This paper presents an alternative for UAVs hover regulation by means of ADRC approach. The propo... more This paper presents an alternative for UAVs hover regulation by means of ADRC approach. The proposed control strategy was designed using Lyapunov's method to ensure asymptotic stability for vertical motion of the vehicle, including a sliding-observer for disturbance detection. The proposed controller was implemented in a Bebop 2 quadrotor using ROS middleware, and it was compared with a PD controller, showing successful results for altitude regulation even in presence of disturbances and unconsidered dynamics.
Mexico Electromechanical system includes a graphic user interface to select loads and display tar... more Mexico Electromechanical system includes a graphic user interface to select loads and display target trajectories. Time response of electromechanicsl system was simulated with a Feedforward + PD controller. Elbow muscles response, kinetic variables and time required to perform measurements were examined in 3 healthy adults during the active elbow movement at 2 loads and compared to a pulley machine. Measurement time of the electromechanical system was reduced to 57.29%; the angular velocity was similar in both systems; the brachioradialis muscle response in the electromechanical system was different for flexion and extension. The results indicate that the electromechanical system accomplish the objectives, but it is suggested to implement a feedback controller with disturbance handling as the simulated.
We propose an artificial vision algorithm for a semi-autonomous brain-computer interface (BCI). T... more We propose an artificial vision algorithm for a semi-autonomous brain-computer interface (BCI). The interface was designed in such a way that users are able to manipulate a robotic arm to pick up an object from a table and place it in one of two possible locations indicated as goal disks, and the manipulation is performed without any concern about continuous control of the final effector. The implemented algorithm is used to obtain, in real time, the positions of the object and the disks in reference to the robot frame. The main techniques used in the proposed algorithm were color segmentation and homography transformation. The implementation of the algorithm allows to obtain the positions of all the items in the table, and it successfully performs pick and place tasks, setting the items on different positions across the table.
This paper presents the combination of a robust controller with robust visual feedback for autono... more This paper presents the combination of a robust controller with robust visual feedback for autonomous flight of a quadrotor. The proposed control strategy estimates the disturbances and rejects them in real time using visual information from the on-board camera and a robust observer to detect disturbances. The navigation, vision and control algorithms are executed in parallel threads using Robotics Operating System in order to enhance the control system performance. The proposed visual control technique was tested by giving a target pose to the quadrotor, which was successfully reached with acceptable performance even in presence of wind disturbances that are intrinsic of outdoors environments. The satisfactory results showed that the proposed combination of robust control and visual feedback techniques provide a suitable performance for autonomous navigation of quadrotors.
This paper presents the implementation of active disturbance rejection control approach for motio... more This paper presents the implementation of active disturbance rejection control approach for motion control of humanoid robots. The main contribution of this work is the usage of sliding and quasi-sliding observers to enhance the performance of such control strategy. The presented control schemes are based on different modelling techniques of humanoid robots by means of both continuous and discrete time analysis. Such controllers were tested in simulation and in real time experiments using a NAO humanoid robot and showing successful results in walking and balancing control in comparison with a linear Luenberger-observer based similar controller.
The aim of this paper is to present all the phases needed by a humanoid to perform bipedal locomo... more The aim of this paper is to present all the phases needed by a humanoid to perform bipedal locomotion, from walking pattern generation to joint control, passing through kinematic and dynamic modeling. All these phases are applied to the HRP-4 humanoid for simulation experiments. Amongst the different gait pattern generation methods, the 3D LIP method is chosen, while to obtain the inverse kinematic model the geometric approach is applied, and the classical Euler-Lagrange method is implemented to derive the dynamics of the planar (Sagittal) biped walking. A feedback linearization control was developed for joint references tracking. Finally, simulation results are obtained to analyze the performance of designed controllers.
Pädi Boletín Científico de Ciencias Básicas e Ingenierías del ICBI
En este trabajo se presenta el desarrollo de un sistema de navegación autónoma para un robot móvi... more En este trabajo se presenta el desarrollo de un sistema de navegación autónoma para un robot móvil basado en una cámara RGB-D, que le permita navegar en un entorno estructurado, pero desconocido, formado por pasillos y las salas que tengan sus entradas accesibles desde dichos pasillos. Entre las diferentes aplicaciones que podría tener un robot móvil con este sistema de navegación están: el mapeo, la desinfección en interiores (SARS-CoV-2) con luz ultravioleta, la inspección autónoma de estos recintos y otras muchas que requieran visitar de manera autónoma las salas a las que se tenga acceso desde un pasillo o conjunto de pasillos interconectados. Se implementó dicho sistema tanto en simulación como en tiempo real, obteniendo para ambos casos los resultados esperados al navegar por un recinto.
