代做ELE000114M、代写Engineering设计编程、代写Python，c/c++程序语言

日期：2021-01-14 10:33

Department of Electronic Engineering

Assessments 2019/20

ELE00114M

Swarm Robotics

This assessment (Essay) contributes 100% of the assessment for this module.

Clearly indicate your Exam Number on every separate piece of work submitted.

Unless the assessment specifies a group submission, you should assume all submissions are

individual and therefore should be your own work.

All assessment submissions are subject to the Department’s policy on plagiarism and, wherever

possible, will be checked by the Department using Turnitin software.

Submission is via VLE and is due by 12:00 on 21 January 2021. Please try and submit early

as any late submissions will be penalised.

Please remember that if this is your first year of study, you need to complete the mandatory

Academic Integrity Tutorial http://www.york.ac.uk/integrity/

ELE000114M

Assessment

This assessment is concerned with implementing a swarm foraging algorithm in the

V-REP/CoppeliaSim robot simulator, and analysing its performance in comparison to a simple

foraging algorithm that is provided to you.

Foraging task

The environment for the foraging task consists of a walled arena containing: a swarm of foot-bot

robots, a ‘nest’ area, and various foraging sites. The robots begin in the nest area, and must

explore their environment to locate the foraging sites, which are indicated by round patches on

the floor. When a robot drives over a foraging site, it will automatically collect a ‘food item’ that

must be returned to the nest area.

A robot can find its way back to the nest by performing phototaxis towards lights that are

situated above the nest area. The robot’s ground sensors may also be used to sense the grey

floor representing the nest. When a robot carrying a food item returns to the nest area, that food

item is automatically removed from the simulation, and recorded as collected. The robot may

then return to foraging for other food items.

Each foraging site contains food items of a different value, and each robot may only carry one

food item at a time. The aim of the foraging task is for the swarm to maximise the value of food

items returned to the nest within a given time period.

Implementation

Download the assessment pack from the course website, which contains:

● A V-REP/CoppeliaSim scene .ttt file that defines the environment and robot model;

● a controller script for the puck robot that implements a simple foraging algorithm; and

● functions that log data to a file and draw informative overlays.

You must write a Lua or Python controller for your own swarm foraging algorithm that

outperforms the simple foraging algorithm provided in the assessment pack.

Your solution should improve foraging performance via cooperation between robots, rather than

simply improving the performance of individual robots. You may take inspiration from existing

swarm foraging algorithms that exist in the literature, provided that you cite them in your report.

Higher marks will be awarded to solutions that prioritise foraging sites based on their food item

value and distance from the nest, in order to maximise the value of food times collected.

Finally, your implementation must not use centralised control or global sensing/communication.

Only decentralised control and local sensing/communication are permitted.

ELE000114M

Analysis

Define one or more suitable metrics that quantify the behaviour of the swarm, then carry out a

series of experiments to analyse the performance of the two foraging algorithms.

You should select appropriate statistical tests to demonstrate whether your own swarm foraging

algorithm outperforms the simple foraging algorithm provided in the assessment pack. The

design of your experiments must ensure a fair comparison between the two foraging algorithms.

For any experiments carried out, you should perform repeat runs with different random

initialisations.

Marking criteria

This assessment is marked out of 100, based on code and a report that are subject to the

marking criteria in this document.

Code

The marking criteria for your code (Lua controller and analysis scripts) are as follows:

Code quality (5 marks): Your code should follow good programming practices such as

commenting, proper structure, and use of descriptive variable names.

Report

You should prepare a 4-page report that documents your implementation and analysis. You may

use as many additional pages as you wish for appendices. You should format your submission

using the IEEE style (in LaTeX, LibreOffice, or Word), templates for which can be downloaded

from:

http://www.ieee.org/conferences_events/conferences/publishing/templates.html

The marking criteria for your report are as follows:

Readability, structure, and referencing (5 marks): Your report should be written in clear

English, be well-structured and easy to read, and make appropriate use of references.

Abstract, introduction, and conclusion (5 marks): Include a brief abstract; an informative

introduction, which clearly explains the context and motivation for the work; and a conclusion

that demonstrates good critical evaluation of the work undertaken.

ELE000114M

Literature review (15 marks): Provide a short literature review of swarm robotics foraging

algorithms, including a critical analysis of existing approaches. If your implementation was

inspired by an existing approach, you must cite and discuss it in your review.

Implementation (25 marks): For both the simple foraging algorithm and your own foraging

algorithm: describe the implementation of the algorithm in the V-REP/CoppeliaSim robot

simulator using a combination of text, diagrams, and/or pseudocode. Explain how the controller

for an individual robot in the swarm works, and how this results in a collective foraging

behaviour. You should also explain how your own foraging algorithm differs from the simple one

provided for you, justifying each of your design decisions.

Qualitative analysis of foraging (5 marks): Critically evaluate the two different foraging

algorithms, based on qualitative observations. Discuss how the placement of foraging sites may

hinder the performance of either algorithm, and what can be done to overcome any issues.

Quantitative analysis of foraging (25 marks): Clearly define any metrics that you have used

to quantify the behaviour of the swarm, and justify your reasoning. Describe your experimental

method in sufficient detail that it would be possible for someone else to reproduce your results.

Present the results of your experimental analysis of the two foraging algorithms, using plots

and/or tables of data where appropriate. You should aim to demonstrate rigour in your

experimental method, and the appropriate use of statistical tests to substantiate any claims.

Evidence of improved performance (5 marks): Your analysis should demonstrate that your

own foraging algorithm consistently outperforms the simple foraging algorithm provided in the

assessment pack.

Bridging the reality gap (10 marks): For both of the foraging algorithms, discuss how well the

simulated implementations would transfer to the real foot-bot robots. Describe any modifications,

or additional infrastructure, that would be required for successful deployment on real robots.

Submission guidelines

You must submit your report as a PDF file. If the specified page limits are not obeyed, your

submission will be penalised.

You must also submit your code, including any analysis scripts and data used to generate the

figures in your report. You should clearly state, with a “readme” file submitted with your code,

how to run your code. If the code does not run on a standard teaching machine, we reserve the

right to award zero marks to the section associated with the coding of the solution.

Your examination number must be written on the front of your report and each answer page, and

in each file of your code (as a comment). You must not identify yourself in any other way (in

either your report, or code).

ELE000114M

Any queries on this assessment should be addressed to Mark A Post <mark.post@york.ac.uk>.

Replies to email queries will be published online (and available to all students) on the course

website.