This is simply related to the fact that all proper Rietveld program anticipate (if no standard deviation of the measured intensity value is indicate) that the intensity values pertaining to a given diffraction angle follow the Poisson statistics. In view of that there CANNOT be negative intensities. A typical cause is background subtraction, which you should not do (at least, if you do not consider that in the standard deviation).

Dear professor Andreas Leineweber, I appreciate your reply. I didn't subtract the background, and I double-checked to be sure, and the problem is still there.

Thanks to Dr. Maryam Shabani, who taught me how to fix this error by simply shifting the background to upper than zero.

Dear Hessam Addin Nadernia , One thing you can do to overcome "negative intensities in scan data points" is to select your scan, go to the object inspector and look for "Shifts". There is an option there ("Add [cts]") where you can add counts to your scan until it is all positive.