This paper shows that the aerodynamic effects can be compensated in a quadrotor system by means of a control allocation approach using neural networks.Thus,the system performance can be improved by replacing the class...This paper shows that the aerodynamic effects can be compensated in a quadrotor system by means of a control allocation approach using neural networks.Thus,the system performance can be improved by replacing the classic allocation matrix,without using the aerodynamic inflow equations directly.The network training is performed offline,which requires low computational power.The target system is a Parrot MAMBO drone whose flight control is composed of PD-PID controllers followed by the proposed neural network control allocation algorithm.Such a quadrotor is particularly susceptible to the aerodynamics effects of interest to this work,because of its small size.We compared the mechanical torques commanded by the flight controller,i.e.,the control input,to those actually generated by the actuators and established at the aircraft.It was observed that the proposed neural network was able to closely match them,while the classic allocation matrix could not achieve that.The allocation error was also determined in both cases.Furthermore,the closed-loop performance also improved with the use of the proposed neural network control allocation,as well as the quality of the thrust and torque signals,in which we perceived a much less noisy behavior.展开更多
Photonic technology combined with artificial intelligence plays a key role in the development of the latest smart system trends,integrating cutting-edge technology with machine learning models.This paper proposes a tr...Photonic technology combined with artificial intelligence plays a key role in the development of the latest smart system trends,integrating cutting-edge technology with machine learning models.This paper proposes a transmission-reflection analysis based system using dielectric nanoparticle-doped fiber combined with artificial intelligence to address one of the major problems in the distributed sensing approach:reducing the cost while maintaining high spatial resolution to close the gap between distributed sensors and the general public.Machine learning-based models are designed to classify the perturbed positions when the same force is used and force regression when different forces are applied on each position.The results show an accuracy of 99.43%in the position classification of multiple disturbances and an rms error of 1.53 N in the force regression,which represents 5%of the force range.In addition,a smart environment using the current system is proposed,which presented 100%accuracy in identifying the positions of different persons in the environment.This smart environment enables remote home care of patients with high reliability,intelligent decision-making,and a predictive capability.展开更多
文摘This paper shows that the aerodynamic effects can be compensated in a quadrotor system by means of a control allocation approach using neural networks.Thus,the system performance can be improved by replacing the classic allocation matrix,without using the aerodynamic inflow equations directly.The network training is performed offline,which requires low computational power.The target system is a Parrot MAMBO drone whose flight control is composed of PD-PID controllers followed by the proposed neural network control allocation algorithm.Such a quadrotor is particularly susceptible to the aerodynamics effects of interest to this work,because of its small size.We compared the mechanical torques commanded by the flight controller,i.e.,the control input,to those actually generated by the actuators and established at the aircraft.It was observed that the proposed neural network was able to closely match them,while the classic allocation matrix could not achieve that.The allocation error was also determined in both cases.Furthermore,the closed-loop performance also improved with the use of the proposed neural network control allocation,as well as the quality of the thrust and torque signals,in which we perceived a much less noisy behavior.
基金Coordenacao de Aperfeicoamento de Pessoal de Nível Superior(309737/2021-4,310668/2021-2,310709/2021-0,84336650,2021-WMR44,304049/2019–0)Financiadora de Estudos e Projetos(2784/20)+1 种基金Agence Nationale de la Recherche(ANR-17-CE08-0002)Fundacao para a Ciência e a Tecnologia(2021.00667.CEECIND,PTDC/EEI-EEE/0415/2021,UIDB/50025/2020,UIDP/50025/2020)。
文摘Photonic technology combined with artificial intelligence plays a key role in the development of the latest smart system trends,integrating cutting-edge technology with machine learning models.This paper proposes a transmission-reflection analysis based system using dielectric nanoparticle-doped fiber combined with artificial intelligence to address one of the major problems in the distributed sensing approach:reducing the cost while maintaining high spatial resolution to close the gap between distributed sensors and the general public.Machine learning-based models are designed to classify the perturbed positions when the same force is used and force regression when different forces are applied on each position.The results show an accuracy of 99.43%in the position classification of multiple disturbances and an rms error of 1.53 N in the force regression,which represents 5%of the force range.In addition,a smart environment using the current system is proposed,which presented 100%accuracy in identifying the positions of different persons in the environment.This smart environment enables remote home care of patients with high reliability,intelligent decision-making,and a predictive capability.
