scholarly journals Artificial Intelligence-Based Optimal Grasping Control

Sensors ◽  
2020 ◽  
Vol 20 (21) ◽  
pp. 6390 ◽  
Author(s):  
Dongeon Kim ◽  
Jonghak Lee ◽  
Wan-Young Chung ◽  
Jangmyung Lee
Keyword(s):  
Neural Network ◽  
Contact Force ◽  
Arrival Time ◽  
Contact Point ◽  
Pressure Sensors ◽  
Air Pressure ◽  
Sensor Data ◽  
Robot Hand ◽  
Force Estimation ◽  
Grasping Force

A new tactile sensing module was proposed to sense the contact force and location of an object on a robot hand, which was attached on the robot finger. Three air pressure sensors are installed at the tip of the finger to detect the contacting force at the points. To obtain a nominal contact force at the finger from data from the three air pressure sensors, a force estimation was developed based upon the learning of a deep neural network. The data from the three air pressure sensors were utilized as inputs to estimate the contact force at the finger. In the tactile module, the arrival time of the air pressure sensor data has been utilized to recognize the contact point of the robot finger against an object. Using the three air pressure sensors and arrival time, the finger location can be divided into 3 × 3 block locations. The resolution of the contact point recognition was improved to 6 × 4 block locations on the finger using an artificial neural network. The accuracy and effectiveness of the tactile module were verified using real grasping experiments. With this stable grasping, an optimal grasping force was estimated empirically with fuzzy rules for a given object.

scholarly journals A learning-based tip contact force estimation method for tendon-driven continuum manipulator

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Fan Feng ◽  
Wuzhou Hong ◽  
Le Xie
Keyword(s):  
Neural Network ◽  
Contact Force ◽  
Recurrent Neural Network ◽  
Estimation Method ◽  
Training Data ◽  
Force Sensing ◽  
Force Estimation ◽  
Advantages And Disadvantages ◽  
Continuum Manipulator ◽  
Three Degree Of Freedom

AbstractAlthough tendon-driven continuum manipulators have been extensively researched, how to realize tip contact force sensing in a more general and efficient way without increasing the diameter is still a challenge. Rather than use a complex modeling approach, this paper proposes a general tip contact force-sensing method based on a recurrent neural network that takes the tendons’ position and tension as the input of a recurrent neural network and the tip contact force of the continuum manipulator as the output and fits this static model by means of machine learning so that it may be used as a real-time contact force estimator. We also designed and built a corresponding three-degree-of-freedom contact force data acquisition platform based on the structure of a continuum manipulator designed in our previous studies. After obtaining training data, we built and compared the performances of a multi-layer perceptron-based contact force estimator as a baseline and three typical recurrent neural network-based contact force estimators through TensorFlow framework to verify the feasibility of this method. We also proposed a manually decoupled sub-estimators algorithm and evaluated the advantages and disadvantages of those two methods.

Application of Neural Network to Teaching of Massage using Multi-fingered Robot Hand

2002 ◽  
Vol 14 (2) ◽  
pp. 162-169 ◽  
Author(s):  
Hideo Kitagawa ◽  
◽  
Tomomi Terai ◽  
Panya Minyong ◽  
Kazuhiko. Terashima ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Motion Control ◽  
Humanoid Robot ◽  
Pressure Sensors ◽  
Robot Hand ◽  
Human Beings

A massage expert using a humanoid robot hand is constructed, and motion control using neural networks (NNs), teaching of fingertip pressure in massage of the robot hand, and its reproduction of force are reported. Force exerted when an expert massages shoulders was detected using distribution pressure sensors, taught to the robot hand, and reproduced by it. It massaged with pressure safe for human beings and similar to an expert.

