2022 IB Diploma Extended Essays

ROS explanation Holonomic refers to the relationship between controllable and the degrees of freedom available by a system. If the relationship between controllable degrees of freedom and the degrees of freedom available is equal then the system is said to be holonomic. A holonomic system like the caster wheel or omni-wheel offer the manoeuvrability where they can move freely in any direction. And as such the controllable degrees of freedom is equal to total degrees of freedom. Non Holonomic is if the controllable degree of freedom is less than the total degrees of freedom. A robot like car has three degrees of freedom which include the position of the two axes and thereorientation. However you can only control two degrees of freedom which in cars are acceleration/braking or turning angle from the steering wheel. A ROS package is designed for this investigation. The purpose of the ROS package is intimidate a car-like robot in autonomous navigation. As such different ROS nodes are designed to be able to communicate with the sensors and the robot computer. The packages are added from teb_local_planner. This ROS package is used to create an online optimal local trajectory planner for mobile robot control and of autonomous navigation. The initial robot trajectory is computed by a global planner is further optimised during runtime w.r.t.. The runtime describes the set of software instructions that are executed while the program is being run, particularly the instructions that aren’t written in specifically for the program but are necessary for the execution of the code. The runtime w.r.t. causes a minimisation of the trajectory execution time (time-optimal solution), this implies avoidance from obstacles and conformity with kinodynamic constraints such as velocity and acceleration of the robot. The optimal trajectory of the robot is then computed by solving the scalarized multi-objective optimisation problem. Weights are then added to the optimisation problem in order to specify the behaviour of the autonomy needed. This approach to solve the optimal trajectory is known as Timed-Elastic-Band. These local path planners such as the Timed-Elastic-Band are often stuck within a locally optimal trajectory, meaning they aren’t able to avoid obstacles. As such an extension to the local planner is added. For this a subset of all applicable trajectories from different typologies is optimised in parallel . This results in the local planner being able to switch into the globally optimal trajectory from the subset. As such the robot is then able to avoid obstacles. The typologies are constructed by the robot through the idea of homology classes within ROS. The teb_local_planner implementation supports the kinematics of the non-holomic robots. As such