Facial reconstruction

Search LJMU Research Online

Browse Repository | Browse E-Theses

Performance Evaluation of AquaFeL-PSO Informative Path Planner under Different Contamination Profiles

Jara Ten Kathen, M, Peralta, F, Johnson, P, Jurado Flores, I and Gutierrez Reina, D (2023) Performance Evaluation of AquaFeL-PSO Informative Path Planner under Different Contamination Profiles. In: Data Analytics and Computational Intelligence: Novel Models, Algorithms and Applications. Studies in Big Data, 132 . Springer, pp. 405-431. ISBN 978-3-031-38324-3

[img] Text
Performance Evaluation of AquaFeL-PSO Informative Path Planner under Different Contamination Profiles.pdf - Accepted Version
Restricted to Repository staff only until 13 September 2025.

Download (2MB)

Abstract

The use of Autonomous Surface Vehicles allows streamlining the task of monitoring the water quality parameters of water resources, reducing costs and time spent. In addition, the vehicles are capable of taking water measurements in hard-to-reach areas. This chapter evaluates the performance of the AquaFeL-PSO monitoring system in different contamination profiles. The AquaFeL-PSO is an algorithm based on Particle Swarm Optimization, Gaussian processes and Federated Learning technique. The operation mode of the algorithm is two-phase, the exploration phase, which is responsible for covering the largest possible area of the water resource in order to detect contaminated areas, and the exploitation phase, which is responsible for accurately characterizing the water quality parameters in the contaminated areas. This system is evaluated using different benchmark functions in order to observe the performance of the system under different contamination profiles. The results show that the AquaFeL-PSO was able to generate models of the water quality parameters in eight of the ten contamination profiles. In addition, it had the best performance in detecting pollution peaks and generating the model of the water quality parameters of the entire water resource.

Item Type: Book Section
Uncontrolled Keywords: Informative Path Planning, Particle Swarm Optimization, Fed-erated Learning, Autonomous Surface Vehicles, Machine Learning, WaterMonitoring, Multi-modal Problems (21) (PDF) Performance Evaluation of AquaFeL-PSO Informative Path Planner under Different Contamination Profiles.
Subjects: T Technology > TA Engineering (General). Civil engineering (General)
Divisions: Engineering
Publisher: Springer
SWORD Depositor: A Symplectic
Date Deposited: 25 Apr 2023 09:22
Last Modified: 26 Sep 2023 11:07
DOI or ID number: 10.1007/978-3-031-38325-0_17
URI: https://researchonline.ljmu.ac.uk/id/eprint/19161
View Item View Item