Akademik Veri Yönetim Sistemi
Tez Türü: Doktora
Tezin Yürütüldüğü Kurum: Çukurova Üniversitesi, Mühendislik Fakültesi, Makina Müh., Türkiye
Tez Danışmanı: Mehmet İlteriş Sarıgeçili
Tezin Onay Tarihi: 2024
Tezin Dili: İngilizce
Desteklendiği Program: Öğretim Üyesi Yetiştirme Programı (ÖYP)
Özet:
In this thesis study, two-degrees-of-freedom, pneumatically
actuated wrist/forearm rehabilitation robot called CWRR has ben modified to
make it suitable for force/torque interaction control. For this aim, the
mechanical and control equipment of the CWRR robot has been redesigned,
manufactured and renewed. Extra sensors have been added to maximize the control
performance of the robot. After the completition of the modernization process,
flow modeling of the renewed precision pneumatic control valves has been
carried out. It has been determined that proportional directional control
valves did not comply with the existing flow models presented in the
literature, and a new and effective compressible flow model has been proposed
for proportional directional control valves. As the next stage, experimental
studies have been carried out to identify the friction force/torque of the
renewed pneumatic cylinders. At this point, novel automation algorithms for
friction force/torque identification that will provide faster and most accurate
friction parameter identification has been designed and implemented. In order
to increase the safety of the interacting human user and to make the robot more
compatible, an algorithm that detects human movement intentions has been
developed and incorporated into the controller structure. Afterwards, a fuzzy
logic adaptive PID controller structure has been designed that can track the
target interaction torque successfully during the executed workout. For this
purpose, fuzzy logic rules have been determined correctly, membership functions
and the boundaries of these functions have been determined according to
experimental data and experience of the expert. In order for the proposed robot
to take correct control actions even under disturbing effects, algorithms have
been developed that can reject both internal and external disturbance effects.
To demonstrate the superior performance of the proposed controller, two
different controllers have also been tested on the robot. One of these
controllers is the Default controller that is built-in the valve by the
manuafcturer and the other one is the conventional PID structure designed in
series with this Default controller. These three different controllers have
been tested in the real-time hardware-in-the loop environmet under four
different scenarios wherein disturbance-free or under-disturbance conditions
were evaluated. It has been confirmed by the tests that the proposed Cascade
Fuzzy Adaptive PID structure have exhibited outstanding control performance in
all scenarios compared to its counterparts.