This content is not included in
your SAE MOBILUS subscription, or you are not logged in.
Information Based Selection of Neural Networks Training Data for S.I. Engine Mapping
Technical Paper
2001-01-0561
ISSN: 0148-7191, e-ISSN: 2688-3627
Annotation ability available
Sector:
Event:
SAE 2001 World Congress
Language:
English
Abstract
The paper deals with the application of two techniques for the selection of the training data set used for the identification of Neural Network black-box engine models; the research starts from previous studies on Sequential Experimental Design for regression based engine models. The implemented methodologies rely on the Active Learning approach (i.e. active selection of training data) and are oriented to drive the experiments for the Neural Network training. The methods allow to select the most significant examples leading to an improvement of model generalization with respect to a heuristic choice of the training data. The data selection is performed making use of two different formulation, originally proposed by MacKay and Cohn, based on the Shannon's Statistic Entropy and on the Mean Error Variance respectively. These techniques have been applied to assist the training of artificial Neural Networks for the estimation of engine torque and exhaust emissions of an S.I. engine, to be embedded into a powertrain dynamic model for the optimal design of engine control strategies (O.D.E.C.S.), now in use at Magneti Marelli.
Authors
Citation
Arsie, I., Marotta, F., Pianese, C., and Rizzo, G., "Information Based Selection of Neural Networks Training Data for S.I. Engine Mapping," SAE Technical Paper 2001-01-0561, 2001, https://doi.org/10.4271/2001-01-0561.Also In
Electronic Engine Controls: Modeling, Controls, Obd and Neural Networks
Number: SP-1585; Published: 2001-03-05
Number: SP-1585; Published: 2001-03-05
References
- Arsie I. Pianese C. Rizzo G. ‘Models for the Prediction of Performance and Emissions in a Spark Ignition Engine - A Sequentially Structured Approach’ SAE Paper 980779 , pub. on SAE 1998 Transactions, ‘Journal of Engines’ 106 1065 1079
- Arsie I. Pianese C. Rizzo G. ‘Development and Identification of a Hierarchical System of Models for Rapid Prototyping of SI Engines’ Proc. of the 6th IEEE Mediterranean Conference on “Control Systems” Alghero, Italy June 9-11 1998 Theory and Practice of Control and Systems 1998 World Scientific Publishing Ltd. London UK 182 188
- Pianese C. Rizzo G. ‘Interactive Optimization of Internal Combustion Engine Tests by means of Sequential Experiment Design’ Proc. of the 3rd Biennal Joint Conference on ‘Engineering Systems Design and Analysis - Esda 96’ Petroleum Division of ASME Montpellier, France July 1 - 4 1996 80 57 64
- Arsie I. Marotta F. Pianese C. Rizzo G. ‘Identification of Spark Ignition Engine Models based on Neural Network via Experimental Design Techniques’ Proc. of the 12th IFAC SYSID 2000, Symposium on ‘System Identification’ June 21-23 2000 Santa Barbara, CA
- Eschenauer H. Koski J. Osyczka A. ‘Multicriteria Design Optimization’ Springler - Verlag 1990
- Mowll D. Robinson D.R. Pilley A.D. 1996 Bayesian Experimental Design and its Application to Engine Research and Development SAE Paper 961157 in SP - 1177
- Mowll D. Robinson D.R. Pilley A.D. Optimising Engine Performance and Emissions Using Bayesian Techniques SAE Paper 971612 in SP - 1275
- Grimaldi C.N. Mariani F. On Board Diagnosis of Internal Combustion Engines: A New Model Definition and Experimental Validation SAE Paper 970211
- Arsie I. Pianese C. Rizzo G. ‘Enhancement of Control Oriented Engine Models Using Neural Network’ Proc. of the 6th IEEE Mediterranean Conference on “Control Systems” Alghero, Italy June 9-11 1998 Theory and Practice of Control and Systems - 1998 World Scientific Publishing Ltd. London UK 465 471
- Arsie I. Flora R. Pianese C. Rizzo G. Serra G. Development and Validation of a Model for mechanical Efficiency in a Spark Ignition Engine SAE Paper 1999-01-0905
- MacKay D.J.C. ‘Information - Based Objective Functions for Active Data Selection’ Neural Computation 4 4 590 604 1992
- Plutowski M. ‘Selecting Training Examples for Neural Network Learning’ University of California San Diego 1994
- Cohn D.A. ‘Neural Network Exploration Using Optimal Experiment Design’ AI Memo No.1491, CBCL Paper No. 99 1994
- Van Gorp J. ‘Nonlinear Identification With Neural Networks and Fuzzy Logic’ september 2000 Vrije Universiteit Brussel
- Amari S. Noboru M. Muller K.R. Finke M. Howard H.Y. Asymptotic Statistical Theory of Overtraining and Cross Validation IEEE Transactions of Neural Networks 8 5 september 1997 985 996
- Arsie I. Flora R. Pianese C. Rizzo G. Serra G. ‘A Computer Code for S.I. Engine Control and Powertrain Simulation’ SAE Paper 2000-01-0938 , on SP-1501 185 198
- Demuth H. Beal M. ‘Matlab Neural Network Toolbox’ The Mathworks 1997
- Patterson D.W. ‘ Artificial Neural Networks - Theory and Applications ’ Prentice Hall 1995
- Hecht - Nielsen R. ‘ Neurocomputing ’ Addison - Wesley 1987
- Fukumizu K. ‘Dynamics of Batch Training in Multilayer Networks - The existence of Overtraining’ 1998 BSIS Technical Report No.98-1 RIKEN Brain Science Institute Hirosowa, Wako-shi, Saitama, JP
- Bard J. ‘ Nonlinear Parameter Estimation ’ Academic Press 1990
- Fukumizu K. ‘Statistical Active Learning in Multilayer Perceptrons’ IEEE Transactions on Neural Network XX 1999
- Sung K.K. Niyogi P. ‘A formulation for Active Learning with Applications to Object Detection’ AI Memo No. 1438, CBCL Paper No. 119 1996
- Sahani. M. ‘Interactively Exploring A Neural code by Active Learning’ Poster in Neural Information and Coding meeting March 16 - 19 1997
- Cohn D.A. ‘Active Learning with Statistical Models’ AI Memo No. 1522, CBCL Paper No. 110 1995
- Lawrence S. Lee G. C. Tsoi A.C. ‘What size Neural Network gives Optimal Generalization? Convergence Properties of Back Propagation’ Technical Report UMIACS - TR - 96 - 22 e CS - TR - 3617 1996