In this work we model the observed evolution in comoving number density of Lyman-alpha blobs (LABs) as a function of redshift, and try to find which mechanism of emission is dominant in LAB. Our model calculates LAB emission both from cooling radiation from the intergalactic gas accreting onto galaxies and from star formation (SF). We have used dark matter (DM) cosmological simulation to which we applied empirical recipes for Ly$alpha$ emission produced by cooling radiation and SF in every halo. In difference to the previous work, the simulated volume in the DM simulation is large enough to produce an average LABs number density. At a range of redshifts $zsim 1-7$ we compare our results with the observed luminosity functions of LABs and LAEs. Our cooling radiation luminosities appeared to be too small to explain LAB luminosities at all redshifts. In contrast, for SF we obtained a good agreement with observed LFs at all redshifts studied. We also discuss uncertainties which could influence the obtained results, and how LAB LFs could be related to each other in fields with different density.