Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Powiadomienia systemowe

Sesja wygasła!

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: inverse dynamic problem

Sortuj według:

Ogranicz wyniki do:

Application of direct and inverse kinematics and dynamics in motion planning of manipulator links

Delyová Ingrid, Hroncová Darina, Frankovský Peter, Sivák Peter, Kostka Ján, Neumann Vojtech

International Journal of Applied Mechanics and Engineering

2023

Vol. 28, no. 3

53--64

For the synthesis of manipulators and robots, an accurate analysis of the movements of the individual links is essential. This thesis deals with motion planning of the effector of a multi-linked manipulator. An important topic in this area is the orientation and position of links and kinematic pairs in space. In particular, attention should be paid to the position of their endpoint as well as other significant points. Trajectory planning allows the manipulator to perform complex tasks, such as picking and placing objects or following a particular path in space. Overall, trajectory planning of a multibody manipulator involves a combination of direct and inverse kinematics calculations, as well as control theory and optimization techniques. It is an important process enabling manipulators to perform complex tasks such as assembly, handling and inspection. In the design of robot kinematic structures, simulation programs are currently used for their kinematic and dynamic analysis. The proposed manipulator was first solved by inverse kinematics problem in Matlab. Subsequently, the trajectories of the end-effector were determined in Matlab by a direct kinematics problem. In Simulink, using the SimMechanics library, the inverse problem of dynamics was used to determine the trajectories of the moments.

Using Tikhonov regularization to improve estimation of robot position based on uncertain robot model obtained by neural network

Możaryn J., Kurek J. E.

Pomiary Automatyka Kontrola

2009

R. 55, nr 3

198-204

A method for improvement of a position estimation of a robot manipulator based on model with uncertain parameters is presented. To calculate the position of the robot there was designed the robot model using artificial neural networks with structure of the mathematical model in the form of Lagrange-Elder equations. The Tikhonov regularization was then used to improve the approximation of the robot's position. The example of the position of the robot PUMA 560 with 6 degrees of freedom calculation with proposed method is presented. Obtained results indicate significant improvement of the estimation.

W pracy przedstawiono metodę poprawy estymacji położeń robota na podstawie modelu robota o niedokładnych parametrach. Do wyznaczania położenia robota zaprojektowano model robota z wykorzystaniem sztucznych sieci neuronowych o strukturze modelu matematycznego w formie równań Lagrange'a-Eulera. W celu poprawy estymacji położeń na podstawie wyznaczonego modelu zastosowano regularyzację Tikhonowa. Zaproponowana metoda została przedstawiona na przykładzie odtwarzania położeń robota PUMA 560. Otrzymane wyniki wskazują na znaczną poprawę dokładności.