代写program、代做C++设计编程

日期：2025-07-15 07:49

Fakult¨at Verkehrswissenschaften ”Friedrich List“

Institut f¨ur Wirtschaft und Verkehr

Professur f¨ur Verkehrsbetriebslehre und Logistik

Decision Support in Transportation Logistics

Summer Term 2025

Detailed Specification of the Homework Assigment

“Fuel-Optimal Supermarket Supply from a Central Warehouse”

release date: 15.04.2025

submission deadline: 21.07.2025 - 08:00

Abstract

This document outlines the homework assignent for the summer term 2025 as part of the

lecture “Decision Support in Transportation Logistics”. The basic scenario assumes that

different vehicle fleets are to be dispatched to deliver goods from a central warehouse to a

collection of up to approx. 30 supermarket sites in a given city at least fuel consumption.

For this, a CVRP-like scenario must be solved to identify vehicle routes.

In particular, this homework assignment requires the completion of four tasks. The

first task comprises the composition of XML-coded scenario files. The second task asks

for the example calculation of the fuel consumption for a given route and a given vehicle.

The third task comprises the preparation and execution of a collection of model-based

decision making experiments using a MILP-solver together with the specific software

VRP++. Finally, the fourth task asks for documentation and presentation of the findings

from the different tasks.

Remarks & Instructions

• Every implementation related to the following four tasks must be done in C++.

• Two additional software libraries are needed: VRP++4.02. You have to install

both packages on your computer.

• This homework assignment is based on the paper Kopfer, H.W., Sch¨onberger, J. &

Kopfer, H. Reducing greenhouse gas emissions of a heterogeneous vehicle fleet. Flex

Serv Manuf J 26, 221–248 (2014). https: // doi. org/ 10. 1007/ s10696-013-9180-9.

Download the paper and read it as soon as possible. It is essential that you fully

understand the proposed decision model of the EVRP-VC.

• Q-%-A-sessions via Zoom will be announced in OPAL. Please consult the schedule

of the module regularly.

• If you need individual support, please book a consultation time slot here: https://

bildungsportal.sachsen.de/opal/auth/RepositoryEntry/48899129348. The list

of available time slots will be updated every month.

• More detailed information will be provided during the term according to the progress

of the lecture series.

1

Task 1 - Specification of the Fleet Deployment Sce nario Data

This task comprises the detailed specification of the five investigated VRP-scenarios.

The final specification is unique for every student since the request portfolio composition

depends upon your individual student ID (“Immatrikulations-Nummer”).

A B C D E F G

Table 1: Selection of the city for investigation

In order to find out in which city you have to route trucks for the supermarket supply

write your student ID into Tab. 1 (one number per field). The number contained in

column E determines the city.

number in column E

in Table. 1

City

9 Hanover (ca. 130 locations)

8 Amsterdam (ca. 271 locations)

7 Warszawa (ca. 300 locations)

6 Bremen (ca. 114 locations)

5 Praha (ca. 250 locations)

4 Dresden (ca. 130 locations)

3 Frankfurt am Main (ca. 245 locations)

2 Munich (ca. 450 locations)

1 D¨usseldorf (ca. 220 locations)

0 Stuttgart (ca. 150 locations)

Table 2: Available cities

Now, proceed to Tab. 2 and label the row that corresponds to the number stored in

column E in Tab. 1. Use the overpass-turbo API-website of Openstreetmap to retrieve

the stored addresses of the supermarkets available in your city1

. Export the selected raw

data of the supermarket locations. Continue with the selection of the first 31 nodes from

the list of all found nodes. The first node will become the central depot of the scenarios

but the next 30 nodes form the customer locations for the scenarios.

Assign the delivery quantities 1, 2, 3, 4, 5, 4, 3, 2, 1, 2, 3, 4, 5, 4, 3... to the delivery

sites following the order of the requests in the exported list. These quantities determine

the number of standard euro-pallets packed for each outlet.

We are going to inspect the impacts of the vehicle fleet. Therefore, it is necessary to

determine five different fleets. We assume that only semitrailers with a capacity of 36

euro-pallets each can be used. Currently, the fleet manager can use up to 5 trucks with

these configurations. This fleet is called the default fleet (fleet 0).

1Supermarkets are tagged in Openstreetmap by node[shop=supermarket];

2

The fleet manager is now instructed to reduce the total consumed fuel of the fleet

during the fulfillment of all customer requests. Therefore, he wants to replace the orig inal objective function of the CVRP (total travel distance) by the objective function

that determines the total fuel consumption. This leads to a variant of the CVRP called

the Emission Minimization Vehicle Routing Problem with Vehicle Categories (EVRP VC) introduced in Kopfer, H.W., Sch¨onberger, J. & Kopfer, H. Reducing greenhouse

gas emissions of a heterogeneous vehicle fleet. Flex Serv Manuf J 26, 221–248 (2014).

https: // doi. org/ 10. 1007/ s10696-013-9180-9

The aforementioned paper introduces four additional different types of vehicles (Table

1 in the manuscript) to setup hetergoneneous fleets. The already used vehicles with the

capacity of 36 euro-pallets fall into the category V C40. The maximal number of euro pallets to be loaded in the three different vehicle categories is: 7 (V C12), 4 (V C7.5) and 2

(V C3.5). We assume that such an euro-pallet weights 1.7 tons in average.

Our fleet manager wants to find out how many fuel consumption can be saved if he

adds vehicles of the three new categories V C12, V C7.5 and V C3.5 to the fleet. In particular,

he add one V C12-vehicle (fleet 1), two V C12-vehicles (fleet 2), two V C12-vehicles and one

V C7.5-vehicle (fleet 3), and two V C12-vehicles, one V C7.5-vehicle, one V C3.5-vehicle (fleet

4) to the default fleet.

