Parameter identification using fuzzy neurons: Application to drones and induction motors

Abigail María Elena Ramírez Mendoza, José Roberto Covarrubias Fabela, Luis Antonio Amezquita-Brooks, Diana Hernández Alcántara

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Recently the use of unmanned aerial vehicles (UAV) has been extended in diverse applications. Within the most predominant configurations are rotary wing vehicles with multiple rotors. These vehicles normally present unstable dynamic behaviour in open loop, and therefore it is necessary to implement appropriate control systems. One of the most important requirements for the design of such control systems is the aerodynamic model of the multi-rotor propulsion system. On the other hand, parametric identification is widely used in the industrial context, such as control of electric machines. In particular for the case of induction motors, the knowledge of the model parameters is essential to tune the controllers properly. Among the most successful identification algorithms in recent applications are those based on neural networks. Within this context, an innovative type of neuronal networks based on adaptive fuzzy spiking neurons (AFSNs) has captured the attention of the community due to its fuzzy neuronal characteristics for unipolar and bipolar signals, such as its fuzzy learning algorithm, a sigmoidal activation function, refractory time, axonal delay, and spikes generation. This article explores for the first time the possibility of using AFSNs for the identification of the multi-rotor UAV propulsion subsystem and the parameters of the electric subsystem of an induction motor. The results show that AFSNs are capable of identifying both systems with a high degree of precision. This opens up the possibility of using AFSNs in both, experimental aerodynamics applications for UAVs and industrial control applications.

Original languageEnglish
Pages (from-to)75-81
JournalDyna (Spain)
Volume93
Issue number1
DOIs
Publication statusPublished - 1 Jan 2018

Fingerprint

Induction motors
Neurons
Identification (control systems)
Unmanned aerial vehicles (UAV)
Rotors
Propulsion
Aerodynamics
Control systems
Electric machinery
Refractory materials
Learning algorithms
Time delay
Chemical activation
Neural networks
Controllers
Drones

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Mendoza, Abigail María Elena Ramírez ; Fabela, José Roberto Covarrubias ; Amezquita-Brooks, Luis Antonio ; Alcántara, Diana Hernández. / Parameter identification using fuzzy neurons: Application to drones and induction motors. In: Dyna (Spain). 2018 ; Vol. 93, No. 1. pp. 75-81.
@article{fd1ee2591f724c3a94ed6a756dcb1eb6,
title = "Parameter identification using fuzzy neurons: Application to drones and induction motors",
abstract = "Recently the use of unmanned aerial vehicles (UAV) has been extended in diverse applications. Within the most predominant configurations are rotary wing vehicles with multiple rotors. These vehicles normally present unstable dynamic behaviour in open loop, and therefore it is necessary to implement appropriate control systems. One of the most important requirements for the design of such control systems is the aerodynamic model of the multi-rotor propulsion system. On the other hand, parametric identification is widely used in the industrial context, such as control of electric machines. In particular for the case of induction motors, the knowledge of the model parameters is essential to tune the controllers properly. Among the most successful identification algorithms in recent applications are those based on neural networks. Within this context, an innovative type of neuronal networks based on adaptive fuzzy spiking neurons (AFSNs) has captured the attention of the community due to its fuzzy neuronal characteristics for unipolar and bipolar signals, such as its fuzzy learning algorithm, a sigmoidal activation function, refractory time, axonal delay, and spikes generation. This article explores for the first time the possibility of using AFSNs for the identification of the multi-rotor UAV propulsion subsystem and the parameters of the electric subsystem of an induction motor. The results show that AFSNs are capable of identifying both systems with a high degree of precision. This opens up the possibility of using AFSNs in both, experimental aerodynamics applications for UAVs and industrial control applications.",
author = "Mendoza, {Abigail Mar{\'i}a Elena Ram{\'i}rez} and Fabela, {Jos{\'e} Roberto Covarrubias} and Amezquita-Brooks, {Luis Antonio} and Alc{\'a}ntara, {Diana Hern{\'a}ndez}",
year = "2018",
month = "1",
day = "1",
doi = "10.6036/8439",
language = "English",
volume = "93",
pages = "75--81",
journal = "Dyna (Spain)",
issn = "0012-7361",
publisher = "Federacion de Asociaciones de Ingenieros Industriales de Espana",
number = "1",

