Yazar "Azeem, Hafiz Huzaifa" seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim Yok
    Öğe
    Estimating Environment Parameters for Teleoperation System with Time Delay
    (IEEE, 2022) Majeed, Fatimah Jabbar; Azeem, Hafiz Huzaifa; Baran, Fray A.
    This paper demonstrates how recursive least squares (RLS) can be used to estimate the parameters of remote environment contact force which can be used to render a virtual environment on the operator side in a bilateral teleoperation system. Proper and fast estimation of the remote environment impedance plays a crucial role in the realization of local force controllers for time delayed teleoperation systems. Addressing that challenge, in this paper, three variants of RLS estimators are implemented and compared against three different impedance models. The algorithms are tested in the simulation environment making use of a recorded real experiment data set. The force reconstruction performances are compared to evaluate the implemented models and estimators. Based on the simulation results, one of the estimators and one of the models are selected for experimental validation on a single degree of freedom motion control system. The results obtained from the experiments confirm how the estimated forces match with that of the actual force responses and provide promising potential for further application in local force controllers of the teleoperation systems.
  • Küçük Resim
    Öğe
    Time-delayed teleoperation with virtual environment reconstruction
    (İstanbul Bilgi Üniversitesi, 2021) Azeem, Hafiz Huzaifa; Baran, Abdurrahman Eray
    ABSTRACT: Teleoperations have achieved a milestone in the field of robotics, performing ex tremely difficult tasks in hazardous environments such as undersea, nuclear sites, space exploration and tele-surgery. Their ability to perform tasks previously performed by a master device and to execute commands through a distant slave has opened many doors in the field of advanced robotics. When the distance between the master and the slave is long, the communication channel is subjected to time delays. In this thesis, our motivation was to find a way to reduce the effects of the time delay problem by estimating the parameters of remote environment contact force and render ing a virtual environment on the master side in a bilateral teleoperation system. First, Deep Learning Algorithm, Long Short Term Memory (LSTM) was used to estimate the parameters of the remote environment. After careful consideration, it was concluded that LSTM was not suitable for real-time implementation. Following the progress, the use of Recursive Least Squares (RLS) is demonstrated for estimating the parameters of remote environment contact force and rendering a virtual environment on the master side in a bilateral teleoperation system. Proper and fast estimation of the remote environment impedance plays a crucial role in the realization of local force controller for time delayed teleoperation systems. Three different variants of RLS estimator were implemented and compared against three different impedance models. The algorithms were initially tested in the sim ulation environment making use of a recorded real experiment data set. The force reconstruction performances are compared to evaluate the implemented models and estimators. Based on the simulation results, one of the estimators and one of the mod els are selected for experimental validation on a single degree of freedom motion control system. In a set of real experiments performed, the estimated force was rendered on the master side as a virtual environment, this way a local force feedback control loop was estab lished. The local force control loop was based on the predicted environment using the estimated parameters of RLS running in real time. Therefore, the master side was able to feel same force felt at the slave side without any time delay.