第10章 线性规划.pptVIP

10.5.3整数规划 纯整数规划：所有决策变量要求取非负整数（这时引进的松弛变量和剩余变量可以不要求取整数）。 全整数规划：除所有决策变量要求取非负整数外，系数aij和常数bi也要求取整数（这时引进的松弛变量和剩余变量也必须是整数）。 混合整数规划：只有一部分的决策变量要求取非负整数，另一部分可以取非负实数。 0－1整数规划：所有决策变量只能取 0 或 1 两个整数。 整数规划求解方法 分支定界法 割平面法 Figure 4.3. Graphical solution of Example 4.3. * The basic idea behind the duality theory is that every linear programming problem has an associated linear program called its dual such that a solution to the original linear program also gives a solution to its dual. Thus, whenever a linear program is solved by the simplex method, we are actually getting solutions for two linear programming problems * Dantzig’s simplex method and its variants, start with an extreme point solution in the feasible region and at each iteration, move to an adjacent extreme point with an improved objective value, always traversing along the boundary of the feasible region. Figure 4.6 illustrates the sequence of points after each iteration (shown as ‘‘dots’’) by the simplex method, where the arrows along the boundary show the path. Unlike the simplex method, the interior point methods, popularly known as Karmarkar’s algorithm, traverse the interior of the feasible region as shown in Fig. 4.6. * In practice, many LP problems do require integer solutions for some of the variables. For instance, it is not possible to employ a fractional number of workers or produce a fractional number of cars. The term (linear) integer programming refers to the class of LP problems wherein some or all of the decision variables are restricted to be integers. But solutions of integer programming problems are generally difficult, too time consuming, and expensive. Hence, a practical approach is to treat all integer variables as continuous and solve the associated linear program by the simplex method. We may be fortunate to get some of the values of the variables as integers automatically, but when the simplex method produces fractional solutions for some integer variables, they are generally rounded off to