Application of bidirectional rapidly exploring random trees (BiRRT) algorithm for collision-free trajectory planning of free-floating space manipulator

Abstract

On-orbit servicing and active debris removal missions will rely on the use of unmanned satellite equipped with a manipulator. Capture of the target object will be the most challenging phase of these missions. During the capture manoeuvre, the manipulator must avoid collisions with elements of the target object (e.g., solar panels). The dynamic equations of the satellite-manipulator system must be used during the trajectory planning because the motion of the manipulator influences the position and orientation of the satellite. In this paper, we propose application of the bidirectional rapidly exploring random trees (BiRRT) algorithm for planning a collision-free trajectory of a manipulator mounted on a free-floating satellite. A new approach based on pseudo-velocities method (PVM) is used for construction of nodes of the trajectory tree. Initial nodes of the second tree are selected from the set of potential final configurations of the system. The proposed method is validated in numerical simulations performed for a planar case (3-DoF manipulator). The obtained results are compared with the results obtained with two other trajectory planning methods based on the RRT algorithm. It is shown that in a simple test scenario, the proposed BiRRT PVM algorithm results in a lower manipulator tip position error. In a more difficult test scenario, only the proposed method was able to find a solution. Practical applicability of the BiRRT PVM method is demonstrated in experiments performed on a planar air-bearing microgravity simulator where the trajectory is realised by a manipulator mounted on a mock-up of the free-floating servicing satellite.

Sub-optimal motion planning of one-chained, two-input nonholonomic systems

“…Methods can be divided with respect to a quality of obtained solutions (not optimized, sub-or optimal ones) and how they incorporate extra constraints (on controls, state, obstacle avoidance) -explicit constraints or introduced via penalty into a cost function [2]. A model of a robot plays a crucial role in planning algorithms [3,4]. Discrete models tend to consider a robot as an automaton that changes its state at some asynchronous time stamps while continuous models describe a robot with differential equations.…”

“…Discrete models tend to consider a robot as an automaton that changes its state at some asynchronous time stamps while continuous models describe a robot with differential equations. Continuous models frequently result from constraints imposed on a robot (for example in the Pfaff form like for simple mobile robots: the unicycle or the kinematic car [1] or tractor pulling some trailers, or free-floating robots [3]). When the constraints are holonomic one, a dimension of a configuration space can be reduced by the number of constraints while for nonholonomic systems the constraints are not integrable and resulting systems (despite being controllable) are difficult to control because a number of controls is smaller than a configuration space dimension.…”

Application of the Obstacle Vector Field Method for Trajectory Planning of a Planar Manipulator in Simulated Microgravity

“…Recently, Rybus (2022) proposed the obstacle vector field (OVF) method. In the classical APF method, a scalar potential field is used, while in the OVF method, a vector field is constructed.…”

“…The guiding vector field is generated by solving a scalar boundary value problem and it has two components: one responsible for driving the gripper toward the desired position and the other one responsible for guiding the gripper around obstacles. A detailed discussion of similarities and differences between these methods and the OVF method was presented by Rybus (2022).…”

“…The OVF method was validated in numerical simulations performed for a planar 2 Degrees of Freedom (DoF) manipulator mounted on a free-floating chaser satellite (Rybus, 2022). This method allows to obtain the desired position of the gripper, but it does not allow to obtain the desired orientation of the gripper.…”