Neural Network Based Transfer Learning for Robot Path Generation

Robot Path

Abstract In this paper, an artificial neural network (ANN) based transfer learning approach of inverse displacement analysis of robot manipulators is studied. ANNs with different structures are applied utilizing data from different end effector paths of a manipulator for training purposes. Four transfer learning methods are proposed by applying pretrained initial parameters. Final training results of ANN with transfer learning are compared with those of ANN with random initialization. To investigate the rate of convergence of data fitting comprehensively, different values of performance targets are defined. The computing epochs and performance measures are compared. It is presented that, depending on the structure of ANN, the proposed transfer learning methods can accelerate the training process and achieve higher accuracy. Depending on the method, the transfer learning improves the performance differently.

Neural Network Based Transfer Learning of Manipulator Inverse Displacement Analysis

Volume 10: 44th Mechanisms and Robotics Conference (MR) ◽

10.1115/detc2020-22508 ◽

2020 ◽

Author(s):

Houcheng Tang ◽

Leila Notash

Keyword(s):

Neural Network ◽

Neural Networks ◽

Transfer Learning ◽

Learning Approach ◽

Training Process ◽

Displacement Analysis ◽

Performance Targets ◽

The Neural Networks ◽

Random Initialization ◽

And Performance

Abstract In this paper, a neural network based transfer learning approach of inverse displacement analysis of robot manipulators is studied. Neural networks with different structures are applied utilizing data from different configurations of a manipulator for training purposes. Then the transfer learning was conducted between manipulators with different geometric layouts. The training is performed on both the neural networks with pretrained initial parameters and the neural networks with random initialization. To investigate the rate of convergence of data fitting comprehensively, different values of performance targets are defined. The computing epochs and performance measures are compared. It is presented that, depending on the structure of neural network, the proposed transfer learning can accelerate the training process and achieve higher accuracy. For different datasets, the transfer learning approach improves their performance differently.

Reusing Source Task Knowledge via Transfer Approximator in Reinforcement Transfer Learning

Symmetry ◽

10.3390/sym11010025 ◽

2018 ◽

Vol 11 (1) ◽

pp. 25 ◽

Cited By ~ 1

Author(s):

Qiao Cheng ◽

Xiangke Wang ◽

Yifeng Niu ◽

Lincheng Shen

Keyword(s):

Neural Network ◽

Transfer Learning ◽

Mapping Rule ◽

Learning Methods ◽

Target Task ◽

Artificial Neural ◽

The One

Transfer Learning (TL) has received a great deal of attention because of its ability to speed up Reinforcement Learning (RL) by reusing learned knowledge from other tasks. This paper proposes a new transfer learning framework, referred to as Transfer Learning via Artificial Neural Network Approximator (TL-ANNA). It builds an Artificial Neural Network (ANN) transfer approximator to transfer the related knowledge from the source task into the target task and reuses the transferred knowledge with a Probabilistic Policy Reuse (PPR) scheme. Specifically, the transfer approximator maps the state of the target task symmetrically to states of the source task with a certain mapping rule, and activates the related knowledge (components of the action-value function) of the source task as the input of the ANNs; it then predicts the quality of the actions in the target task with the ANNs. The target learner uses the PPR scheme to bias the RL with the suggested action from the transfer approximator. In this way, the transfer approximator builds a symmetric knowledge path between the target task and the source task. In addition, two mapping rules for the transfer approximator are designed, namely, Full Mapping Rule and Group Mapping Rule. Experiments performed on the RoboCup soccer Keepaway task verified that the proposed transfer learning methods outperform two other transfer learning methods in both jumpstart and time to threshold metrics and are more robust to the quality of source knowledge. In addition, the TL-ANNA with the group mapping rule exhibits slightly worse performance than the one with the full mapping rule, but with less computation and space cost when appropriate grouping method is used.

Neural Network-Based Transfer Learning of Manipulator Inverse Displacement Analysis

Journal of Mechanisms and Robotics ◽

10.1115/1.4050622 ◽

2021 ◽

Vol 13 (3) ◽

Author(s):

Houcheng Tang ◽

Leila Notash

Keyword(s):

Neural Network ◽

Neural Networks ◽

Transfer Learning ◽

Robot Manipulators ◽

Learning Approach ◽

Displacement Analysis ◽

Performance Targets ◽

The Neural Network ◽

The Neural Networks ◽

And Performance

Abstract In this paper, the feasibility of applying transfer learning for modeling robot manipulators is examined. A neural network-based transfer learning approach of inverse displacement analysis of robot manipulators is studied. Neural networks with different structures are applied utilizing data from different configurations of a manipulator for training purposes. Then, the transfer learning was conducted between manipulators with different geometric layouts. The training is performed on both the neural networks with pretrained initial parameters and the neural networks with random initialization. To investigate the rate of convergence of data fitting comprehensively, different values of performance targets are defined. The computing epochs and performance measures are compared. It is presented that, depending on the structure of the neural network, the proposed transfer learning can accelerate the training process and achieve higher accuracy. For different datasets, the transfer learning approach improves their performance differently.

