Thanks to developing technology, the linear motor applications have been getting popularity these days. Linear motors applications have been chosen in order to increase system performance and efficiency. In this paper, non-sinusoidal magnetic flux of the secondary surface was measured by measurement coils placed in the secondary block. After magnetic flux values placed end coils at entry and exit side of primary, they were remeasured. Measured magnetic flux values decomposed two components (end effect and main signals) by means of discrete wavelet packet transform and artificial neural network that operating with end coils and without end coils. Third level wavelet packet decomposition of measured magnetic values was carried out by Simulink wavelet filter bank and then decomposition signals of magnetic flux were obtained. To get obtain end effect and main signals, which use as target data for ANN model were synthesized respectively Copyright (C) 2011 Praise Worthy Prize S.r.l. - All rights reserved.