Niagara transform question

To achieve a particle effect, I must convert the particle’s world coordinates through clip to prev clip to the clip space coordinates of the previous frame, and finally convert them back to world coordinates. Or use the camera’s matrix to offset the camera transform of the previous frame. The problem is that doing so results in the particle’s transform lagging behind. Is there any way to re-add the transform back?

I know it’s sound a little bit crazy, but in my test it’s the only way to get the “correct” motion vector for skeletal mesh.Guys, really need your help here.