I spent a summer working at an accelerator physics lab, where a constant problem was degaussing large beam steering magnets. We always used a DC current to drive the magnet through smaller Hysteresis loops, until the magnetization is roughly zero. Is there a better way?