Professional Documents
Culture Documents
Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com Volume 2, Issue 12, December 2013 ISSN 2319 - 4847
Comparative Analysis amongst Prediction of Field Findings through Empirical Model and Optimized Neural Network Mathematical Model for Human Powered Flywheel Motor
Arati R. Lende1 and J. P. Modak2
1
Ex Assistant. Professor, DMIETR, Sawnagi (M), Wardha, MH, India 2 Emeritus Professor and Dean (R&D), PCE, Nagpur, MH, India
Abstract
Neural Network is one of the most emerging tools in fitting function, future forecasting, pattern recognition, etc. This article signifies the effective methodology of Neural Network modeling for prediction of field findings. J. P. Modak and their associates had already carried out lots of investigations over development of applications utilizing Human Powered Flywheel Motor as an energy source. The applications tried so far are mostly rural based such as brick making, low head water lifting, wood turning, wood strip cutting, electricity generation, etc. The productivity of the above mentioned applications had great affection towards rider thereby affecting quality and quality of production. The paper evaluates optimum artificial neural network (ANN) parameters for prediction of experimental findings accurately through sequential variation of each ANN parameter. The most favorable values of each parameter are selected and a mathematical model is extracted in the course of it. This document also compares the prediction amongst ANN based mathematical model and traditionally generated empirical model.
Figure 1 Schematics of Human Powered Flywheel Motor R= Rider M = mechanism (01-OA-B-02-01) BSC = Big Sprocket Chain Drive SSC = Small Sprocket Chain Drive GSR = Gear of Speed Rise PSR = Pinion of Speed Rise FW= Flywheel CH = Chain CS = Counter Shaft FS = Flywheel Shaft
Page 423
The observations recorded during the experimentation are as below Table 3: Experimental observations
Independent Variables Log (I/RT2) 1 23 50 82 141 200 -7.4270 -7.1792 -7.2694 -6.1549 -5.9717 -7.2694 Log (ME) 0.00 0.0662 0.0600 0 0 0 Log (G) 0.3010 0.0010 0.0010 0.301 0.301 0.0569 Dependent variable Log ( T) 3.6305 3.5570 3.5004 3.0767 2.8587 3.4107
1.3 Empirical Model The experimental Independent variables were reduced by evaluating dimensionless pi terms by Buckingham pi theorem and a mathematical equation was generated by traditional method to predict the experimental findings. The equation [12] is as shown. T = 1.288 ( I/RT2)-0.46 (ME)-0.87 (G)0.40
Page 424
Page 425
Page 426
Page 427
Page 428
Page 429
Page 430
Page 431
Page 432
Figure 37 Percentage errors in predication with learning Function learnh The training program T28 was selected on the basis of minimum percentage error occurred in prediction of evidences.
5. COMPARING PREDICTION OF EVIDENCES WITH ANN MATHEMATICAL MODEL AND EMPIRICAL MODEL
The figures shown below give comparative analysis amongst prediction of evidences with ANN mathematical model and previously drawn empirical model.
Figure 38 Comparison between experimental evidences (red) and prediction through empirical model (blue)
Page 433
Figure 39 Comparison between experimental evidences (red) and prediction through ANN mathematical model (blue)
Figure 40 Percentage error in prediction through ANN mathematical model (red) and Empirical equation (blue) The figure 38 shows the prediction of experimental evidences (red) with previously drawn empirical model (blue) which is much deflected compared to the experimental findings. On the contrary artificial neural network based mathematical model show much better results as in figure 39. The figure 40 compares percentage error in prediction at every stage of the experimentation which also implies that neural prediction (red) is much better than the empirical model (blue).
6. CONCLUSION
The paper carries a systematic method of optimization of artificial neural network model and comes out with a audacious solution for prediction of experimental findings. The plots carried for each variation of ANN parameter clearly signifies its effect on prediction of the model. The mathematical model deduced may be utilized for future research in development of a controller for Human Powered Flywheel Motor.
References
[1] Modak J. P. and Askhedkar R. D. Hypothesis for the extrusion of lime flash sand brick using a manually driven Brick making machine, Building Research and Information U.K., V22,NI, Pp 47-54, 1994 [2] Modak J. P. and Bapat A. R. Manually driven flywheel motor operates wood turning machine, Contempory Ergonomics, Proc. Ergonomics Society annual convension13-16April, Edinburg, Scotland, Pp 352-357, 1993. [3] Sohoni V. V., Aware H. V. and Modak J. P. Manual Manufacture of Keyed Bricks, Building Research and Information UK, Vol 25, N6, 1997, 354-364. [4] Modak J. P.Design and Development of Manually Energized Process Machines having Relevance to Village/Agriculture and other productive operations, Application of manually energized flywheel motor for cutting of wood strip, Human Power, send for Publications. [5] H. Schenck Junior Theory of Engineering Experimentation, MC Graw Hill, New York. [6] A. R. Lende, Modelling of pedal driven flywheel motor by use of ANN, M. Tech. Thesis, PCE, Nagpur [7] S. N. Shvanandam, Introduction to Neural Network using Matlab 6.0, McGraw Hill publisher. [8] Stamtios V. Kartaplopoulos , Understanding Neural Networks and Fuzzy Logics, IEEE Press [9] Neural Network Toolbox TM 7 Users Guide R2010a, Mathworks.com [10] Rudra Pratap, Getting Started with Matlab7, Oxford, First Indian Edition 2006. [11] A. R. Bapat, Experimental Optimization of a manually driven flywheel motor, M.E. Thesis, VNIT, Nagpur. [12] A. R. Bapat, Experimentation of Generalized experimental model for a manually driven flywheel motor, PhD Thesis, VNIT, Nagpur.
Page 434
AUTHOR
Prof. (Ms) A. R. Lende received her Bachelor degree in Mechanical Engineering from BDCOE, Sewagram in 2004 and Master degree in Mechanical Engineering Design from PCE, Nagpur in 2007. She had registered her Ph.D. in 2009. She had given her teaching services to MIT, Kothrud, Pune and DMIETR, Wardha for three and two years respectively. The author also has an industrial experience of one year in the field of design and implementations. She also had worked in many mechanical engineering design projects and published papers in that region.
Dr. J. P. Modak is an Emeritus Professor and Dean (R&D) PCE, Nagpur. He has guided number of Ph.D. and PG Projects in the field of Design engineering, mathematical modeling, Artificial Intelligence, Vibration, etc. He is also consultant for many industries. He have delivered number of key not lectures, expert lectures at various conferences and workshops. He became a guiding light for researchers for driving research in systematic manner. The author had contributed through large number of papers in reputed international Journals & Conferences for developing research and path of research.
Page 435