Herlambang, Teguh (2016) Desain dan analisa sistem gerak autonomous underwater vehicle. Doctoral thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
4113301001-desertation thesis.pdf Download (5MB) | Preview |
Preview |
Text
4113301001-desertation-paperpdf.pdf Download (2MB) | Preview |
Preview |
Text
4113301001-desertation-presentationpdf.pdf Download (1MB) | Preview |
Abstract
Autonomous Underwater Vehicle (AUV) merupakan salah satu jenis robot
bawah air yang relatif flexibel untuk eksplorasi bawah laut dan peralatan sistem
pertahanan bawah laut. AUV dikendalikan dan dikemudikan oleh komputer di
atas kapal pendukung untuk melaju dan bergerak dengan enam derajat kebebasan
(6-DOF). Untuk mengendalikan dan mengemudikan AUV dibutuhkan sistem
navigasi, panduan dan kendali gerak. Pada penelitian ini dikembangkan sistem
navigasi, panduan dan kendali gerak menggunakan model linier Segorogeni
AUV. Pemodelan diawali dengan penyusunan formulasi matematika gerakan
AUV, berupa model non-linier 6-DOF yang selanjutnya dilakukan linierisasi
model non-linier 6-DOF dengan matriks Jacobi. Model linier ini digunakan
sebagai platform untuk mengembangkan sistem navigasi, panduan dan kendali
gerak. Untuk menjaga akurasi posisi secara terus menerus diterapkan estimasi
trajektori pada navigasi dan panduan AUV dengan metode estimasi, dengan
metode estimasi Ensemble Kalman Filter (EnKF) dan Ensemble Kalman Filter
Square Root (EnKF-SR). Untuk menjaga kestabilan AUV diterapkan sistem
kendali gerak dengan metode Proportional Integral Derivative (PID), Sliding
Mode Control (SMC) dan Sliding PID (SPID).
Pada penelitian ini dibuat sepuluh kasus lintasan yang harus dilalui AUV.
Lintasan pertama berupa gerakan maju lurus tanpa belok dan menyelam (diving),
lintasan kedua adalah gerakan belok tanpa diving, lintasan ketiga yaitu gerakan
diving tanpa belok, lintasan keempat merupakan gerakan belok dengan diving,
lintasan kelima adalah gerakan memutar tanpa diving, lintasan keenam berupa
gerakan memutar dengan diving, lintasan ketujuh yakni gerakan memutar ellips
tanpa diving, lintasan kedelapan yaitu gerakan memutar ellips dengan diving,
lintasan kesembilan adalah gerakan diving dan gerakan naik (emerging) tanpa
belok, sedangkan lintasan kesepuluh merupakan gerakan diving dan emerging
dengan belok. Berdasarkan metode yang dikembangkan, hasil simulasi
menunjukkan bahwa metode navigasi dan panduan EnKF, EnKF-SR dan KF
untuk model linier memiliki tingkat akurasi yang tinggi dalam memodelkan
kesepuluh lintasan, dan tidak melebihi maksimal error posisi dari desain kriteria,
yaitu 3.0%. Rata-rata error posisi untuk kesepuluh lintasan berkisar 0.1%-
0.99%. Hasil simulasi menunjukkan tingkat akurasi yang tinggi, sehingga dapat
disimpulkan bahwa implementasi dari metode-metode tersebut pada platform
Segorogeni AUV telah dilakukan secara tepat. Berdasarkan pembangkitan
ensemble, hasil simulasi sistem navigasi dan panduan dengan metode EnKF-SR
memiliki tingkat akurasi yang lebih baik daripada EnKF dan KF untuk kesepuluh
lintasan. Simulasi menggunakan metode EnKF-SR dengan membangkitkan 300
ensemble memiliki rata-rata akurasi yang lebih baik daripada dengan membangkitkan 100, 200 atau 400 ensemble. Sedangkan berdasarkan lintasan
kasus yang dibuat, hasil simulasi mendapatkan hasil yang baik dengan metode
EnKF-SR.
Hasil simulasi sistem kendali dengan metode SMC dan SPID menunjukkan
adanya error yang tidak jauh berbeda, yakni kurang dari 5%, serta memiliki
settling time yang cukup cepat, yaitu sekitar 1.0 detik. Sedangkan metode PID
mempunyai error yang lebih besar dari 5% dan settling time yang sangat lama,
sekitar 60 – 80 detik. Hal ini berarti bahwa hasil kendali gerak dengan metode
SMC dan SPID jauh lebih stabil daripada PID. Selanjutnya, analisa kestabilan
yang dilakukan dengan metode Lyapunov terhadap keseluruhan metode kendali
menunjukkan kondisi stabil asimtotik global.
