Viacheslav Khomenko | Université Versailles Saint-Quentin en Yvelines (original) (raw)

Uploads

Papers by Viacheslav Khomenko

Walking robots need precise control for legs articulations because it influence their equilibrium... more Walking robots need precise control for legs articulations because it influence their equilibrium. It is necessary to compensate vibrational effects caused by imperfections in their articulations, like elasticities, mechanical backlashes, frictions and structural deformations and those appearing during shocks with the ground or when forces become significant. Our approach consists in correcting the inputs of the robot control system in a robust way according to variations of functional conditions and robot parameters. To reach this objective, we use adaptive and learning control methods, nonlinear oscillators. The research problematic, objectives, and state of the art are presented in the chapter 1. As it’s necessary to know exact torques of a robot, we present in the chapter 2 an approach of polyarticulated multi-masses systems modeling that takes into account its control system, mechanical transmissions and nonlinearities of transmissions. Experimental validations are carried on the biped robot ROBIAN. The chapter 3 explains the instrumentation of the non-direct articulation accelerations measurement method based on distributed accelerometer measurements on the body of the robot. The chapter 4 concerns experimental validation of compensation and control methods for ROBIAN for different kinds of flexion/extension movements.

The paper displays main results of the Ukrainian-French research project which studies interactio... more The paper displays main results of the Ukrainian-French research project which studies interaction of a robot with its environment. Following applications of computer vision technology for various issues of robotics are shown: robot follows movements of a person, detects speed of the moving object, studies its environment, learns particular manipulations, gets trained to focus attention on the objects of interest and synchronizes its motion with an interacting agent. The article proposes approaches to development of the software interface for the robot control based on the computer vision technology.

The paper present results of the Ukrainian-French scientific activity between Donetsk National Te... more The paper present results of the Ukrainian-French scientific activity between Donetsk National Technical University and French partners in the field of human-robot and robot-environment interactions. The first part of the research shows adaptation of robotic arm movements to the dynamics of interacting person who imposes its own rhythmicity. The adaptation algorithm is based on rhythmic oscillator inspired by biology. The second part of the research considers adaptation of a biped robot internal vibration modes to the environment during flexion-extension vertical movements. These vibrations are extremely unwanted; they appear during robot motion by reason of internal flexibilities in articulations, backlashes and friction and can cause the fall of the robot. It is shown that their efficient attenuation can be achieved by using auto-adaptive oscillator that acts on the robot's knee articulations.

This paper presents experimental results on the learning of synchrony between human and a robot a... more This paper presents experimental results on the learning of synchrony between human and a robot arm when they interact in a handshaking situation. The controller is based on two coupled nonlinear neurons (central pattern generator, CPG) that are able to learn to oscillate at different frequencies depending on a biological inspired learning mechanism. This mechanism uses interaction force measured in the robot joint as a learning signal. Results show that the CPG learns and follows the human rhythm easily and then enjoy it.

masters.donntu.edu.ua

... Vol. 46, No.1. – 2009. – P. 59-73. 3. B. Hamed. Application of a LabVIEW for Real-Time Contro... more ... Vol. 46, No.1. – 2009. – P. 59-73. 3. B. Hamed. Application of a LabVIEW for Real-Time Control of Ball and Beam System // IACSIT International Journal of Engineering and Technology. – Vol.2, No.4. – 2010. – P. 401-407. ... Vol. 1. – P. 306 – 310. 7. NB Almutairi, M. Zribi. ...

Annotation -Generalized model of articulated electromechanical system with elasticites in transmi... more Annotation -Generalized model of articulated electromechanical system with elasticites in transmissions is presented in these materials.

Journal of Mining Science, 2014

ABSTRACT The authors analyze excitability of polyharmonic vibrations in a single-body vibration m... more ABSTRACT The authors analyze excitability of polyharmonic vibrations in a single-body vibration machine. The developed mathematical model of the vibration system accounts for an elastic component element included in the design of the unbalance vibration exciter drive. The operating limits, frequency content and effect of the main design factors on the flow data of the vibration machine are examined. It is found that superharmonic vibration largely contributes to the polyharmonic spectrum at certain frequencies, depending on the stiffness of the elastic component element of the clutch connecting the vibration exciter and the rotary drive.