• Task 1a: Code the aforementioned data into five different xml-scenario files apply ing the VRP++-xml-tagging scheme presented in the lecture materials

• Task 1b: Create a map in openstreetmap.org in which all 31 locations are shown

(the warehouse must be printed differently from the delivery sites).

Task 2 - Example Fuel Calculation

This task aims ensuring that you have understood the way how the fuel consumption is

calculated. For this, we first specify a particular round trip. It starts at the depot and

terminates there. This round trip serves the first three delivery sites from your list. The

node sequence is R+=0 → 1 → 2 → 3 → 0.

• Task 2a: Calculate the fuel consumption for one vehicle of type V C40 if it travels

along R+.

• Task 2b: Calculate the fuel consumption for one vehicle of type V C12 if it travels

along R+.

• Task 2c: Calculate the fuel consumption for one vehicle of type V C7.5 if it travels

along R+.

• Task 2d: Calculate the fuel consumption for one vehicle of type V C3.5 if it travels

along R+.

Now determine the inverse route R− of R+, which is R−=0 → 3 → 2 → 1 → 0

• Task 2e: Calculate the fuel consumption for one vehicle of type V C40 if it travels

along R−.

3

• Task 2f: Calculate the fuel consumption for one vehicle of type V C12 if it travels

along R−.

• Task 2g: Calculate the fuel consumption for one vehicle of type V C7.5 if it travels

along R−.

• Task 2h: Calculate the fuel consumption for one vehicle of type V C3.5 if it travels

along R−.

• Task 2i: Summarize and analyze your observations associated with the inversion

of the round trip?.

Task 3 - Exploring the Limits of Mathmatical Pro gramming

This task is dedicated to answering the question: Can we use linear programming to solve

instances of the of the aforementioned vehicle routing scenario?

To find experimentally founded answers, we incorporation the scip optimization suite

and VRP++in a C++ - project. These libraries are used to setup and solve the mixed

integer linear program proposed in Kopfer, H.W., Sch¨onberger, J. & Kopfer, H. Reducing

greenhouse gas emissions of a heterogeneous vehicle fleet. Flex Serv Manuf J 26, 221–248

(2014). https: // doi. org/ 10. 1007/ s10696-013-9180-9. This tool is used to carry

out the following tasks to inspect the impacts of extending the default fleet:

• Task 3a: Find an optimal solution for the scenario including fleet 0 and the first ten

customer requests. Describe the found solution and print them in openstreetmap

using synchronized problem data.

• Task 3b: Find an optimal solution for the scenario including fleet 1 and the first ten

customer requests. Describe the found solution and print them in openstreetmap

using synchronized problem data.

• Task 3c: Find an optimal solution for the scenario including fleet 2 and the first ten

customer requests. Describe the found solution and print them in openstreetmap

using synchronized problem data.

• Task 3d: Find an optimal solution for the scenario including fleet 3 and the first ten

customer requests. Describe the found solution and print them in openstreetmap

using synchronized problem data.

• Task 3e: Find an optimal solution for the scenario including fleet 4 and the first ten

customer requests. Describe the found solution and print them in openstreetmap

using synchronized problem data.

Next we want to find out if and how the enlargement of the request set impacts the

ability of the MILP-solver to find optimal or suboptimal solutions.

• Task 3f: Now extend the previously investigated set of the first ten customer

requests step by step. Take fleet 4 and combine it with the first 12 customer requests.

4

Next, take fleet 4 and combine it the first 14 customer requests, ... Use the linear

model and solve these instances one after the other. For each instance collect the

following information

– the best found objective function value

– is the best found solution feasible?

– the relative optimality gap?

– the number of deployed vehicles of every category

• Task 3g: Explain the changes of the routes (if there are changes) if you increase

the customer request set.

For all subtasks of task 3 set the solver time limit to 60 minutes.

Task 4 - Documentation and Submission of Your Re sults

The only item you need to submit is a 20-minute video. The length of the video is strict

any exceeding will lead to a failure of this examination.

This video has to be submitted not later than 21st July 2025 at 08:00 (Dresden local

time). Do not expect a deadline extension.

For submitting the video, upload the video file into your TU Dresden-Cloudspace.

Next, share the folder with the video with me. I will then download the video for inspec tion and evaluation. Do not remove or update the video before I advice you to do so.

Only this submission process is authorized. Using another process without the previous

approval by the lecturer will lead to a failure of this examination.

Issues to be considered when producing the requested

video

It is important that you consider all requirements specified below when cre ating and submitting your video. Missing at least fulfilling one of the require ments may result in a failure of the complete task.

• You must be seen in the video while you are commenting (picture in picture)

throughout the complete duration.

• A title slide is necessary on which you indicate your name, student ID, and the title

of your presentation.

• All results of the specified tasks must be answered in the presentation.

• It is necessary to demonstrate that you can compile your executable (the applica tion).

• You have to demonstrate that the application works as expected by showing an

example.

5

• Produce, show and explain at least one map in which you draw the identified best

solution (do not forget to explain which scenario it is)

• Explain/demonstrate how you fetched and processed the results from the experi ments.

• A last summarizing slide is necessary

• There are no background noises (e.g. caused by air conditions or other devices).

• Watermarks or other logos do not appear.

• If you explain source code then avoid a pure ”scrolling” through long listings.

• Respect the maximal allowed video duration.

• Make a script and define the sceneries of your video before you start recording. If

necessary, record shorter sceneries and put them together afterward.

• Be sure that your explanations can be followed without being involved in your

project work.

6