YES

TRS:

  f(f(X)) -> c()
  c() -> d()
  h(X) -> c()

max/plus interpretations on N:

  f_A(x1) = max{2, 1}
  f#_A(x1) = max{0, 2}
  c_A = 2
  c#_A = 1
  d_A = 0
  d#_A = 0
  h_A(x1) = max{2, 1}
  h#_A(x1) = max{2, 0}

precedence:

  f = h > c > d