Optimization of operating parameters and performance evaluation of forced draft cooling tower using response surface methodology (RSM) and artificial neural network (ANN)

Journal of Mechanical Science and Technology ◽

10.1007/s12206-012-0323-9 ◽

2012 ◽

Vol 26 (5) ◽

pp. 1643-1650 ◽

Cited By ~ 14

Author(s):

Ramkumar Ramakrishnan ◽

Ragupathy Arumugam

Keyword(s):

Neural Network ◽

Response Surface Methodology ◽

Performance Evaluation ◽

Response Surface ◽

Cooling Tower ◽

Operating Parameters ◽

Forced Draft ◽

And Performance ◽

Artificial Neural Network Ann

Modeling of daily runoff from a small agricultural watershed using artificial neural network with resampling techniques

Journal of Hydroinformatics ◽

10.2166/hydro.2014.106 ◽

2014 ◽

Vol 17 (1) ◽

pp. 56-74 ◽

Cited By ~ 11

Author(s):

Gurjeet Singh ◽

Rabindra K. Panda ◽

Marc Lamers

Keyword(s):

Neural Network ◽

Agricultural Watershed ◽

Training Data ◽

Optimum Number ◽

Training Process ◽

Data Set ◽

Reported Study ◽

Artificial Neural ◽

Artificial Neural Network Ann

The reported study was undertaken in a small agricultural watershed, namely, Kapgari in Eastern India having a drainage area of 973 ha. The watershed was subdivided into three sub-watersheds on the basis of drainage network and land topography. An attempt was made to relate the continuously monitored runoff data from the sub-watersheds and the whole-watershed with the rainfall and temperature data using the artificial neural network (ANN) technique. The reported study also evaluated the bias in the prediction of daily runoff with shorter length of training data set using different resampling techniques with the ANN modeling. A 10-fold cross-validation (CV) technique was used to find the optimum number of hidden neurons in the hidden layer and to avoid neural network over-fitting during the training process for shorter length of data. The results illustrated that the ANN models developed with shorter length of training data set avoid neural network over-fitting during the training process, using a 10-fold CV method. Moreover, the biasness was investigated using the bootstrap resampling technique based ANN (BANN) for short length of training data set. In comparison with the 10-fold CV technique, the BANN is more efficient in solving the problems of the over-fitting and under-fitting during training of models for shorter length of data set.

Stable and accurate a posteriori LES of 2D turbulence with convolutional neural networks: Backscatter analysis and generalization via transfer learning

10.5194/egusphere-egu21-402 ◽

2021 ◽

Author(s):

Yifei Guan ◽

Ashesh Chattopadhyay ◽

Adam Subel ◽

Pedram Hassanzadeh

Keyword(s):

Neural Network ◽

Transfer Learning ◽

Network Architecture ◽

Turbulence Modeling ◽

Scale Effects ◽

A Priori ◽

A Posteriori ◽

Subgrid Scale ◽

Short Term Prediction

<p>In large eddy simulations (LES), the subgrid-scale effects are modeled by physics-based or data-driven methods. This work develops a convolutional neural network (CNN) to model the subgrid-scale effects of a two-dimensional turbulent flow. The model is able to capture both the inter-scale forward energy transfer and backscatter in both a priori and a posteriori analyses. The LES-CNN model outperforms the physics-based eddy-viscosity models and the previous proposed local artificial neural network (ANN) models in both short-term prediction and long-term statistics. Transfer learning is implemented to generalize the method for turbulence modeling at higher Reynolds numbers. Encoder-decoder network architecture is proposed to generalize the model to a higher computational grid resolution.</p>

Neuron Blockchain Algorithm for Legal Problems in Inheritance of Legacy

Electronics ◽

10.3390/electronics9101595 ◽

2020 ◽

Vol 9 (10) ◽

pp. 1595

Author(s):

Seong-Kyu Kim ◽

Jun-Ho Huh

Keyword(s):

