Virtual Decomposition Control of a Hydraulic Manipulator
Koivumäki, Janne (2012)
2012
Automaatiotekniikan koulutusohjelma
Automaatio-, kone- ja materiaalitekniikan tiedekunta - Faculty of Automation, Mechanical and Materials Engineering
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2012-12-05
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201302081046
https://urn.fi/URN:NBN:fi:tty-201302081046
Tiivistelmä
The controller design and modelling of hydraulically driven robots is a challenging task. This comes inter alia due to inherent nonlinear dynamics associated with hydraulic actuators, highly nonlinear characteristic of the robot dynamics and various uncertainties and disturbances of mathematical models.
In this master’s thesis a new control theory, namely Virtual Decomposition Control (VDC), is studied. The VDC approach is developed especially for precision control of complex robots. In VDC approach the robotic system to be controlled is first virtually decomposed into subsystems. Then, the subsystems dynamics based control can be applied, to make each subsystem qualified to be virtually stable. Finally, the virtual stability of every subsystem results in the stability and convergence of entire robot. An effectiveness of this approach comes from the fact that no matter how complicated a robotic system is the dynamics of the subsystems remain relatively simple with fixed dynamic structures invariant to target systems.
The purpose of this thesis was to study and implement VDC into hydraulic 2-DOF manipulator actuated with hydraulic cylinders. The parameter adaptation for uncertain parameters was not studied in scope of this thesis. The objective of this thesis was also test performance of VDC-controller in practice and compare achieved results to corresponding PID-controller results.
The theory of VDC approach was successfully applied into studied manipulator and the L_2 and L_∞ stability of subsystems were mathematically guaranteed leading to stability of entire system. In experimental measurements certain Cartesian motion trajectory was driven with both VDC- and PID-controller. With VDC- controller roughly 7 times better piston position tracking performance was achieved for first cylinder and about 4.4 times better performance for second cylinder was achieved. Moreover, the very same Cartesian motion trajectory was driven with twice faster and half slower execution times. The stability of PID-controller was lost in both of these cases, whereas VDC-controller managed to drive these trajectories without problems.
In this master’s thesis a new control theory, namely Virtual Decomposition Control (VDC), is studied. The VDC approach is developed especially for precision control of complex robots. In VDC approach the robotic system to be controlled is first virtually decomposed into subsystems. Then, the subsystems dynamics based control can be applied, to make each subsystem qualified to be virtually stable. Finally, the virtual stability of every subsystem results in the stability and convergence of entire robot. An effectiveness of this approach comes from the fact that no matter how complicated a robotic system is the dynamics of the subsystems remain relatively simple with fixed dynamic structures invariant to target systems.
The purpose of this thesis was to study and implement VDC into hydraulic 2-DOF manipulator actuated with hydraulic cylinders. The parameter adaptation for uncertain parameters was not studied in scope of this thesis. The objective of this thesis was also test performance of VDC-controller in practice and compare achieved results to corresponding PID-controller results.
The theory of VDC approach was successfully applied into studied manipulator and the L_2 and L_∞ stability of subsystems were mathematically guaranteed leading to stability of entire system. In experimental measurements certain Cartesian motion trajectory was driven with both VDC- and PID-controller. With VDC- controller roughly 7 times better piston position tracking performance was achieved for first cylinder and about 4.4 times better performance for second cylinder was achieved. Moreover, the very same Cartesian motion trajectory was driven with twice faster and half slower execution times. The stability of PID-controller was lost in both of these cases, whereas VDC-controller managed to drive these trajectories without problems.