I am interested working in the fields of robotics and multibody dynamics, particularity,
- Modeling of complex robotic and multibody systems
- Path planning of space robots
- Design of compliant wheeled robot
- Vision-based control of robots
- Motion planning of legged robot
Reactionless visual servoing
In this work a novel visual servoing controller for a satellite mounted dual-arm space robot. The controller is designed to complete the task of servoing the robot’s end-effectors to the desired pose, while regulating orientation of the base-satellite. Task redundancy approach is utilized to coordinate the servoing process and attitude of the base satellite. The visual task is defined as a primary task, while regulating attitude of the base satellite to zero is defined as a secondary task. The secondary task is formulated as an optimization problem in such a way that it does not affect the primary task, and simultaneously minimizes its cost function. A set of numerical experiments are carried out on a dual-arm space robot showing efficacy of the proposed control methodology.
Step climbing of redundant wheeled robot
This work focuses on development of a redundant wheeled robot for climbing steep obstacles. The proposed design of the robot has five modules, where each module consists of a rigid link and two wheels. The joints connecting these modules are passive whereas the only actuation is provided at the wheels. The use of compliant elements like springs at the passive joints is proposed in this work for climbing steps of varying heights. A framework is presented for choosing appropriate spring stiffness by investigating the static stability of the robot during its ascent phase. In contrast to existing active suspension step climbers, the proposed mechanism does not need knowledge of a step height beforehand. The robot was simulated to climb steps of various heights with constant wheel velocity. Presently, the efforts are being made to implement the proposed algorithm for control of physical prototype of the robot and to achieve wheel torque optimization.
Autonomous capture of tumbling orbiting objects
In this work strategies were presented for point-to-point reactionless manipulation of a satellite mounted dual-arm robotic system for capture of tumbling orbiting objects, such as out-of-commission satellites and space debris. Use of the dual-arm robot could be more effective than the single arm when there is no provision for a grapple fixture or the object is tumbling. The dual arms can also provide dexterous manipulation. As the main objective in capture of orbital objects is to move the end-effector from initial position to the grapple point with desired velocity, the task-level reactionless constraints in terms of end-effector velocities are derived. The trajectory planned using these constraints, however, results in several singular points within the robot’s workspace. In order to overcome this shortcoming, three point-to-point path planning strategies are presented, which improve the reactionless operation of the dual-arm robot. The strategies are illustrated by carrying out simulations for a 6-degree-of-freedom (DOF) dual-arm robotic system mounted on a satellite.
Dynamics o chains with large degrees-of-freedom
This work investigates the dynamic behaviour of serial chains with degrees-of-freedom (DOF) as large as 1,00,000. A recursive solver called Recursive Dynamic Simulator (ReDySim), based on the Newton-Euler formulation and the Decoupled Natural Orthogonal Complement (DeNOC) matrices, was used to model and simulate the dynamics of these systems. Planar, as well as spatial motions of chains with moderate- (DOF≤1,000), large- (1000<DOF≤10,000) and huge- (DOF>10,000) DOF were simulated, and the results were validated by several means, such as: reported results wherever available, results obtained from commercial software, energy checks, etc. The study shows that ReDySim is capable of analyzing serial chains of huge-DOF with acceptable numerical accuracy. The scheme is found to be numerically stable as well as computationally efficient, owing to the linear-time computation of the joint accelerations. Numerical studies were conducted to establish the theoretical basis for better performance of the proposed ReDySim solver. With the demonstrated capabilities of ReDySim, it may be found suitable for a large number of applications involving serial systems with huge-DOF.
Reduced-order Forward Dynamics of Multi-Closed-Loop Systems (In collaboration with Majid Kaul, research scholar at IIT Delhi)
In this work, a reduced-order forward dynamics of multi-closed loop systems is proposed by exploiting inherent kinematic constraints at acceleration level associated with these systems. In the proposed forward dynamics formulation, Lagrange multipliers are calculated sequentially at subsystem level and later treated as external forces to the resulting serial or tree-type systems of the original closed-loop system, for the recursive computation of joint accelerations. The proposed methodology showed improvement in the overall numerical performances in addition to the reduced order computational complexity.
Modular Framework for Dynamic Modelling and Analyses Tree-type Robotic Systems
The aim of research was broadly classified into two parts. In the first part a novel concept of Euler-Angles-Joints (EAJs) was introduced for unified representation of multiple-Degrees-of-Freedom (multiple-DoF) joint. Modular framework for dynamic modelling was then proposed introducing the concept of kinematic modules, where each module is a set of serially connected links. Module-level analytical expressions of the generalized inertia matrix (GIM) were obtained for the first time which in turn helped in obtaining module-level decomposition and explicit inversion of the GIM, and inter-modular recursive algorithms. Comparison of the proposed algorithms was provided in terms of methodology used, advantages and computational complexity with the existing algorithms. Moreover, several module-level physical interpretations were also provided, which provide better insight into the dynamics involved and help in deciding any inconsistency in the dynamic behaviour of robots. The second part of the research focused on the dynamic analysis of various fixed- and floating-base robotics systems. Two model-based control schemes were presented using the proposed algorithms. Dynamics analyses of robotic gripper, multi closed-loop system, biped, hexapod, quadruped, and hyper-DoF system were presented, which would help researcher and practicing engineer to use these results for design, simulation and control.
Simulation and Control of monopod hopping
In this work, the strategies for hopping were applied on simulated planar one legged model to obtain velocity and direction control. The leg consists of three links with PD controlled actuators, the foot is rigid and assumed to be pointed, while the ground is assumed to be compliant. Two types of realistic foot ground interaction models were used which provided continuous forces and allowed slipping of leg on ground. Simulation results showed that the monopod hopping without falling was achieved successfully using both the foot ground interaction models. The monopod was able to hop periodically with the speed of 20 kmph for longer distance without falling. The velocity control and change of stable state were possible using the proposed control methodology for forward and backward hopping.