======================================================================================================
Autonomous Underwater Vehicle (AUV) is one type of underwater robot
which is relatively flexible for undersea exploration and underwater defense
systems. AUV is controlled and driven by computer on the support vessel to
drive and move with six degrees of freedom (6-DOF). To control and navigate
AUV, systems of guidance and motion control are required. In this research, the
systems of navigation, guidance and motion control were developed using the
linear model of Segorogeni AUV. The modeling began with mathematical
formulation of AUV movement, in the form of non-linear models of 6-DOF, then
the non-linear 6-DOF was linearized with Jacobi matrix. This linear model was
used as a platform for developing navigation, guidance and motion control
systems. To maintain the continuous accuracy of the position, estimation of
trajectory was applied to the AUV navigation and guidance with the estimation
method, Ensemble Kalman Filter (EnKF) and Ensemble Kalman Filter Square
Root (EnKF-SR). To maintain the stability of AUV, the motion control systems
by the methods of Proportional Integral Derivative (PID), Sliding Mode Control
(SMC) and Sliding PID (SPID) were applied.
In this research, ten trajectory cases to be passed by AUV were prepared.
The first trajectory was forward straight movement without turning and diving.
The second trajectory was the movement of turning without diving. The third
trajectory was diving movement without turning. The fourth trajectory was
turning movement without diving. The fifth trajectory was twisting movement
without diving. The sixth trajectory was twisting movement with diving. The
seventh trajectory was ellipse like twisting without diving. The eighth trajectory
was ellipse-like twisting movement with diving. The ninth was diving and
remerging movement without turning whereas the tenth trajectory was diving and
emerging movement with turning. Based on the methods developed, the
simulation results showed that the methods of navigation and guidance of EnKF,
EnKF-SR and KF for the linear models had a high degree of accuracy in
modeling the tent trajectories and did not exceed the position maximum error
against the design criteria, which is 3.0%. The average error for the ten
trajectories ranged from 0.1% - 0.99%. The simulation results showed a high
degree of accuracy, so it could be concluded that the implementation of these
methods on the platform of Segorogeni AUV was appropriately carried out. Based ensemble generation, the simulation results showed that the navigation
and guidance systems by EnKF-SR method demonstrated a higher accuracy than
those by EnKF and KF for those ten trajectories. The simulations using EnKFSR
to generate 300 ensembles had an average accuracy, better than those by
generating a 100, 200 or 400 ensembles. Whereas, based on the trajectory cases
made, the simulation results showed that the good results were obtained by using
EnKF-SR method.
The simulation results of the control system by applying the SMC and
SPID methods demonstrated more or less similar error, for instance less than 5%,
and a fairly fast settling time, which is about 1.0 seconds. While the application
of PID method resulted in higher error, greater than 5% and the very long settling
time, about 60-80 seconds. This means that the results of motion control by SMC
and SPID methods were much more stable than those by PID. Furthermore, the
stability analysis conducted by Lyapunov method to the overall control methods
showed the global asymptotically stable condition.
Keywords: AUV, the linear model, Ensemble Kalman Filter (EnKF), Ensemble
Kalman Filter Square Root (EnKF), Kalman Filter (KF), Proportional Integral
Derivative (PID), Sliding Mode Control (SMC), Sliding PID (SPID), Lyapunov.
| Item Type: | Thesis (Doctoral) |
|---|---|
| Additional Information: | RDKe 623.820 5 Her d-1 |
| Uncontrolled Keywords: | AUV, model linier, Ensemble Kalman Filter (EnKF), Ensemble Kalman Filter Square Root (EnKF), Kalman Filter (KF), Proportional Integral Derivative (PID), Sliding Mode Control (SMC), Sliding PID (SPID), Lyapunov. |
| Subjects: | V Naval Science > VM Naval architecture. Shipbuilding. Marine engineering > VM365 Remote submersibles. Autonomous vehicles. |
| Divisions: | Faculty of Marine Technology (MARTECH) > Ocean Engineering > 38001-(S3) PhD Thesis |
| Depositing User: | - Taufiq Rahmanu |
| Date Deposited: | 26 Nov 2019 07:21 |
| Last Modified: | 26 Nov 2019 07:23 |
| URI: | http://repository.its.ac.id/id/eprint/72049 |
Actions (login required)
 |
View Item |