You are on page 1of 2

Prescriptive Analytics with Mathematical Programming – CMSE11356

Individual Assignment 2021/2022

Instructions. Students must complete and submit online this assignment within the 48 hours’
time frame allotted for the assignment. Each activity must be developed within a limit of 300
words; mathematical models do not count within this limit. In total, the assignment must not be
longer than 4 pages. The ideal format is Time News Roman font 12 with simple spacing.

Activity 1. Consider a transportation problem in which we have to transport relief aid from
depots 𝑖 = 1,…,2 to shelters 𝑗 = 1,…,3. The following parameters are known: capacity of relief
depots (bi, in which b1 = 100 and b2 = 200), victims’ needs (dj, in which d1 = 80, d2 =150, and
d3 = 120), and shipping costs (cij, in which cij = bidj/100, for all i and j). Based on this description,
you are required to perform the following tasks. A) Give a feasible linear programming
formulation of the problem when shipping costs need to be minimised. Explain the decision
variables, the objective function and the constraints. B) Without solving your model, show that
it always returns a feasible solution and give any feasible solution. C) What exact method would
you use to find the optimal solution of this problem? Provide an overview of this method using
the appropriate terminology. D) Propose an alternative feasible mathematical programming
formulation based on maximising the amount of relief aid shipped. Discuss the pros and cons
of adopting this solution. E) Assume now we want to find the optimal location for a number of
depots that may supply the given shelters, as well as the best transportation plan, given that the
shelters’ location is fixed. There is a fixed-charge cost of fi to install a depot of capacity bi at
location 𝑖 = 1,…, N. Based on this description, propose a mixed-integer linear programming
formulation to minimise location and transportation costs.
(Maximum 30 marks)

Activity 2. Consider a company that must produce a number I of different products over a
planning horizon composed of T periods to satisfy forecasted customers’ demands of 𝑑𝑖𝑡 units,
for 𝑖 = 1,2, … , 𝐼 and 𝑡 = 1,2, … , 𝑇. Because of the need to install special equipment and tools
at the beginning of the production process, there is a set-up cost of 𝑠𝑖𝑡 (£) incurred in producing
product i in period t. Ideally, the company would save money by manufacturing more products
than the forecasted demand 𝑑𝑖𝑡 , so as to avoid paying the set-up cost every period. In this case,
the excess of production can be stored in the company’s warehouse. To hold product i in
inventory during one period costs on average ℎ𝑖𝑡 (£), including the cost of capital and storage
costs. There is enough space available in the warehouse to store the products. Initially, there is
no inventory of any product. You are required to: A) Propose a mixed-integer programme (MIP)
to plan the production and inventory levels of the products, in order to satisfy demand and
minimise the corresponding set-up and inventory costs. B) Consider now that product i requires
𝑝𝑖 units of capacity to be manufactured, excluding the set-up time that requires 𝑞𝑖 units of
capacity. Assume that the overall production capacity per period is 𝐶𝑎𝑝𝑡 . Propose a new MIP
to include this set of constraints. C) Assume that the manufacturer does not need to satisfy
100% of the demands necessarily, but for each unit of product i required in period t that is not
produced, there is a fine of 𝑓𝑖𝑡 (£). The unsatisfied products in a given period must be satisfied
in the next periods. At the end of the planning horizon, 100% of the demand must be satisfied,
though. Propose a new MIP to include this information. D) Assume now that, as it is rather
expensive and time-consuming to implement a set-up to start the production process, the
company has established a minimum production quantity of 𝑄𝑖𝑡 every time a set-up takes place.
1
Propose a new MIP to include this additional restriction. E) To avoid machine break-downs,
the manufacturer has realised that there is a maximum number of set-ups for each type of
product that should be done over the planning horizon. Assume that this integer number is 𝑅𝑖 ,
and include this information in your MIP.
(Maximum 30 marks)

Activity 3. Consider the LP:


max zp = 8x1 + 6x2
s.t.: 3x1 + 2x2 ≤ 28,
5x1 + 2x2 ≤ 42,
x1 ≤ 8, x2 ≤ 8,
x1 ≥0, x2 ≥ 0.

A) Solve the model geometrically and explain the steps to obtain the optimal solution. B) State
the problem in its standard form and solve it by the simplex method. Show the steps necessary
to obtain the optimal solution.
(Maximum 20 marks)

Activity 4. The PAwMP Corporation produces tables and chairs. A table requires 1 hour of
labour and 9 square board feet of wood, and a chair requires 1 hour of labour and 5 square board
feet of wood. Currently, 6 hours of labour and 45 square board feet of wood are available. Each
table contributes $8 to profit, and each chair contributes $5 to profit. Formulate and solve an
integer programming to maximise PAwMP’ profit using Branch-an-Bound (B&B). Summarise
the search history of the B&B in a tree.
(Maximum 20 marks)

(END OF ASSIGNMENT)

You might also like