Nonintrusive Measurement of Contact Forces During Vibration Dominated Impacts

2004 ◽  
Vol 126 (3) ◽  
pp. 489-497 ◽  
Author(s):  
Satwinder Jit Singh ◽  
Anindya Chatterjee
Keyword(s):  
Contact Force ◽  
Tikhonov Regularization ◽  
Strain Gauge ◽  
Initial Conditions ◽  
Contact Point ◽  
Electrical Contact ◽  
Contact Forces ◽  
Force Estimation ◽  
Contact Duration ◽  
Gauge Data

Impact force estimation is done indirectly through, e.g., strain measurements away from the contact point, because inserting a force transducer between the contacting objects changes the force. Most prior contact force measurements involved a single contact interval. Here we study transverse impacts of a slender beam and a clamped-free plate; contact occurs more than once within one impact. Strain gauge data, electrical contact detection, and a dynamic model of the beam are used to estimate the contact force. The problem of force estimation from strain gauge data is ill-posed, and Tikhonov regularization fails initially. A reduced-order model is then developed using symmetry, and better initial conditions are estimated using a Kalman filter. Subsequently, Tikhonov regularization gives excellent force estimates, empirically supported by the contact duration measurements. Two other methods that explicitly use the contact duration measurements are also given. The first uses Tikhonov regularization within each contact interval, followed by Kalman filtering during noncontact to get initial conditions for the next contact. The second uses truncated Fourier sine series in each contact interval and is, computationally, the simplest. All three methods provide consistent force estimates. Our work complements recent work by Inoue and coworkers where the impulse response of the colliding object was measured separately using a Hopkinson bar, and electrical contact was not monitored.

314 External force estimation for robot hand without grasping-force sensor

2005 ◽  
Vol 2005.42 (0) ◽  
pp. 97-98
Author(s):  
Takeshi TORITA ◽  
Ryoichi SUZUKI ◽  
Nobuaki KOBAYASHI
Keyword(s):  
External Force ◽  
Force Sensor ◽  
Robot Hand ◽  
Force Estimation ◽  
Grasping Force

scholarly journals Neural Network Based Contact Force Control Algorithm for Walking Robots

Sensors ◽  
2021 ◽  
Vol 21 (1) ◽  
pp. 287
Author(s):  
Byeongjin Kim ◽  
Soohyun Kim
Keyword(s):  
Neural Network ◽  
Contact Force ◽  
Force Control ◽  
Control Algorithm ◽  
Position Control ◽  
Linear Quadratic Regulator ◽  
Walking Robots ◽  
Test Model ◽  
Linear Quadratic ◽  
Contact Force Control

Walking algorithms using push-off improve moving efficiency and disturbance rejection performance. However, the algorithm based on classical contact force control requires an exact model or a Force/Torque sensor. This paper proposes a novel contact force control algorithm based on neural networks. The proposed model is adapted to a linear quadratic regulator for position control and balance. The results demonstrate that this neural network-based model can accurately generate force and effectively reduce errors without requiring a sensor. The effectiveness of the algorithm is assessed with the realistic test model. Compared to the Jacobian-based calculation, our algorithm significantly improves the accuracy of the force control. One step simulation was used to analyze the robustness of the algorithm. In summary, this walking control algorithm generates a push-off force with precision and enables it to reject disturbance rapidly.

scholarly journals Remaining Useful Life Prediction Using Temporal Convolution with Attention

AI ◽  
2021 ◽  
Vol 2 (1) ◽  
pp. 48-70
Author(s):  
Wei Ming Tan ◽  
T. Hui Teo
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Time Series Data ◽  
Remaining Useful Life ◽  
Sensor Data ◽  
Series Data ◽  
Multiple Time ◽  
Data Set ◽  
Form Complex ◽  
Useful Life

Prognostic techniques attempt to predict the Remaining Useful Life (RUL) of a subsystem or a component. Such techniques often use sensor data which are periodically measured and recorded into a time series data set. Such multivariate data sets form complex and non-linear inter-dependencies through recorded time steps and between sensors. Many current existing algorithms for prognostic purposes starts to explore Deep Neural Network (DNN) and its effectiveness in the field. Although Deep Learning (DL) techniques outperform the traditional prognostic algorithms, the networks are generally complex to deploy or train. This paper proposes a Multi-variable Time Series (MTS) focused approach to prognostics that implements a lightweight Convolutional Neural Network (CNN) with attention mechanism. The convolution filters work to extract the abstract temporal patterns from the multiple time series, while the attention mechanisms review the information across the time axis and select the relevant information. The results suggest that the proposed method not only produces a superior accuracy of RUL estimation but it also trains many folds faster than the reported works. The superiority of deploying the network is also demonstrated on a lightweight hardware platform by not just being much compact, but also more efficient for the resource restricted environment.

Sign in / Sign up

Export Citation Format

Share Document