You are on page 1of 2

Logistics Modelling, Prof.

Djamila Ouelhadj

Coursework

Due Date: 31 January 2019 by 3.30pm,


Report and Excel Files Submission on MOODLE

This coursework is worth 25% of the total unit mark. This is an individual
assignment. Group work and/or copying of work or reports are not permitted
and will be dealt with under the University rules on plagiarism.

What you need to submit


1. Submit your report on the unit MOODLE website. Provide the list of references
you have used in your coursework. The font size used must be 10 or greater.
Your report should be no longer than the specified word limits for each part.
2. Submit your Excel files on the unit MOODLE website.

Problem: Soft Drinks Distribution and Routing

A soft drinks wholesaler distributes drinks to coffee shops (retailers) primarily located
in the South East of England. The soft drinks are packaged in quantities of 12 bottles
of 500ml each. A retailer can only order in packs of 12 bottles.
A homogeneous fleet of 9 vans deliver the soft drinks to 60 retailers. Each retailer has
a specific quantity of demand for the soft drinks, and each van has a limited capacity
for carrying 100 packs of soft drinks. The vans depart from the distribution centre to
service the retailers with known locations and demands, and then return to the
distribution centre. The locations of the retailers are expressed using (x, y)
coordinates on the Euclidean plane. The coordinates of the locations and the demands
are described in Table 1 and in the file “Soft Drinks-Distribution-Routing.xls” on
Moodle. The main objective is to assign the vans to routes to deliver the soft drinks to
the retailers so as to minimise the total distance travelled by the vans, subject to the
following hard constraints:
• Each van departs from and returns to the distribution centre.
• Each retailer is visited exactly once by a single van.
• Each van has a limited capacity of 100 packs.
• The total load on any van associated with a given route must not exceed the
capacity of the van.
• The number of vans available at the distribution centre is 9.

1. Formulate mathematically the optimisation model of the Soft Drinks Distribution


routing problem to minimise the total distance travelled by the vans. Describe clearly
in detail your optimisation model: fitness function, decision variables, and
constraints. Provide any references in the literature you have used.
[Maximum 300 words excluding references]
[10 marks]

1

Logistics Modelling, Prof. Djamila Ouelhadj

2. Consider the data described in the file “Soft Drinks-Distribution-Routing.xls” on


Moodle. The data describes the coordinates of the locations and the demands of
retailers, coordinates of the location of the distribution centre, capacity of the vans,
and the number of vans available at the distribution centre.

Set up the Excel spreadsheet and optimisation models on Evolver and explain
clearly the implementation of the decision variables, the constraints, and the fitness
function. Indicate which Evolver solving method you have used to solve the problem
and the initial solution you have used.
Solve the problem to find the best minimum distance travelled by experimenting
with the following parameters:
(a) Population size: 30
(b) Crossover and mutation rates:
(i) Crossover rate = 0.8, mutation rate = 0.01.
(ii) Crossover rate = 0.9, mutation rate = 0.02.
(iii) Crossover rate = 0.6, mutation rate = 0.5.
(iv) Crossover rate = 0.3, mutation rate = 0.8.
(c) The stopping criterion for all the experiments should be set to 30000
generations.

For each combination provide a detailed description of the best solution: total
distance travelled, set of routes to be travelled by the vans, computation time
required to generate the best solution. For each van, specify the retailers to visit
and the order in which they have to be visited.
Provide a graphical illustration of the routes of the vans for the initial solution and
the overall best solution.
[Maximum 1000 words]
[50 marks]

3. The best solution, b-best, found in the literature for the data provided is 1035.
Analyse your results and discuss the effect of the crossover and mutation rates on
the quality of the best solutions for the different combinations and how these
solutions compare to b-best. Provide a graphical illustration to compare the quality
of the best solutions found for the different crossover and mutation rate
combinations and how they compare to b-best.
[Maximum 500 words]
[20 marks]

4. Solve the problem using the Tabu search (TS) method with OptQuest and
the same stopping condition as GA, 30000 iterations. Discuss and argument the
difference in the solution quality and computation time of both methods, and how
their solutions compare to b-best.
[Maximum 500 words]
[20 marks]

You might also like