technologyforchapter7andchapter13continuousoptimi.docVIP

Technology for Chapter 7 and Chapter 13: Continuous Optimization (Numerical Techniques) Chapter 7: Discrete Single Variable Techniques Lab covers the following techniques: Dichotomous Golden Section Maple: Programs were written for each of these techniques. DICHOTOMOUS SEARCH ALGORITHM- Maximization Dr. William P. Fox, Naval Postgraduate School, Monterey, CA 93953 restart; DICHOTOMOUS:=proc(f::procedure,a::numeric,b::numeric,T::numeric,Ep::numeric) local x1,x2; x1:=(a+b)/2-Ep; x2:=(a+b)/2+Ep; printf(The interval [a,b] is [% 4.2f,% 4.2f]and user specified tolerance level is% 6.5f.\n,a,b,T); ### WARNING: %x or %X format should be %y or %Y if used with floating point arguments printf(The first 2 experimental endpoints are x1= % 6.3f and x2 = % 6.3f. \n,x1,x2); printf( \n); printf( \n); N:=ceil((ln(T/(b-a))/ln(0.5))); printf( Iteration x(1) x(2) f(x1) f(x2) Interval \n); iterate(f,a,b,N,x1,x2,Ep); val:=f(mdpt); printf( \n); printf( \n); printf(The midpoint of the final interval is% 9.6f and f(midpoint) = % 7.3f. \n,mdpt, val); printf( \n); printf( \n); ### WARNING: %x or %X format should be %y or %Y if used with floating point arguments printf(The maximum of the function is % 7.3f and the x value = % 9.6f \n,fkeep,xkeep); printf( \n); printf( \n); end: iterate:=proc(f::procedure,a::numeric,b::numeric, N::posint,x1::numeric,x2::numeric,Ep::numeric) local x1n,x2n,an,bn,i,fx1,fx2,j,f1,f2,fmid; global mdpt,fkeep,xkeep; i:=1; x1n(1):=x1; x2n(1):=x2; an(1):=a; bn(1):=b; i:=1; for j from 1 to N+1 do fx1(i):=f(x1n(i)); fx2(i):=f(x2n(i)); if fx1(i)=fx2(i) then an(i+1):=x1n(i); bn(i+1):=bn(i); x1n(i+1):=(an(i+1)+bn(i+1))/2-Ep; x2n(i+1):=(an(i+1)+bn(i+1))/2+Ep; else an(i+1):=an(i); bn(i+1):=x2n(i); x1n(i+1):=(an(i+1)+bn(i+1))/2-Ep; x2n(i+1):=(an(i+1)+bn(i+1))/2+Ep; fi; printf( % 3.0f % 11.4f % 10.4f % 10.4f %10.4f [% 6.4f, % 6.4f]\n,i,x1n(i),x2n(i),fx1(i),fx2(i),an(i),bn