Human body has nonlinear dynamics and many degrees of freedom. In this study, we propose a wearab... more Human body has nonlinear dynamics and many degrees of freedom. In this study, we propose a wearable sensors network for quantitative measurements of anthropomorphic robots and human movements. It has been shown that the proposed network of acceleration sensors and force sensors is able to capture main characteristic of a body during the fundamental physical interaction activities of a human being. Thus, jumping, vertical flexion-extension, walking and handshaking were studied in our experiments. Wearable sensors network is combined with the Kinect sensor by Microsoft to obtain the bodies segments positions and orientations, articulation angles. Experimental results show the effectiveness of the idea of combining Kinect and sensor systems. Accelerometer sensor, Microsoft Kinect, force sensitive resistor, handshake, jump, walk, flexion-extension movement, anthropomorphic robot, human

This paper proposes a low-cost system, based on the method of Electrical Impedance Tomography (EI... more This paper proposes a low-cost system, based on the method of Electrical
Impedance Tomography (EIT), for data acquisition from soft conductive fabric, for the
purposes of designing of robots artificial skin. A simple multiplexer/demultiplexer circuit is
used for retrieving the resistance field from the pair-wised electrodes which inject the
electrical current and the electrodes which measure the output voltage from the conductive
fabric. A microcontroller governs the injection of current, voltage output patterns and the ...

Handshaking is an important component of social interaction between people in many cultures. Thus... more Handshaking is an important component of social interaction between people in many cultures. Thus, for further applications in human/humanoid-robot interaction it is important to understand and measure the characteristics of a handshake during interaction between humans. In this paper, a new wearable sensor network to measure a handshake is described. It consists of a set of several sensors (accelerometers, gyroscopes and force sensors) attached to the glove, and of a microcontroller for signal acquisition and conditioning. The paper focuses on the applicability and qualitative analysis of the proposed architecture of sensors through several experiments of handshaking between two human subjects. The results show that the proposed system allows reproducible experiments to quantify handshake characteristics such as duration and strength of the grip, vigor and rhythmicity of a handshake.

A new non-invasive low cost method for accelerations measurement in biped locomotion mechanisms i... more A new non-invasive low cost method for accelerations measurement in biped locomotion mechanisms is presented. The aim is to measure all the joint accelerations of the locomotion system during the walking that is difficult because of the nonaccessibility of several joint axes. Our measurement approach is non-direct using two triaxial accelerometers aligned on the main axis of each body, that ensures measurement of five accelerations per joint, two angular and three linear. By considering negligible the vertical rotation of a leg in the absolute frame during walking, the vertical rotational acceleration of the hip is measurable by two accelerometers aligned along the sagittal axis. For tibia and thigh, the accelerometers are aligned along vertical axes. The calculation of accelerations in articulation references is achieved by solving a system of Newton mechanics equations. The theoretical statements were validated experimentally on the robot for vertical flexion-extension movement.

Talks by Viacheslav Khomenko

1. Recap: feed-forward artificial neural network 2) Temporal dependencies 3) Recurrent neural net... more 1) Recap: feed-forward artificial neural network
2. Temporal dependencies
3. Recurrent neural network architectures
4. RNN training
5. New RNN architectures
6. Practical considerations
7. Neural models for locomotion
8. Application of RNNs

articles by Viacheslav Khomenko

The paper provides a practical solution to a real-time text/shape differentiation problem for onl... more The paper provides a practical solution to a real-time text/shape differentiation problem for online handwriting input. The proposed structure of the classification system comprises stroke grouping and stroke classification blocks. A new set of features is derived that has low computational complexity. The method achieves 98.5 {\%} text/shape classification accuracy on a benchmark dataset. The proposed stroke grouping machine learning approach improves classification robustness in relation to different input styles. In contrast to the threshold-based techniques, this grouping adaptation enhances the overall discriminating accuracy of the text/shape recognition system by 11.3 {\%}. The solution improves system's response on a touch-screen device.

inproceedings by Viacheslav Khomenko

An efficient algorithm for recurrent neural network training is presented. The approach increases... more An efficient algorithm for recurrent neural network training is presented. The approach increases the training speed for tasks where a length of the input sequence may vary significantly. The proposed approach is based on the optimal batch bucketing by input sequence length and data parallelization on multiple graphical processing units. The baseline training performance without sequence bucketing is compared with the proposed solution for a different number of buckets. An example is given for the online handwriting recognition task using an LSTM recurrent neural network. The evaluation is performed in terms of the wall clock time, number of epochs, and validation loss value.

