RT info:eu-repo/semantics/article
T1 Glasius bio-inspired neural networks based UV-C disinfection path planning improved by preventive deadlock processing algorithm
A1 Rodrigo, Daniel Vicente
A1 Sierra Garcia, Jesús Enrique
A1 Santos, Matilde
K1 Complete coverage path planning
K1 Mobile robot
K1 UV-C
K1 Deadlocks
K1 Escape routes
K1 Electrotecnia
K1 Electrical engineering
K1 Ingeniería mecánica
K1 Mechanical engineering
AB The COVID-19 pandemic made robot manufacturers explore the idea of combining mobile robotics with UV-Clight to automate the disinfection processes. But performing this process in an optimum way introduces somechallenges: on the one hand, it is necessary to guarantee that all surfaces receive the radiation level to ensurethe disinfection; at the same time, it is necessary to minimize the radiation dose to avoid the damage of theenvironment. In this work, both challenges are addressed with the design of a complete coverage path planning(CCPP) algorithm. To do it, a novel architecture that combines the glasius bio-inspired neural network (GBNN),a motion strategy, an UV-C estimator, a speed controller, and a pure pursuit controller have been designed.One of the main issues in CCPP is the deadlocks. In this application they may cause a loss of the operation, lackof regularity and high peaks in the radiation dose map, and in the worst case, they can make the robot to getstuck and not complete the disinfection process. To tackle this problem, in this work we propose a preventivedeadlock processing algorithm (PDPA) and an escape route generator algorithm (ERGA). Simulation resultsshow how the application of PDPA and the ERGA allow to complete complex maps in an efficient way wherethe application of GBNN is not enough. Indeed, a 58% more of covered surface is observed. Furthermore, twodifferent motion strategies have been compared: boustrophedon and spiral motion, to check its influence onthe performance of the robot navigation.
PB Elsevier
SN 0965-9978
YR 2023
FD 2023-01
LK http://hdl.handle.net/10259/7483
UL http://hdl.handle.net/10259/7483
LA eng
DS Repositorio Institucional de la Universidad de Burgos
RD 21-nov-2024