Nonlinear Model Predictive Control Berbasis Metode Pseudospektral Chebyshev-Lagrange Pada Kendali Haluan Kapal Untuk Menghindari Tabrakan

Situmorang, Agustinus Gabriel (2026) Nonlinear Model Predictive Control Berbasis Metode Pseudospektral Chebyshev-Lagrange Pada Kendali Haluan Kapal Untuk Menghindari Tabrakan. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002221012-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Tabrakan kapal merupakan kecelakaan maritim dengan frekuensi tertinggi, sehingga menuntut sistem kendali yang responsif terhadap dinamika nonlinier dan batasan operasional. Penelitian ini mengusulkan perancangan kendali haluan kapal menggunakan metode Nonlinear Model Predictive Control (NMPC) untuk menyelesaikan masalah pelacakan lintasan dan penghindaran tabrakan terhadap objek dinamis. Untuk mengatasi beban komputasi pada NMPC, permasalahan optimasi kendali diformulasikan dalam fungsi objektif kontinu dan diselesaikan menggunakan metode pseudospektral Chebyshev-Lagrange. Metode ini mendiskritisasi sistem melalui titik kolokasi Chebyshev Gauss-Lobatto serta kuadratur Clenshaw-Curtis, sehingga mentransformasi masalah menjadi Nonlinear Programming (NLP). Integrasi mekanisme penghindaran tabrakan dilakukan dengan menerapkan aturan International Regulations for Preventing Collision at Sea (COLREGS) yang mencakup skenario overtaking, head-on, dan crossing sebagai batasan sistem. Simulasi dievaluasi dalam dua kondisi, yaitu kondisi tanpa halangan (dengan variasi horizon prediksi Tp) dan kondisi dengan halangan. Pada kondisi tanpa halangan, metode yang diusulkan dibandingkan dengan NMPC konvensional. Hasil simulasi menunjukkan bahwa nilai Root Mean Square Error (RMSE) pada NMPC berbasis metode pseudospektral Chebyshev-Lagrange tidak memiliki perbedaan signifikan dibandingkan NMPC konvensional, namun mampu mengurangi waktu komputasi hingga 83,91%. Sementara itu, pada kondisi dengan halangan, sistem berhasil menghindari tabrakan dengan objek dinamis secara aman sesuai peraturan COLREGS, di mana jarak minimum yang tercapai pada ketiga skenario tetap memenuhi batasan yang ditetapkan.
======================================================================================================================================
Ship collision is recorded as a maritime accident with the highest frequency, thus demanding a control system that is responsive to nonlinear dynamics and operational constraints. This research proposes the design of ship heading control using the Nonlinear Model Predictive Control (NMPC) method to solve the problems of trajectory tracking and collision avoidance against dynamic objects. To overcome the computational burden on NMPC, the control optimization problem is formulated in a continuous objective function and solved using the Chebyshev-Lagrange pseudospectral method. This method discretizes the system through Chebyshev-Gauss-Lobatto collocation points and Clenshaw-Curtis quadrature, thereby transforming the problem into Nonlinear Programming (NLP). The integration of the collision avoidance mechanism is carried out by applying the International Regulations for Preventing Collision at Sea (COLREGS) rules which include overtaking, head-on, and crossing scenarios as system constraints. Simulations are evaluated in two conditions, namely the condition without obstacles (with variations in the prediction horizon Tp) and the condition with obstacles. In the condition without obstacles, the proposed method is compared with conventional NMPC. Simulation results show that the Root Mean Square Error (RMSE) value in NMPC based on the Chebyshev-Lagrange pseudospectral method does not have a significant difference compared to conventional NMPC, but is able to reduce computational time by up to 83,91%. Meanwhile, in the condition with obstacles, the system successfully avoids collisions with dynamic objects safely according to COLREGS regulations, where the minimum distance achieved in the three scenarios still meets the established constraints.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Kendali Haluan Kapal, Metode Pseudospektral Chebyshev-Lagrange, Nonlinear Model Predictive Control, Penghindaran Tabrakan, Collision Avoidance, Chebyshev-Lagrange Pseudospectral Method, Nonlinear Model Predictive Control, Ship Heading Control.
Subjects:	Q Science > Q Science (General) > Q180.55.M38 Mathematical models T Technology > TJ Mechanical engineering and machinery > TJ217.6 Predictive Control T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL152.8 Vehicles, Remotely piloted. Autonomous vehicles.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Agustinus Gabriel Situmorang
Date Deposited:	21 Jul 2026 02:28
Last Modified:	21 Jul 2026 02:28
URI:	http://repository.its.ac.id/id/eprint/135174

Actions (login required)

View Item