Walking robots need precise control for legs articulations because it influence their equilibrium... more Walking robots need precise control for legs articulations because it influence their equilibrium. It is necessary to compensate vibrational effects caused by imperfections in their articulations, like elasticities, mechanical backlashes, frictions and structural deformations and those appearing during shocks with the ground or when forces become significant. Our approach consists in correcting the inputs of the robot control system in a robust way according to variations of functional conditions and robot parameters. To reach this objective, we use adaptive and learning control methods, nonlinear oscillators. The research problematic, objectives, and state of the art are presented in the chapter 1. As it’s necessary to know exact torques of a robot, we present in the chapter 2 an approach of polyarticulated multi-masses systems modeling that takes into account its control system, mechanical transmissions and nonlinearities of transmissions. Experimental validations are carried on the biped robot ROBIAN. The chapter 3 explains the instrumentation of the non-direct articulation accelerations measurement method based on distributed accelerometer measurements on the body of the robot. The chapter 4 concerns experimental validation of compensation and control methods for ROBIAN for different kinds of flexion/extension movements.

The paper displays main results of the Ukrainian-French research project which studies interactio... more The paper displays main results of the Ukrainian-French research project which studies interaction of a robot with its environment. Following applications of computer vision technology for various issues of robotics are shown: robot follows movements of a person, detects speed of the moving object, studies its environment, learns particular manipulations, gets trained to focus attention on the objects of interest and synchronizes its motion with an interacting agent. The article proposes approaches to development of the software interface for the robot control based on the computer vision technology.

The paper present results of the Ukrainian-French scientific activity between Donetsk National Te... more The paper present results of the Ukrainian-French scientific activity between Donetsk National Technical University and French partners in the field of human-robot and robot-environment interactions. The first part of the research shows adaptation of robotic arm movements to the dynamics of interacting person who imposes its own rhythmicity. The adaptation algorithm is based on rhythmic oscillator inspired by biology. The second part of the research considers adaptation of a biped robot internal vibration modes to the environment during flexion-extension vertical movements. These vibrations are extremely unwanted; they appear during robot motion by reason of internal flexibilities in articulations, backlashes and friction and can cause the fall of the robot. It is shown that their efficient attenuation can be achieved by using auto-adaptive oscillator that acts on the robot's knee articulations.

This paper presents experimental results on the learning of synchrony between human and a robot a... more This paper presents experimental results on the learning of synchrony between human and a robot arm when they interact in a handshaking situation. The controller is based on two coupled nonlinear neurons (central pattern generator, CPG) that are able to learn to oscillate at different frequencies depending on a biological inspired learning mechanism. This mechanism uses interaction force measured in the robot joint as a learning signal. Results show that the CPG learns and follows the human rhythm easily and then enjoy it.

masters.donntu.edu.ua

... Vol. 46, No.1. – 2009. – P. 59-73. 3. B. Hamed. Application of a LabVIEW for Real-Time Contro... more ... Vol. 46, No.1. – 2009. – P. 59-73. 3. B. Hamed. Application of a LabVIEW for Real-Time Control of Ball and Beam System // IACSIT International Journal of Engineering and Technology. – Vol.2, No.4. – 2010. – P. 401-407. ... Vol. 1. – P. 306 – 310. 7. NB Almutairi, M. Zribi. ...

Annotation -Generalized model of articulated electromechanical system with elasticites in transmi... more Annotation -Generalized model of articulated electromechanical system with elasticites in transmissions is presented in these materials.

Journal of Mining Science, 2014

ABSTRACT The authors analyze excitability of polyharmonic vibrations in a single-body vibration m... more ABSTRACT The authors analyze excitability of polyharmonic vibrations in a single-body vibration machine. The developed mathematical model of the vibration system accounts for an elastic component element included in the design of the unbalance vibration exciter drive. The operating limits, frequency content and effect of the main design factors on the flow data of the vibration machine are examined. It is found that superharmonic vibration largely contributes to the polyharmonic spectrum at certain frequencies, depending on the stiffness of the elastic component element of the clutch connecting the vibration exciter and the rotary drive.

