Pemodelan dan Pengendalian Two-Wheeled Self Balancing-Robot dengan Metode Pengendali Linear Quadratic Regulator

Shiddieqy, Ilham Maulana Ash (2019) Pemodelan dan Pengendalian Two-Wheeled Self Balancing-Robot dengan Metode Pengendali Linear Quadratic Regulator. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 02111540000071-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Penelitian tentang robot saat ini sudah berkembang sangat pesat di seluruh dunia. Salah satu pengaplikasian robot yang paling populer adalah two-wheeled self-balancing robot yang digunakan manusia dalam hal transportasi yang praktis, ramah lingkungan, dan dapat melewati berbagai medan tertentu. Pada dasarnya two-wheeled self-balancing robot merupakan inverted pendulum yang dihubungkan dengan dua roda. Perlu adanya pengendalian terhadap inverted pendulum agar dapat mempertahankan posisinya pada titik equilibrium terhadap tanah secara optimal. Pada penelitian ini, dalam perancangan sistem kendali full state - feedback akan dilakukan pada two-wheeled self-balancing robot. Perhitungan dilakukan dengan menurunkan persamaan gerak dan selanjutnya linearisasi dilakukan pada titik equilibrium. Lalu, parameter dari sistem didapatkan dari perangkat lunak dan dari pengukuran yang kemudian dimasukkan dalam persamaan. Langkah selanjutnya, merancang sistem kendali dengan metode Linear Quadratic Regulator untuk mencari gain dengan melakukan pembobotan pada matriks Q dan R sehingga didapatkan nilai pembobotan yang optimal. Terdapat tiga variasi sistem kendali dan tiga variasi pengujian yaitu pemberian disturbance, sudut pitching awal, dan input berjalan, lalu dilanjutkan dengan analisis respon. Hasil pengujian menunjukkan two-wheeled self-balancing robot, seluruh variasi pengendali LQR yang diuji dapat mempertahankan robot pada kondisi stabil. Variasi pengendali 2 memiliki stabilitas yang paling baik dan variasi pengendali 3 memiliki stabilitas yang paling buruk. Dengan memperbesar pembobotan matriks Q, nilai gain akan meningkat yang memberikan dampak baik dalam menjaga stabilitas robot untuk tetap pada titik equilibrium. Sebaliknya dengan memperbesar pembobotan matriks R, nilai gain akan menurun yang berakibat pada melemahnya stabilitas robot. Selain itu, metode LQR yang didesain memiliki respon yang lebih baik jika dibandingkan dengan pengendali PID.
===================================================================================================================================
Research on robots is now growing very rapidly throughout the world. One of the most popular applications of robots is two-wheeled self-balancing robot that humans use in terms of practical, environmentally friendly transportation, and can pass through a variety of specific fields. Basically, a two-wheeled self-balancing robot is an inverted pendulum that is connected with two wheels. There is a need for control of the inverted pendulum in order to maintain its position at the equilibrium point to the ground optimally. In this study, the design of a full state-feedback control system was carried out on two-wheeled self-balancing robots. Calculations are carried out by reducing the equation of motion and then linearization is carried out at the equilibrium point. Then, the parameters of the system are obtained from the software and from the measurements which are then included in the equation. The next step is to design a control system using the Linear Quadratic Regulator method to find gain by weighting the Q and R matrices so that the optimal weighting value is obtained. There are three variations of control systems and three variations of testing, namely giving disturbances, initial pitching angles, and walking inputs, then followed by analysis of responses. The test results show two-wheeled self-balancing robots, all variations of LQR controllers tested can maintain the robot in a stable condition. Variation 2 has the best stability and variation 3 has the worst stability. By increasing the weighting of the Q matrix, the gain value will increase which has a good impact on maintaining the stability of the robot to remain at the equilibrium point. Conversely, by increasing the weighting of the matrix R, the gain value will decrease which results in the weakening of the stability of the robot. In addition, the LQR method designed has a better response compared to PID controllers.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Inverted Pendulum, Two-Wheeled Self-Balancing Robot, State-feedback, LQR, PID
Subjects:	T Technology > TJ Mechanical engineering and machinery T Technology > TJ Mechanical engineering and machinery > TJ223.P76 Programmable controllers
Divisions:	Faculty of Industrial Technology > Mechanical Engineering > 21201-(S1) Undergraduate Thesis
Depositing User:	Ilham Maulana Ash Shiddieqy
Date Deposited:	22 Jul 2026 01:24
Last Modified:	22 Jul 2026 01:24
URI:	http://repository.its.ac.id/id/eprint/67167

Actions (login required)

View Item