2016 IEEE Colombian Conference on Robotics and Automation (CCRA), 2016
Musculoskeletal Models are useful for designing rehabilitation technology (RT). MSMS software has... more Musculoskeletal Models are useful for designing rehabilitation technology (RT). MSMS software has unique features which let simulate RT along with different controllers. Unfortunately there are technical differences across platforms which yield to different results. Therefore benchmarking of simulations has been proposed. Here we present the implementation of a Shoulder and Elbow musculosketal (SE) model based on the MoBL ARMS model. SE model consists of 5 rigid segments linked with 5 degrees of freedom driven by 12 muscles. Muscle fascicle length (FL) and active fiber force (AFF) were calculated during simulation of 2 single joint shoulder and elbow movements. RMS errors were calculated between results. Differences between models were smallest for muscle FL (RMSE≤0.0037) and AFF (RMSE≤22.0155). RMS errors were biggest for FL of deltoid posterior muscle (RMSE=0.1165) and for AFF of supraspinatus muscle (RMSE=164.97) during shoulder abduction. But both models showed similar outputs.
A NEW REFORMULATION OF PHILIPS ALGORITHM FOR THE COMPUTATION OF DISCRETE IMAGE MOMENTS IS PRESENT... more A NEW REFORMULATION OF PHILIPS ALGORITHM FOR THE COMPUTATION OF DISCRETE IMAGE MOMENTS IS PRESENTED. AS PHILIPS METHOD, THE REFORMULATION PRODUCES THE SAME EXACT RESULTS BUT IN A FASTER MANNER. MOMENT CONTRIBUTIONS DUE TO THE PRESENCE OF HOLES IN THE SDHAPE /WHICH ARE NOT TAKEN INTO ACCOUNT, TO OUR KNOWLEDGE, INTO ANY BOWDARY BASED METHOD FOR MOMENT CALCULATION INCLUIDING PHILIPS) ARE ALSO INTRODUCED INTO THE NEW METHOD.
This paper presents an alternative for UAVs hover regulation by means of ADRC approach. The propo... more This paper presents an alternative for UAVs hover regulation by means of ADRC approach. The proposed control strategy was designed using Lyapunov's method to ensure asymptotic stability for vertical motion of the vehicle, including a sliding-observer for disturbance detection. The proposed controller was implemented in a Bebop 2 quadrotor using ROS middleware, and it was compared with a PD controller, showing successful results for altitude regulation even in presence of disturbances and unconsidered dynamics.
Mexico Electromechanical system includes a graphic user interface to select loads and display tar... more Mexico Electromechanical system includes a graphic user interface to select loads and display target trajectories. Time response of electromechanicsl system was simulated with a Feedforward + PD controller. Elbow muscles response, kinetic variables and time required to perform measurements were examined in 3 healthy adults during the active elbow movement at 2 loads and compared to a pulley machine. Measurement time of the electromechanical system was reduced to 57.29%; the angular velocity was similar in both systems; the brachioradialis muscle response in the electromechanical system was different for flexion and extension. The results indicate that the electromechanical system accomplish the objectives, but it is suggested to implement a feedback controller with disturbance handling as the simulated.
We propose an artificial vision algorithm for a semi-autonomous brain-computer interface (BCI). T... more We propose an artificial vision algorithm for a semi-autonomous brain-computer interface (BCI). The interface was designed in such a way that users are able to manipulate a robotic arm to pick up an object from a table and place it in one of two possible locations indicated as goal disks, and the manipulation is performed without any concern about continuous control of the final effector. The implemented algorithm is used to obtain, in real time, the positions of the object and the disks in reference to the robot frame. The main techniques used in the proposed algorithm were color segmentation and homography transformation. The implementation of the algorithm allows to obtain the positions of all the items in the table, and it successfully performs pick and place tasks, setting the items on different positions across the table.
This paper presents the combination of a robust controller with robust visual feedback for autono... more This paper presents the combination of a robust controller with robust visual feedback for autonomous flight of a quadrotor. The proposed control strategy estimates the disturbances and rejects them in real time using visual information from the on-board camera and a robust observer to detect disturbances. The navigation, vision and control algorithms are executed in parallel threads using Robotics Operating System in order to enhance the control system performance. The proposed visual control technique was tested by giving a target pose to the quadrotor, which was successfully reached with acceptable performance even in presence of wind disturbances that are intrinsic of outdoors environments. The satisfactory results showed that the proposed combination of robust control and visual feedback techniques provide a suitable performance for autonomous navigation of quadrotors.