Human body has nonlinear dynamics and many degrees of freedom. In this study, we propose a wearab... more Human body has nonlinear dynamics and many degrees of freedom. In this study, we propose a wearable sensors network for quantitative measurements of anthropomorphic robots and human movements. It has been shown that the proposed network of acceleration sensors and force sensors is able to capture main characteristic of a body during the fundamental physical interaction activities of a human being. Thus, jumping, vertical flexion-extension, walking and handshaking were studied in our experiments. Wearable sensors network is combined with the Kinect sensor by Microsoft to obtain the bodies segments positions and orientations, articulation angles. Experimental results show the effectiveness of the idea of combining Kinect and sensor systems. Accelerometer sensor, Microsoft Kinect, force sensitive resistor, handshake, jump, walk, flexion-extension movement, anthropomorphic robot, human

This paper proposes a low-cost system, based on the method of Electrical Impedance Tomography (EI... more This paper proposes a low-cost system, based on the method of Electrical
Impedance Tomography (EIT), for data acquisition from soft conductive fabric, for the
purposes of designing of robots artificial skin. A simple multiplexer/demultiplexer circuit is
used for retrieving the resistance field from the pair-wised electrodes which inject the
electrical current and the electrodes which measure the output voltage from the conductive
fabric. A microcontroller governs the injection of current, voltage output patterns and the ...

Handshaking is an important component of social interaction between people in many cultures. Thus... more Handshaking is an important component of social interaction between people in many cultures. Thus, for further applications in human/humanoid-robot interaction it is important to understand and measure the characteristics of a handshake during interaction between humans. In this paper, a new wearable sensor network to measure a handshake is described. It consists of a set of several sensors (accelerometers, gyroscopes and force sensors) attached to the glove, and of a microcontroller for signal acquisition and conditioning. The paper focuses on the applicability and qualitative analysis of the proposed architecture of sensors through several experiments of handshaking between two human subjects. The results show that the proposed system allows reproducible experiments to quantify handshake characteristics such as duration and strength of the grip, vigor and rhythmicity of a handshake.

A new non-invasive low cost method for accelerations measurement in biped locomotion mechanisms i... more A new non-invasive low cost method for accelerations measurement in biped locomotion mechanisms is presented. The aim is to measure all the joint accelerations of the locomotion system during the walking that is difficult because of the nonaccessibility of several joint axes. Our measurement approach is non-direct using two triaxial accelerometers aligned on the main axis of each body, that ensures measurement of five accelerations per joint, two angular and three linear. By considering negligible the vertical rotation of a leg in the absolute frame during walking, the vertical rotational acceleration of the hip is measurable by two accelerometers aligned along the sagittal axis. For tibia and thigh, the accelerometers are aligned along vertical axes. The calculation of accelerations in articulation references is achieved by solving a system of Newton mechanics equations. The theoretical statements were validated experimentally on the robot for vertical flexion-extension movement.

1. Recap: feed-forward artificial neural network 2) Temporal dependencies 3) Recurrent neural net... more 1) Recap: feed-forward artificial neural network
2. Temporal dependencies
3. Recurrent neural network architectures
4. RNN training
5. New RNN architectures
6. Practical considerations
7. Neural models for locomotion
8. Application of RNNs

The paper provides a practical solution to a real-time text/shape differentiation problem for onl... more The paper provides a practical solution to a real-time text/shape differentiation problem for online handwriting input. The proposed structure of the classification system comprises stroke grouping and stroke classification blocks. A new set of features is derived that has low computational complexity. The method achieves 98.5 {\%} text/shape classification accuracy on a benchmark dataset. The proposed stroke grouping machine learning approach improves classification robustness in relation to different input styles. In contrast to the threshold-based techniques, this grouping adaptation enhances the overall discriminating accuracy of the text/shape recognition system by 11.3 {\%}. The solution improves system's response on a touch-screen device.

An efficient algorithm for recurrent neural network training is presented. The approach increases... more An efficient algorithm for recurrent neural network training is presented. The approach increases the training speed for tasks where a length of the input sequence may vary significantly. The proposed approach is based on the optimal batch bucketing by input sequence length and data parallelization on multiple graphical processing units. The baseline training performance without sequence bucketing is compared with the proposed solution for a different number of buckets. An example is given for the online handwriting recognition task using an LSTM recurrent neural network. The evaluation is performed in terms of the wall clock time, number of epochs, and validation loss value.