Analisis Kinematika Balik pada Kendali Robot Lengan Dental Light Berbasis Pengolahan Citra Digital Berdasarkan Isyarat Tangan

Abstract

Abstrak

Pemanfaatan kemajuan teknologi dalam bidang rekayasa robot manipulator dan pengolahan citra diterapkan dalam penelitian ini.Robot lengan enam DOF digunakan dalam membantu tindakan medis dokter dengan mengautomatisasi gerakan dari dental light berdasarkan isyarat jari tangan dokter lewat penggunaan pengolahan citra.Isyarat jari tangan dikenali sebagai instruksi untuk kemudian diolah sebagai koordinat kartesian.Koordinat yang didapat digunakan untuk mencari sudut sendi manipulator menggunakan algoritma kinematika balik metode analisis aljabar.Penyelesaian diawali dengan menentukan notasi Denavit-Hartenberg dari manipulator dilanjutkan dengan pencarian matriks transformasi homogen tiap sendi.Matriks-matriks diolah untuk mendapatkan solusi sudut sendi dari manipulator.Untuk mempermudah pengendalian ditambahkan subsistem focus pedal yang digunakan untuk mengendalikan gerak dari dua sendi terakhir bagian end-effector.

Uji coba sistem dilakukan terhadap kinerja algoritma kinematika balik dengan hasil selisih pada angka  jika dibandingkan dengan hasil metode referen lain.Uji coba hardware menunjukkan performansi yang baik dengan selisih nilai pada range -1 sampai 2 mm. Waktu eksekusi satu kali looping algoritma kinematika balik berada pada range 207-208 ms, sedangkan eksekusi program subsistem focus pedal berada pada range 36-38 ms. Komunikasi instruksi antar ketiga subsistem berjalan baik pada jarak 1 sampai 3.5 meter.

Kata kunci— pengolahan citra digital, DOF, algoritma kinematika balik, metode analisis aljabar, notasi Denavit-Hartenberg

Abstract

 The use of technological advances in the field of robot manipulator engineering and image processing was applied in this study. Robotic arm with six DOF is used for helping dental procedures by automatizing the movement of dental light based on dentist hand gesture using image processing. The hand gesture is recognized as an instruction for later use processed as cartesian coordinate. The coordinate is used to find the angle of the manipulator joints uses inverse kinematics algorithm with algebraic analysis method. The solution itself begins by determining Denavit-Hartenberg notation from the manipulator followed by searching of homogenous transformation matrix of each joint. Then, the matrix being calculated to obtain joint displacement solution from the manipulator. For easier control, one focus pedal sub system was added to control the movement of the last two joints of the end-effector.

System testing was done by testing the performance of inverse kinematics algorithm. The difference of those two methods is only  which shows the accuracy of the algorithm. Algorithm testing by hardware shows the good performance too with the difference between axis input value and axis testing value was in -1 to 2 mm range. The time needed for one execution of inverse kinematics algorithm was in 207-208 ms range and the execution of focus pedal instruction was in 36-38 ms range. Communication of instruction form focus pedal sub system ran well in the range of 1 metre to 3.5 metre.

Keywords—digital image processing, DOF, inverse kinematics algorithm, analysis algebra method, Denavit-Hartenberg notation