Investigation of Prediction Models for Forces Calculation in Linear Induction Motor with Data-Based System Identification Algorithms

Created by W.Langdon from gp-bibliography.bib Revision:1.8901

@InProceedings{Lv:2019:IEMDC,

• author = "Gang Lv and Dihui Zeng and Tong Zhou and Michele Degano",
• title = "Investigation of Prediction Models for Forces Calculation in Linear Induction Motor with Data-Based System Identification Algorithms",
• booktitle = "2019 IEEE International Electric Machines Drives Conference (IEMDC)",
• year = "2019",
• pages = "1752--1756",
• month = "12-15 " # may,
• address = "San Diego",
• keywords = "genetic algorithms, genetic programming, Training, Support vector machines, SVM, Analytical models, Induction motors, Predictive models, Prediction algorithms, Classification algorithms, Finite element analysis, System identification, Random forests, Artificial neural networks, ANN, asymmetric secondary, symbolic regression, linear induction motor, thrust, vertical force, transversal force",
• DOI = "10.1109/IEMDC.2019.8785143",
• isbn13 = "978-1-5386-9351-3",
• abstract = "We investigates the prediction models of the thrust, vertical and transversal forces in the linear induction motors (LIMs) with the laterally asymmetric secondary. The models aim at presenting an analytical process for obtaining dynamic estimation model that takes account of the nonlinear effects in the analysis of the motors, e.g. magnetic saturation, end and edge effects. First, a number of simulation results of a prototype machine are generated by means of finite element method (FEM) for different conditions. The results, which mainly contains the values of the thrust, vertical and transverse forces, are classified as a function of the slip-frequency and the secondary displacement and divided into two sets: training set and test set. Different types of the identification algorithms for the prediction model are investigated: linear regression (LR), support vector machines (SVMs), symbolic regression using genetic programming (GP), random forests (RFRs), and artificial neural networks (ANNs), The prediction models with these algorithms are then optimized by the training set, and their accuracy is then validated by the test set. Finally, a discussion on the most optimal algorithm for the prediction model is given.",
• notes = "Also known as \cite{8785143}",

}

Genetic Programming entries for Gang Lv Dihui Zeng Tong Zhou Michele Degano

Citations