Neural Network ◽

Deep Neural Network ◽

Non Fiction ◽

Legal Problems ◽

Artificial Neurons ◽

The Will ◽

Artificial Neural ◽

And Performance ◽

Artificial Network

This paper discusses the worldwide trend of aging as the lifespan of humans increases. Nonetheless, most people do not write wills, which results in many legal problems after their death. There are many reasons for this including the problem of the validity of their heritage possibly not being legally certified. Wills can be divided into two categories, i.e., testimony and documents. A lawyer in the middle should notarize them, however, instead of providing these notarized services, we propose more transparent algorithms, blockchain shading, and smart country functions. Architectures are designed based on a neural network, the blockchain deep neural network (DNN), and deep neural network-based units are built with a necessary artificial neural network (ANN) base. A heritage inherited blockchain architecture is designed to communicate between nodes based on the minimum distance algorithm and multichannel protocol. In addition, neurons refer to the nerve cells that make up the nervous system of an organism, and artificial neurons are an abstraction of the functions of dendrite, soma, and axon that constitute the neurons of an organism. Similar to the neurons in organisms, artificial neural algorithms such as the depth-first search (DFS) algorithm are expressed in pseudocode. In addition, all blockchain nodes are equipped with verified nodes. A research model is proposed for an artificial network blockchain that is needed for this purpose. The experimental environment builds the server and network environments based on deep neural networks that require verification. Weights are also set for the required verification and performance. This paper verifies the blockchain algorithm equipped with this non-fiction preprocessor function. We also study the blockchain neuron engine that can safely construct a block node for a suicide blockchain. After empirical testing of the will system with artificial intelligence and blockchain, the values are close to 2 and 10 and the distribution is good. The blockchain node also tested 50 nodes more than 150 times, and we concluded that it was suitable for actual testing by completing a demonstration test with 4500 TPS.

Applications of Artificial Neural Network for the Prediction of Pool Boiling Curves

10th International Conference on Nuclear Engineering, Volume 3 ◽

10.1115/icone10-22487 ◽

2002 ◽

Cited By ~ 3

Author(s):

Guanghui Su ◽

K. Fukuda ◽

K. Morita

Keyword(s):

Neural Network ◽

Pool Boiling ◽

Convergence Criterion ◽

Training Process ◽

Transient Heating ◽

Wall Superheat ◽

Artificial Neural ◽

Best Fit

Artificial neural network (ANN) has the advantage that the best-fit correlations of experimental data will no longer be necessary for predicting unknowns from the known parameters. The ANN was applied to predict the pool boiling curves in this paper. The database of experimentel data presented by Berenson, Dhuga et al., and Bui and Dhir etc. were used in the analysis. The database is subdivided in two subsets. The first subset is used to train the network and the second one is used to test the network after the training process. The input parameters of the ANN are: wall superheat ΔTw, surface roughness, steady/transient heating/transient cooling, subcooling, Surface inclination and pressure. The output parameter is heat flux q. The proposed methodology allows us to achieve the accuracy that satisfies the user’s convergence criterion and it is suitable for pool boiling curve data processing.

MODELING METHOD USING COMBINED ARTIFICIAL NEURAL NETWORK

International Journal of Computational Intelligence and Applications ◽

10.1142/s1469026811003057 ◽

2011 ◽

Vol 10 (02) ◽

pp. 189-198

Author(s):

YANGPO SONG ◽

XIAOQI PENG

Keyword(s):

Neural Network ◽

Traditional Method ◽

Modeling Method ◽

Early Stopping ◽

Training Process ◽

Artificial Neural ◽

Optimal Modeling ◽

Error Simulation

To improve the modeling performance — such as accuracy and robustness — of artificial neural network (ANN), a new combined ANN and corresponding optimal modeling method are proposed in this paper. The combined ANN consists of two parallel sub-networks, and methods such as "early stopping" and "data resampling" are jointly used in training process to reduce the sensitivity of the modeling performance to its structure. To achieve better performance, the structure of combined ANN is proposed to be adjusted dynamically according to the information of expectation error and real error. Simulation experimental results verify that the optimal modeling method using combined ANN can achieve much better performance than the traditional method.

ARTIFICIAL NEURAL NETWORK BASED ONLINE SENSOR CALIBRATION AND COMPENSATION

International Journal of Computing ◽

10.47839/ijc.6.3.454 ◽

2014 ◽

pp. 74-78

Author(s):

Shakeb A. Khan ◽

Tarikul Islam ◽

Gulshan Husain

Keyword(s):

Neural Network ◽

Inverse Modeling ◽

Sensor Response ◽

Modeling Technique ◽

Sensor Calibration ◽

Online Sensor ◽

Artificial Neural ◽

And Performance ◽

Artificial Neural Network Ann

This paper presents an artificial neural network (ANN) based generalized online method for sensor response linearization and calibration. Inverse modeling technique is used for sensor response linearization. Multilayer ANN is used for inverse modeling of sensor. The inverse model based technique automatically compensates the associated nonlinearity and estimates the measurand. The scheme is coded in MATLAB® for offline training and for online measurement and successfully implemented using NI PCI-6221 Data Acquisition (DAQ) card and LabVIEW® software. Manufacturing tolerances, environmental effects, and performance drifts due to aging bring up a need for frequent calibration, this ANN based inverse modeling technique provides greater flexibility and accuracy under such conditions.