The velocity approximation given by the algorithm is m1 = n1lim, where n1 = Nep1/Ndt1. By the nature of the counting, Nep1 can be any whole number while Ndt1 is a natural number. If the velocity is higher than lim, there are more Iep impulses than dt impulses, so Ndt1 is always one; on the contrary, if the velocity is smaller than lim, there are more dt impulses than Iep, so Nep1 is one. Possible values for n1, when the velocity is high, are 1,2,3,…, and for low velocities are ½, ⅓, ¼,….
Helping one of our deacons who is in Memphis studying robotics to proofread a paper the general language of which seems to be English, when I can make it out.