Chapter 4 SimplexSearch for LP.ppt

EXERCISE: COMPUTING WITH BASIS INVERSE Basic Variable Sequence in Revised Simplex Computing Reduced Costs by Pricing EXERCISE: COMPUTING REDUCED COSTS BY PRICING (1) Compute the corresponding basis inverse. (2) Compute the associated pricing vector. (3) Compute the reduced costs on nonbasic variables without generating any simplex directions. ALGORITHM 5C: REVISED SIMPLEX SEARCH FOR LP Top Brass Example 5.3 THE SIMPLEX ALGORITHM Standard Display Standard-form data of Top Brass Trophy produce an initial table, with columns for each of the variables and the right-hand side: Initial Basic Solution Simplex Directions At current Top Brass solution, the simplex direction increasing when is fixed =1 At current Top Brass solution, the simplex direction increasing when is fixed =1 EXERCISE: CONSTRUCTING SIMPLEX DIRECTIONS Improving Simplex Directions and Reduced Costs Step Size and the Minimum Ratio Rule EXERCISE: DETERMINING THE MAXIMUM SIMPLEX STEP Updating the Basis After each move of simplex search, the nonbasic variable generating the chosen simplex direction enters the basic, and any one of the basic variables fixing step size leaves the basis. ALGORITHM 5A: RUDIMENTARY SIMPLEX SEARCH FOR LINEAR PROGRAMS Top Brass Example Stopping and Global Optimality When simplex Algorithm 5A is applied to a linear program in standard form, it either stops with a correct indication of unboundedness or produces a globally optimal solution. A current basic feasible solution is optimal if and only if no simplex direction is improving. 5.4 TWO PHASE SIMPLEX Starting Basis in the Two Phase Simplex A starting basis for simplex can be obtained by making basic one variable for each constraint row, with that variable having its only nonzero coefficient in the row and coefficient sign matching that of the corresponding right-hand side. Where no standard-form variable meets these requirements, an artificial is introduc