Find the greatest value of #abc# for positive values of #a,b,c# subject to #ab +bc+ca=12# ?
1 Answer
Explanation:
Here's a method using a Lagrangian...
Let:
#{ (a=e^u), (b=e^v), (c=e^w) :}#
where
That takes care of the requirement that
We want to maximise:
#f(u,v,w) = e^ue^ve^w#
subject to:
#g(u,v,w) = e^ue^v+e^ve^w+e^we^u = 12#
Define the Lagrangian:
#L(u,v,w,lambda) = f(u,v,w)-lambda(g(u,v,w)-12)#
#color(white)(L(u,v,w,lambda)) = e^ue^ve^w-lambda(e^ue^v+e^ve^w+e^we^u-12)#
At the maxima and minima we have:
#grad L(u,v,w,lambda) = bb(0)#
That is, all of the partial derivatives of
#0 = (del L)/(del u) = e^u(e^ve^w-lambda (e^v+e^w))#
#0 = (del L)/(del v) = e^v(e^we^u-lambda (e^w+e^u))#
#0 = (del L)/(del w) = e^w(e^ue^v-lambda (e^u+e^v))#
#0 = (del L)/(del lambda) = -(e^ue^v+e^ve^w+e^we^u-12)#
That is:
#{ (bc-lambda(b+c) = 0), (ca-lambda(c+a) = 0), (ab-lambda(a+b)=0), (ab+bc+ca=12) :}#
We find:
#lambda = (bc)/(b+c) = (ca)/(c+a) = (ab)/(a+b)#
Hence:
#c = (lambdab)/(b-lambda) = (lambdaa)/(a-lambda)#
Hence:
#lambdaab-lambda^2b = lambdaab-lambda^2a#
Hence
Hence
Footnote
The Lagrangian method used here is a standard method for finding maxima and minima of continuous functions of several variables subject to constraints.
In this particular example, there may be an alternative algebraic method involving