grant DPI2016-81002-R The material in this paper was not presented at any conference. This paper was recommended for publication in revised form by Associate Editor Tingshu Hu under the direction of Editor Andre L. Tits.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Drive Force and Longitudinal Dynamics Estimation in Heavy-Duty Vehicles

  • Girbés, Vicent
  • Hernández, Daniel
  • Leopoldo Armesto|||0000-0003-0979-4428
  • Dols Ruiz, Juan Francisco|||0000-0003-1815-1360
  • Sala, Antonio|||0000-0002-5691-8772
[EN] Modelling the dynamic behaviour of heavy vehicles, such as buses or trucks, can be very useful for driving simulation and training, autonomous driving, crash analysis, etc. However, dynamic modelling of a vehicle is a difficult task because there are many subsystems and signals that affect its behaviour. In addition, it might be hard to combine data because available signals come at different rates, or even some samples might be missed due to disturbances or communication issues. In this paper, we propose a non-invasive data acquisition hardware/software setup to carry out several experiments with an urban bus, in order to collect data from one of the internal communication networks and other embedded systems. Subsequently, non-conventional sampling data fusion using a Kalman filter has been implemented to fuse data gathered from different sources, connected through a wireless network (the vehicle¿s internal CAN bus messages, IMU, GPS, and other sensors placed in pedals). Our results show that the proposed combination of experimental data gathering and multi-rate filtering algorithm allows useful signal estimation for vehicle identification and modelling, even when data samples are missing., This research was funded by the Spanish Ministry of Economy and European Union, grant DPI2016-81002-R, by Generalitat Valenciana, grant APOSTD/2017/055, and by the European Union too through the European Social Fund (ESF). The authors want to acknowledge public transport operators Autos Vallduxense S.L. and EMT Valencia for their support to the project by lending their vehicles and bus drivers. The authors are also thankful to Instituto de Diseño y Fabricación and SAFETRANS project (DPI2013-42302-R) research teams, for lending most of the electronic devices and sensors used in the experimentation.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Performance-oriented quasi-LPV modeling of nonlinear systems

  • Robles, Ruben
  • Bernal, Miguel
  • Sala, Antonio|||0000-0002-5691-8772
[EN] Polytopic quasi¿linear parameter¿varying (quasi¿LPV) models of nonlinear processes allow the usage linear matrix inequalities (LMIs) to guarantee some performance goal on them (in most cases, locally, over a so¿called modeling region). In order to get a finite number of LMIs, nonlinearities are embedded on the convex hull of a finite set of linear models. However, for a given system, the quasi¿LPV representations are not unique, yielding different performance bounds depending on the model choice. To avoid such drawback, earlier literature on the topic used annihilator¿based approaches, which require gridding on the modeling region, and nonconvex BMI conditions for controller synthesis; optimal performance bounds are obtained, but with a huge computational burden. This paper proposes building a model by minimizing the projection of the nonlinearities onto directions, which are deleterious for performance. For a small modeling region, these directions are obtained from LMIs with the linearized model. Additionally, these directions will guide the selection of the polytopic embedding's vertices. The procedure allows gridding¿free LMI controller synthesis, as in standard LPV setups, with a very reduced performance loss with respect to the aforementioned BMI+gridding approaches, at a fraction of the computational cost., Spanish ministry of Economy and European Union (AEI/FEDER, UE), Grant/Award Number: DPI2016-81002-R; Generalitat Valenciana, Grant/Award Number: GRISOLIA/2014/006; Instituto Tecnologico de Sonora (ITSON), Grant/Award Number: ITSON PFCE 2017-18 and PROFAPI CA-18 ITSON 2018-1062
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Force Control Improvement in Collaborative Robots through Theory Analysis and Experimental Endorsement

  • Pérez-Ubeda, Rodrigo
  • Zotovic Stanisic, Ranko|||0000-0002-4746-2101
  • Gutiérrez, S. C.|||0000-0002-9106-9120
[EN] Due to the elasticity of their joints, collaborative robots are seldom used in applications with force control. Besides, the industrial robot controllers are closed and do not allow the user to access the motor torques and other parameters, hindering the possibility of carrying out a customized control. A good alternative to achieve a custom force control is sending the output of the force regulator to the robot controller through motion commands (inner/outer loop control). There are different types of motion commands (e.g., position or velocity). They may be implemented in different ways (Jacobian inverse vs. Jacobian transpose), but this information is usually not available for the user. This article is dedicated to the analysis of the effect of different inner loops and their combination with several external controllers. Two of the most determinant factors found are the type of the inner loop and the stiffness matrix. The theoretical deductions have been experimentally verified on a collaborative robot UR3, allowing us to choose the best behaviour in a polishing operation according to pre-established criteria., The authors are grateful for the financial support of the Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE), to the research work here published. Rodrigo Perez-Ubeda is grateful to the Ph.D. Grant CONICYT PFCHA/DOCTORADO BECAS CHILE/2017-72180157.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Fitted Q-Function Control Methodology Based on Takagi-Sugeno Systems

