a) Let e be the unit element of the monoid (A,∗). Then
f(0)+g(e)+h(0)⋅e=0∗e=0andf(e)+g(0)+h(e)⋅e=e∗0=0,
so that f(e)=g(e)=f(0)=g(0)=h(e)⋅e=h(0)⋅e=0, whence e=e∗e=f(e)+g(e)=0.
Then 0∗x=x and x∗0=x, for any x≥0, and we obtain
f(0)+g(x)+h(0)⋅x=xandf(x)+g(0)+h(x)⋅x=x,
so that f(x)=x(1−h(x)) and g(x)=x(1−h(0)), for any x≥0. Since f(x), g(x)≥0, it follows that h(x)∈[0,1], ∀x≥0.
x∗y=x+y−x⋅h(x)−y⋅h(0)+h(x)⋅∣x−y∣,∀x,y≥0.
xh(x)−yh(y)=h(0)(x−y)+(h(x)−h(y))⋅∣x−y∣,∀x,y≥0.
Since h is bounded, it follows that limx→yxh(x)=yh(y), for any y≥0, so that the function p:A→A, p(x)=xh(x), is continuous. But then h is continuous on (0,∞).
Also, for any y>0 we have
x→ylimx−yp(x)−p(y)=h(0),
so that there are a=h(0) and b≥0 such that p(y)=ay+b=h(0)y+b,∀y>0. Then b=limy→0p(y)=p(0)=0. But then yh(y)=p(y)=yh(0) for any y>0 and it follows that h(y)=h(0),∀y>0. The function h is thus constant, hence continuous.
b) Let k=h(0). Then h(x)=k and f(x)=g(x)=x(1−k), for any x≥0, and x∗y=(x+y)(1−k)+k∣x−y∣,∀x,y≥0. Then (1∗1)∗2=1∗(1∗2)⟹k(4k−2)=0, so that k∈{0,21}.
For k=0 we have that f=g=idA and x∗y=x+y,∀x,y≥0.
For k=21 we have that f(x)=g(x)=2x,∀x≥0 and x∗y=2x+y+2∣x−y∣=max(x,y),∀x,y≥0.