YES

TRS:

  div(X,e()) -> i(X)
  i(div(X,Y)) -> div(Y,X)
  div(div(X,Y),Z) -> div(Y,div(i(X),Z))

max/plus interpretations on N:

  div_A(x1,x2) = max{0, x1, x2}
  div#_A(x1,x2) = max{0, x1, x2}
  e_A = 0
  e#_A = 0
  i_A(x1) = max{0, x1}
  i#_A(x1) = max{0, x1}

precedence:

  div = e = i