  • Diaz-Iza, Henry Paul
  • Leopoldo Armesto|||0000-0003-0979-4428
  • Sala, Antonio|||0000-0002-5691-8772
[EN] This paper presents a combined identification/ Q-function fitting methodology that involves identification of a Takagi-Sugeno model, computation of (sub)optimal controllers from linear matrix inequalities (LMIs), and subsequent data-based fitting of the Q-function via monotonic optimization. The LMI-based initialization provides a conservative solution, but it is a sensible starting point to avoid convergence/local-minima issues in raw data-based fitted Q-iteration or Bellman residual minimization. An inverted-pendulum experimental case study illustrates the approach., This work was supported in part by the Spanish Ministry of Economy and European Union (AEI/FEDER, UE) under Grant DPI2016-81002-R and in part by the Government of Ecuador through the Ph.D. Grant SENESCYT.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Constraint-aware learning of policies by demonstration

  • Leopoldo Armesto|||0000-0003-0979-4428
  • Sala, Antonio|||0000-0002-5691-8772
  • Moura, Joao
  • Ivan, Vladimir
  • Erden, Mustafa Suphi
  • Vijayakumar, Sethu
[EN] Many practical tasks in robotic systems, such as cleaning windows, writing, or grasping, are inherently constrained. Learning policies subject to constraints is a challenging problem. In this paper, we propose a method of constraint-aware learning that solves the policy learning problem using redundant robots that execute a policy that is acting in the null space of a constraint. In particular, we are interested in generalizing learned null-space policies across constraints that were not known during the training. We split the combined problem of learning constraints and policies into two: first estimating the constraint, and then estimating a null-space policy using the remaining degrees of freedom. For a linear parametrization, we provide a closed-form solution of the problem. We also define a metric for comparing the similarity of estimated constraints, which is useful to pre-process the trajectories recorded in the demonstrations. We have validated our method by learning a wiping task from human demonstration on flat surfaces and reproducing it on an unknown curved surface using a force- or torque-based controller to achieve tool alignment. We show that, despite the differences between the training and validation scenarios, we learn a policy that still provides the desired wiping motion., The author(s) disclosed receipt of the following financial support for the research, auth/orship, and/or publication of this article: This work was supported by the Spanish Ministry of Economy and the European Union (grant number DPI2016-81002-R (AEI/FEDER, UE)), the European Union Horizon 2020, as part of the project Memory of Motion - MEMMO (project ID 780684), and the Engineering and Physical Sciences Research Council, UK, as part of the Robotics and AI hub in Future AI and Robotics for Space - FAIR-SPACE (grant number EP/R026092/1), and as part of the Centre for Doctoral Training in Robotics and Autonomous Systems at Heriot-Watt University and the University of Edinburgh (grant numbers EP/L016834/1 and EP/J015040/1)
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Metodología de programación dinámica aproximada para control óptimo basada en datos, Approximate Dynamic Programming Methodology for Data-based Optimal Controllers

  • Díaz, Henry
  • Leopoldo Armesto|||0000-0003-0979-4428
  • Sala, Antonio|||0000-0002-5691-8772
[EN] In this article, we present a methodology for learning data-based approximately optimal controllers, within the context of learning and approximate dynamic programming. There are previous solutions in dynamic programming that use linear programming in discrete state space, but cannot be applied directly to continuous space. The objective of the methodology is to calculate data-based optimal controllers for continuous state space, these controllers are obtained by a lower estimation of the accumulated cost through functional approximators with linear parameterization. This is solved non-iteratively with linear programming, but it requires to provide appropriate conditions for regressor regularization and to introduce a cost of leaving the region with valid data, in order to obtain satisfactory results (avoiding unrestricted or poorly conditioned solutions)., [ES] En este artículo se presenta una metodología para el aprendizaje de controladores óptimos basados en datos, en el contexto de la programación dinámica aproximada. Existen soluciones previas en programación dinámica que utilizan programación lineal en espacios de estado discretos, pero que no se pueden aplicar directamente a espacios continuos. El objetivo de la metodología es calcular controladores óptimos para espacios de estados continuos, basados en datos, obtenidos mediante una estimación inferior del coste acumulado a través de aproximadores funcionales con parametrización lineal. Esto se resuelve de forma no iterativa con programación lineal, pero requiere proporcionar las condiciones adecuadas de regularización de regresores e introducir un coste de abandono de la región con datos válidos, con el fin de obtener resultados satisfactorios (evitando soluciones no acotadas o mal condicionadas)., Agradecemos al Ministerio de Economía de España, la Unión Europea DPI2016-81002-R (AEI/FEDER, UE), y al Gobierno de Ecuador (Beca SENESCYT) la financiación recibida para la línea de investigación objeto de este trabajo.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Design and manufacturing of a protoype of a lightweight robot arm