}

Parameter identification using fuzzy neurons: Application to drones and induction motors. / Mendoza, Abigail María Elena Ramírez; Fabela, José Roberto Covarrubias; Amezquita-Brooks, Luis Antonio; Alcántara, Diana Hernández.

In: Dyna (Spain), Vol. 93, No. 1, 01.01.2018, p. 75-81.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Parameter identification using fuzzy neurons: Application to drones and induction motors

AU - Mendoza, Abigail María Elena Ramírez

AU - Fabela, José Roberto Covarrubias

AU - Amezquita-Brooks, Luis Antonio

AU - Alcántara, Diana Hernández

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Recently the use of unmanned aerial vehicles (UAV) has been extended in diverse applications. Within the most predominant configurations are rotary wing vehicles with multiple rotors. These vehicles normally present unstable dynamic behaviour in open loop, and therefore it is necessary to implement appropriate control systems. One of the most important requirements for the design of such control systems is the aerodynamic model of the multi-rotor propulsion system. On the other hand, parametric identification is widely used in the industrial context, such as control of electric machines. In particular for the case of induction motors, the knowledge of the model parameters is essential to tune the controllers properly. Among the most successful identification algorithms in recent applications are those based on neural networks. Within this context, an innovative type of neuronal networks based on adaptive fuzzy spiking neurons (AFSNs) has captured the attention of the community due to its fuzzy neuronal characteristics for unipolar and bipolar signals, such as its fuzzy learning algorithm, a sigmoidal activation function, refractory time, axonal delay, and spikes generation. This article explores for the first time the possibility of using AFSNs for the identification of the multi-rotor UAV propulsion subsystem and the parameters of the electric subsystem of an induction motor. The results show that AFSNs are capable of identifying both systems with a high degree of precision. This opens up the possibility of using AFSNs in both, experimental aerodynamics applications for UAVs and industrial control applications.

AB - Recently the use of unmanned aerial vehicles (UAV) has been extended in diverse applications. Within the most predominant configurations are rotary wing vehicles with multiple rotors. These vehicles normally present unstable dynamic behaviour in open loop, and therefore it is necessary to implement appropriate control systems. One of the most important requirements for the design of such control systems is the aerodynamic model of the multi-rotor propulsion system. On the other hand, parametric identification is widely used in the industrial context, such as control of electric machines. In particular for the case of induction motors, the knowledge of the model parameters is essential to tune the controllers properly. Among the most successful identification algorithms in recent applications are those based on neural networks. Within this context, an innovative type of neuronal networks based on adaptive fuzzy spiking neurons (AFSNs) has captured the attention of the community due to its fuzzy neuronal characteristics for unipolar and bipolar signals, such as its fuzzy learning algorithm, a sigmoidal activation function, refractory time, axonal delay, and spikes generation. This article explores for the first time the possibility of using AFSNs for the identification of the multi-rotor UAV propulsion subsystem and the parameters of the electric subsystem of an induction motor. The results show that AFSNs are capable of identifying both systems with a high degree of precision. This opens up the possibility of using AFSNs in both, experimental aerodynamics applications for UAVs and industrial control applications.

UR - http://www.scopus.com/inward/record.url?scp=85042539186&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85042539186&partnerID=8YFLogxK

U2 - 10.6036/8439

DO - 10.6036/8439

M3 - Article

VL - 93

SP - 75

EP - 81

JO - Dyna (Spain)

JF - Dyna (Spain)

SN - 0012-7361

IS - 1

ER -