This paper presents the implementation of active disturbance rejection control approach for motio... more This paper presents the implementation of active disturbance rejection control approach for motion control of humanoid robots. The main contribution of this work is the usage of sliding and quasi-sliding observers to enhance the performance of such control strategy. The presented control schemes are based on different modelling techniques of humanoid robots by means of both continuous and discrete time analysis. Such controllers were tested in simulation and in real time experiments using a NAO humanoid robot and showing successful results in walking and balancing control in comparison with a linear Luenberger-observer based similar controller.
The aim of this paper is to present all the phases needed by a humanoid to perform bipedal locomo... more The aim of this paper is to present all the phases needed by a humanoid to perform bipedal locomotion, from walking pattern generation to joint control, passing through kinematic and dynamic modeling. All these phases are applied to the HRP-4 humanoid for simulation experiments. Amongst the different gait pattern generation methods, the 3D LIP method is chosen, while to obtain the inverse kinematic model the geometric approach is applied, and the classical Euler-Lagrange method is implemented to derive the dynamics of the planar (Sagittal) biped walking. A feedback linearization control was developed for joint references tracking. Finally, simulation results are obtained to analyze the performance of designed controllers.
Pädi Boletín Científico de Ciencias Básicas e Ingenierías del ICBI
En este trabajo se presenta el desarrollo de un sistema de navegación autónoma para un robot móvi... more En este trabajo se presenta el desarrollo de un sistema de navegación autónoma para un robot móvil basado en una cámara RGB-D, que le permita navegar en un entorno estructurado, pero desconocido, formado por pasillos y las salas que tengan sus entradas accesibles desde dichos pasillos. Entre las diferentes aplicaciones que podría tener un robot móvil con este sistema de navegación están: el mapeo, la desinfección en interiores (SARS-CoV-2) con luz ultravioleta, la inspección autónoma de estos recintos y otras muchas que requieran visitar de manera autónoma las salas a las que se tenga acceso desde un pasillo o conjunto de pasillos interconectados. Se implementó dicho sistema tanto en simulación como en tiempo real, obteniendo para ambos casos los resultados esperados al navegar por un recinto.
2016 IEEE Colombian Conference on Robotics and Automation (CCRA), 2016
Musculoskeletal Models are useful for designing rehabilitation technology (RT). MSMS software has... more Musculoskeletal Models are useful for designing rehabilitation technology (RT). MSMS software has unique features which let simulate RT along with different controllers. Unfortunately there are technical differences across platforms which yield to different results. Therefore benchmarking of simulations has been proposed. Here we present the implementation of a Shoulder and Elbow musculosketal (SE) model based on the MoBL ARMS model. SE model consists of 5 rigid segments linked with 5 degrees of freedom driven by 12 muscles. Muscle fascicle length (FL) and active fiber force (AFF) were calculated during simulation of 2 single joint shoulder and elbow movements. RMS errors were calculated between results. Differences between models were smallest for muscle FL (RMSE≤0.0037) and AFF (RMSE≤22.0155). RMS errors were biggest for FL of deltoid posterior muscle (RMSE=0.1165) and for AFF of supraspinatus muscle (RMSE=164.97) during shoulder abduction. But both models showed similar outputs.
A NEW REFORMULATION OF PHILIPS ALGORITHM FOR THE COMPUTATION OF DISCRETE IMAGE MOMENTS IS PRESENT... more A NEW REFORMULATION OF PHILIPS ALGORITHM FOR THE COMPUTATION OF DISCRETE IMAGE MOMENTS IS PRESENTED. AS PHILIPS METHOD, THE REFORMULATION PRODUCES THE SAME EXACT RESULTS BUT IN A FASTER MANNER. MOMENT CONTRIBUTIONS DUE TO THE PRESENCE OF HOLES IN THE SDHAPE /WHICH ARE NOT TAKEN INTO ACCOUNT, TO OUR KNOWLEDGE, INTO ANY BOWDARY BASED METHOD FOR MOMENT CALCULATION INCLUIDING PHILIPS) ARE ALSO INTRODUCED INTO THE NEW METHOD.
Uploads
Papers by Juan Manuel Ibarra Zannatha