  • Gutiérrez, S. C.|||0000-0002-9106-9120
  • Zotovic Stanisic, Ranko|||0000-0002-4746-2101
  • Navarro-Mas, M.D.|||0000-0003-2022-8857
  • Meseguer, Maria-Desamparados|||0000-0003-0254-7993
[EN] The aim of this work is to manufacture a prototype of a lightweight robot arm with a low cost budget, fully functional. This prototype is used to test and fix the elements for driving and controlling. During the development process, several tests and studies were performed, such as, strength simulations, dimensional effects after a post-process treatment with acetone, adjustment of control parameters to improve the accuracy, testing of behaviour of transmissions, etc. The prototype must have a low weight overall and a right operation. The results and conclusions, related with material, reinforcements, geometry/shape of the parts, etc., become recommendations for the manufacture of the final lightweight robot arm. The one that will not be prototype., The authors are grateful to the financial support of Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Study of the application of a collaborative robot for machining tasks

  • Pérez-Ubeda, R.
  • Gutiérrez, S. C.|||0000-0002-9106-9120
  • Zotovic Stanisic, Ranko|||0000-0002-4746-2101
  • Lluch-Cerezo, Joaquín|||0000-0003-1973-0188
[EN] The importance of collaborative robots is increasing very fast in the industry. They have several advantages over the 'classical' robot arms: they may work side-by-side with humans, their environment needs less adaptation, they may be easily transported, etc. Their joints are more elastic than those in classical robots. For this reason, they are less suited for machining. In this work, a collaborative robot, a sensor of 6 Degree of Freedom (DOF) and a spindle with flex-shaft attachment are used to perform milling operations on soft materials. An inner/outer loop control is being developed to control the movements and the cutting forces. The experiments have been designed to evaluate the capability of the robot with milling operations with different parameters. An analysis of the dimensions and the finished surface will be carried out. The contribution of this article is to determine the possibilities and limitations of the collaborative robots in machining applications, with external control of forces., The authors are grateful for the financial support of the Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE). This work was funded by the CONICYT PFCHA/DOCTORADO BECAS CHILE/2017 - 72180157
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Clothoid-Based Three-Dimensional Curve for Attitude Planning

  • Girbés, Vicent
  • Vanegas, Gloria
  • Leopoldo Armesto|||0000-0003-0979-4428
This work was supported by Generalitat Valenciana under the postdoctoral grant APOSTD/2017/055. The authors are also grateful to the financial support of Spanish Ministry of Economy and European Union, grant DPI2016-81002-R (AEI/FEDER, UE). Thanks to Sergio Garcia-Nieto and Franklin Samaniego for their help to setup the simulator.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Optimisation of transient and ultimate inescapable sets with polynomial boundaries for nonlinear systems

  • Sala, Antonio|||0000-0002-5691-8772
  • Pitarch, José Luis
[EN] This paper addresses the problem of bounding the trajectories of nonlinear systems (transient and ultimate bounds) from initial conditions in given sets, when subject to possibly nonvanishing disturbances constrained by some finite-interval integral bounds, with a suitable controller. The so-called robustly inescapable sets are determined from such initial conditions and disturbance bounds. In order to get numerical results, the approach considers embedding the nonlinear dynamics in a convex combination of polynomials, and solving sum-of-squares (SOS) problems on them, optimising some inescapable-set size parameters. Determination of approximate (locally) optimal solutions usually requires an iterative evaluation of SOS problems, because of products of decision variables. (C) 2016 Elsevier Ltd. All rights reserved., The research leading to these results has received funding from the European Union (FP7/2007-2013 no 604068) and from the Spanish Government (MINECO/FEDER DPI2015-70975-P, DPI2016-81002). The material in this paper was not presented at any conference. This paper was recommended for publication in revised form by Associate Editor Graziano Chesi under the direction of Editor Richard Middleton.