DYNAMIC MODELING AND CONTROL OF MOBILE SOFT CONTINUUM MANIPULATOR

Abstract

Bionic robots, imply the robots which are inspired by biological life forms. The manipulator robots inspired from elephant trunk are good example of bionic robots. These are continuous robots with bending backbones and high degrees of freedom. Ideally, the compliant nature of such manipulator delivers infinite degrees of freedom, which makes it difficult to control them at the space task. Bionic manipulators are becoming very popular these days due to light weight, higher compliance. The applications of bionic manipulators are found in material transport, pick and place operations. Their application is also explored in areas of skeletal trauma treatment, endoscopy and minimally invasive surgery and inspection. Modelling and control of a bionic manipulator is a challenging task due to number of redundancies present in the system. This thesis presents a detailed bond graph model of bionic manipulator. Initially a kinematic model is developed for planar bionic manipulator and cooperative planar bionic manipulator. Then the bond graph model for both the cases are developed. Then a three-dimensional multi-section bionic manipulator and cooperative bionic manipulators are considered. For both the cases first, a kinematic model is developed and then using that the bond graph models are developed. The three-dimensional multi-section bionic manipulator is modeled as a hybrid manipulator, i.e., a serial manipulator whose links/segments are formed by a parallel manipulator. The base of the mobile manipulator as well as the rotational motion of end-effector also have been modelled. The developed bond graph model is validated with MappleSim software. Further, to validate the modelling an example of bionic manipulator Robotino XT has been considered. For structure like bionic manipulator a model-based control scheme is appropriate as it can control the manipulator with greater accuracy. A model-based control scheme is developed for the bionic manipulator. The control scheme uses the inverse dynamics of manipulator segments to calculate the command torque components required for trajectory tracking. Proportional- Integral-Derivative (PID) controllers are then used along with the model-based controller for accurate trajectory tracking. First, model-based control of cooperative planar bionic manipulator has been considered. Then, model-based control of spatial bionic manipulator and cooperative spatial bionic manipulator have been considered. The control performance is studied for both cases through simulation and animation results of the bionic manipulator. To validate the modelling an example of bionic manipulator Robotino XT has been considered In addition, we developed tip trajectory controller for three-dimensional multi-section bionic manipulator. This is achieved by model-based controller algorithm. Model based control gives better performance than linear control for control of nonlinear system. The control scheme uses the inverse dynamics of manipulator segments to calculate the command torque components required for trajectory tracking. Proportional-Integral-Derivative (PID) controllers are then used along with the model-based controller for accurate trajectory tracking. The control performance is studied through simulation and animation results of the bionic manipulator. Finally, this thesis discusses the fault tolerant control of mobile manipulator using the redundancy in the system. First a Closed Loop Inverse Kinematics (CLIK) algorithm has been introduced. Then the trajectory control of spatial continuum manipulator under actuator failure has been studies. Also, the fault tolerant control of bionic manipulator through reconfiguration has been done.