Akademik Veri Yönetim Sistemi
Tezin Türü: Doktora
Tezin Yürütüldüğü Kurum: Kocaeli Üniversitesi, Mühendislik Fakültesi, Mekatronik Mühendisliği, Türkiye
Tezin Onay Tarihi: 2021
Tezin Dili: İngilizce
Öğrenci: Zied Ben Hazem
Danışman: Zafer Bingül
Özet:
RIPS is one of the fundamental problems in the control
theory field. To verify the modern control theory, RIPS may be considered as a
better example in control engineering. It is the best model for the attitude
control such as space booster, rocket, satellite, aircraft stabilization in the
turbulent air-flow, humanoid robots, etc…. The RIPS is a highly non-linear and
open-loop unstable system that makes the control more challenging. It is an
intriguing subject from the control point of view due to its intrinsic nonlinearity.
The RIPS include a nonlinearity due to the frictions in the joints. Common
control approaches require a good knowledge of the frictions in the joints of
the system and accurate friction estimation to obtain the desired performances
of feedback controllers. However, the frictions have high non-linear values,
which result in steady-state errors, limit cycles, and poor performance of the
system. It has an influence on the system's response, and it should be
considered seriously. Therefore, friction estimation has the potential to
ameliorate the quality and dynamic behavior of the system.
One of the aims of this thesis is to estimate the
nonlinear frictions in the triple link rotary inverted pendulum. In this
research, novel NFFEMs are developed to estimate the joint friction
coefficients of three link rotary pendulum and compared with AFEMs. The
different versions of AFEMs and NFFEMs are generated based on each of the
following friction estimation models: NCFM, LFM, and NLFM. The aim of this
friction study is to obtain joint friction models which depend on both velocity
and acceleration in a large range of motion trajectory that involves difficult
and sudden large changes. In the proposed NFFEMs, joint velocities and
accelerations of the TLRIP are used as the input variables of the NF system
trained by using a RBNN. Several experiments are conducted on the TLRIP system
to verify the NFFEMs. In order to determine the estimation performance of the
friction models, total RMSEs between position simulation results obtained from
each joint friction model and encoders in the experimental setup are computed.
Based on the position RMSEs, the NFFEMs produce better estimation results than
the AFEMs. Among the novel NFFEMs, the NFNLM gives the best results.
Another aim of this thesis is to develop non-linear
controllers for the stabilization and anti-swing up control problems. PID, LQR
and swing-up based LQR controllers are developed for the stability control of
the SLRIP. Moreover, FLQR and FLQG controllers are developed for the stability
control of DLRIP and TLRIP. The aim of the stability control is to study the dynamic
performance of both FLQR and FLQG controllers and to compare them with the
classical LQR and LQG controllers, respectively. To determine the control
performance of the controllers, Ts, PO, Ess, MP and the total RMSEs of the
joint positions are computed. Furthermore, the dynamic responses of the
controllers were compared based on robustness analysis under internal and
external disturbances. To show the control performance of the controllers,
several simulations were conducted. Based on the comparative results, the
dynamic responses of both FLQR and FLQG controllers produce better results than
the dynamic responses of the classical LQR and LQG controllers, respectively.
Moreover, the robustness results indicate that the FLQR and FLQG controllers
under the internal and external disturbances were effective. Furthermore, an anti-swing up control of the SLRIP,
DLRIP and TLRIP is developed. To determine the control performance of the
anti-swing up controllers, different control parameters are computed, such as
Ts, MP, Ess, and RMSEs of the joint positions. Based on the comparative
results, the LQR controller produces better results than the classical PID for
the SLRIP. Moreover, a novel RBNF-LQR controller is developed for an anti-swing
up control of the DLRIP and TLRIP. The objective of this research is to study
the RBNF-LQR controller and to compare it with the FLQR and the LQR
controllers. In the proposed RBNF-LQR controllers, the positions and velocities
of state variables multiplied by their LQR gains are trained by using RBNNs
architecture. The outputs of the two RBNNs are used as the input variables of
the fuzzy controller. The novel architecture of the RBNF controller is
developed in order to obtain better control performance than the classical
ANFIS. To show the control performance of the anti-swing up controllers,
simulation and experiments results are given and compared. According to the
comparative results, the RBNF-LQR anti-swing up controller produces better
results than FLQR and LQR. Furthermore, the performance of the three
controllers developed was compared based on robustness analysis under external
disturbance. The results obtained here indicate that the RBNF-LQR anti-swing up
controller produces better performance than others in term of vibration
suppression capability.