Gradient Descent with Momentum Based Backpropagation Neural Network for Selection of Industrial Robot

Published on Jan 1, 2016
· DOI :10.1007/978-3-319-30933-0_49
Sasmita Nayak1
Estimated H-index: 1
(Government College),
B.B. Choudhury6
Estimated H-index: 6
(Indira Gandhi Institute of Technology),
Saroj Kumar Lenka13
Estimated H-index: 13
Source
Abstract
Fast development of industrial robots and its utilization by the manufacturing industries for many different applications is a critical task for the selection of robots. As a consequence, the selection process of the robot becomes very much complicated for the potential users because they have an extensive set of parameters of the available robots. In this paper, gradient descent momentum optimization algorithm is used with backpropagation neural network prediction technique for the selection of industrial robots. Through this proposed technique maximum, ten parameters are directly considered as an input for the selection process of robot where as up to seven robot parameter data be used in the existing methods. The rank of the preferred industrial robot evaluates from the perfectly the best probable robot that specifies the most genuine benchmark of robot selection for the particular application using the proposed algorithm. Moreover, the performance of the algorithms for the robot selection is analyzed using Mean Square Error (MSE), R-squared error (RSE), and Root Mean Square Error (RMSE).
📖 Papers frequently viewed together
1999
3 Authors (Jizhong Xiao, ..., Danwei Wang)
5 Citations
1 Citations
References9
Newest
15 CitationsSource
Jun 27, 2011 in ICSECS (International Conference on Software Engineering and Computer Systems)
#1M. Z. Rehman (UTHM: Universiti Tun Hussein Onn Malaysia)H-Index: 6
#2Nazri Mohd Nawi (UTHM: Universiti Tun Hussein Onn Malaysia)H-Index: 16
The traditional Gradient Descent Back-propagation Neural Network Algorithm is widely used in solving many practical applications around the globe. Despite providing successful solutions, it possesses a problem of slow convergence and sometimes getting stuck at local minima. Several modifications are suggested to improve the convergence rate of Gradient Descent Backpropagation algorithm such as careful selection of initial weights and biases, learning rate, momentum, network topology, activation ...
29 CitationsSource
#1R. Venkata Rao (Shri Mata Vaishno Devi University)H-Index: 39
#1R. Venkata Rao (Shri Mata Vaishno Devi University)H-Index: 1
Last. K. K. Padmanabhan (Kumaraguru College of Technology)H-Index: 3
view all 2 authors...
Abstract In the present work, a methodology based on digraph and matrix methods is developed for evaluation of alternative industrial robots. A robot selection index is proposed that evaluates and ranks robots for a given industrial application. The index is obtained from a robot selection attributes function, obtained from the robot selection attributes digraph. The digraph is developed considering robot selection attributes and their relative importance for the application considered. A step b...
123 CitationsSource
#1T.-C. Chu (NTU: National Taiwan University)H-Index: 3
#2Yi-Jyun Lin (NTU: National Taiwan University)H-Index: 1
A fuzzy TOPSIS method for robot selection is proposed, where the ratings of various alternatives versus various subjective criteria and the weights of all criteria are assessed in linguistic terms represented by fuzzy numbers. The values of objective criteria are converted into dimensionless indices to ensure compatibility between the values of objective criteria and the linguistic ratings of subjective criteria. The membership function of each weighted rating is developed by interval arithmetic...
254 CitationsSource
#1Celik Parkan (CityU: City University of Hong Kong)H-Index: 8
#2Ming-Lu Wu (CityU: City University of Hong Kong)H-Index: 17
Abstract Many advanced manufacturers use robots extensively to perform repetitious, difficult, and hazardous tasks with precision. Robots improve quality and productivity dramatically if deployed properly. The process of selecting the most suitable robot among many alternatives involves robots' performance in a number of key areas. Various quantitative methods have been proposed as an aid to selection decision on the choice of robots. This paper demonstrates the use of and compares some of the c...
195 CitationsSource
#1Moutaz Khouja (UNCC: University of North Carolina at Charlotte)H-Index: 33
Abstract The range of manufacturing technologies available to firms has significantly increased in the past 20 years. A potential buyer of a technology such as machine tools, industrial robots, or flexible manufacturing systems is faced with many options in both performance and cost. This paper proposes a decision model for technology selection problems using a two-phase procedure. In phase 1, data envelopment analysis is used to identify technologies that provide the best combinations of vendor...
199 CitationsSource
#1Gin-Shuh Liang (NTHU: National Tsing Hua University)H-Index: 24
#2Mao-Jiun J. Wang (NTHU: National Tsing Hua University)H-Index: 43
Abstract From combing the concepts of fuzzy set theory and hierarchical structure analysis, a robot selection algorithm is proposed. It is used to aggregate decision-makers' fuzzy assessments about criteria weightings and the suitability ratings of a robot vs various selection criteria to obtain fuzzy suitability indices. The suitability ratings are then ranked to select the most suitable robot. By utilizing this decision algorithm, the decision-makers' fuzzy assessments with various rating atti...
122 CitationsSource
#1V. P. Agrawal (IITD: Indian Institute of Technology Delhi)H-Index: 5
#2V. Kohli (IITD: Indian Institute of Technology Delhi)H-Index: 1
Last. Sushil Gupta (IITD: Indian Institute of Technology Delhi)H-Index: 29
view all 3 authors...
Abstract This paper presents an efficient approach for a computer-based solution to the problem of selection of an ‘optimum robot’ specifically to aid industries. The approach breaks new ground by applying Multiple Attribute Decision Making (MADM) ideology to equipment selection (robots) which is of consequence because selection is an implicit aspect of design. Indeed, the approach is also applicable to a robot design problem. The robot selection procedure allows rapid convergence from a very la...
147 CitationsSource
#1Richard P. Paul (Purdue University)H-Index: 29
#1Richard P. Paul (Purdue University)H-Index: 12
Last. Shimon Y. Nof (Purdue University)H-Index: 35
view all 2 authors...
SUMMARY The industrial robot, already an accepted element in modern manufacturing, when coupled with sensors and controlled by a computer, is better modelled as an operator than as a machine. In this paper the basic characteristics of sensor controlled robots are defined : workspace, actuation, positioning, orientation, sensing and decision making. Actual values of these characteristics are given for the Stanford Arm running under the WAVE system. In order to compare a robot to a human operator ...
59 CitationsSource
Cited By3
Newest
#1Untari Novia Wisesty (Telkom University)H-Index: 6
#2Febryanti Sthevanie (Telkom University)H-Index: 3
Last. Rita Rismala (Telkom University)H-Index: 5
view all 3 authors...
Early detection of cancer can increase the success of treatment in patients with cancer. In the latest research, cancer can be detected through DNA Microarrays. Someone who suffers from cancer will experience changes in the value of certain gene expression. In previous studies, the Genetic Algorithm as a feature selection method and the Momentum Backpropagation algorithm as a classification method provide a fairly high classification performance, but the Momentum Backpropagation algorithm still ...
Source
#1Sarfaraz Masood (Jamia Millia Islamia)H-Index: 7
#2M. N. Doja (Jamia Millia Islamia)H-Index: 5
Last. Pravin Chandra (IP: Guru Gobind Singh Indraprastha University)H-Index: 15
view all 3 authors...
The selection of the initial network weights has been a known key aspect affecting the convergence of sigmoidal activation function-based artificial neural networks. In this paper, a new network initialization scheme has been proposed that initializes the network weights such that activation functions in the network are not saturated initially. The proposed method ensures that the initial outputs of the hidden neurons are in the active region which positively impacts the network’s rate of conver...
Source
1 CitationsSource