You saw that the “natural” way to model SureStep’s backlogging problem, with IF functions, leads to a non-smooth model that Solver has difficulty handling. There is another version of the problem that is also difficult for Solver. Suppose SureStep wants to meet all demands on time (no backlogging), but it wants to keep its employment level as constant over time as possible. To induce this, it charges a cost of $1000 each month on the absolute difference between the beginning number of workers and the number after hiring and firing—that is, the absolute difference between the values in rows 17 and 20 of the original spreadsheet model. Implement this extra cost in the model in the “natural” way, using the ABS function. Using demands of 6000, 8000, 5000, and 3000, see how well Solver does in solving this nonsmooth model. Try several initial solutions, and see whether Solver gets the same optimal solution from each of them.
Answer to relevant QuestionsModify the Barney-Jones investment problem so that there is a minimum amount that must be put into any investment, although this minimum can vary by investment. For example, the minimum amount for investment A might be $0, ...In the pension fund problem, suppose there is an upper limit of 60 on the number of bonds of any particular type that can be purchased. Modify the model to incorporate this extra constraint and then optimize. How much more ...In the capital budgeting model in Figure 14.40, we supplied the NPV for each investment. Suppose instead that you are given only the streams of cash inflows from each investment shown in the file S14_49.xlsx. This file also ...Referring to the previous problem, if it is optimal for the company to produce sweatshirts, use SolverTable to see how much larger the fixed cost of sweatshirt machinery would have to be before the company would not produce ...Make up an example, as described in Problem 54, with 20 possible investments. However, do it so that the ratios of NPV to cash requirement are in a very tight range, from 3.0 to 3.2. Then use Solver to find the optimal ...
Post your question