Null Space-Based Behavioral Control Applied to Multirobot Systems
Name: MAURO SERGIO MAFRA MOREIRA
Publication date: 18/12/2025
Advisor:
Name |
Role |
|---|---|
| MARIO SARCINELLI FILHO | Advisor |
Examining board:
| Name |
Role |
|---|---|
| DANIEL KHEDE DOURADO VILLA | Coorientador |
| JOSE LEANDRO FELIX SALLES | Examinador Interno |
| KIMON VALAVANIS | Examinador Externo |
| MARIO SARCINELLI FILHO | Presidente |
| PEDRO CASTILLO | Examinador Externo |
Pages
Summary: This work proposes controllers based on task/behavior prioritization to guide a formation of homogeneous or heterogeneous robots in working environments with large spatial occupancy and cargo transport, such as automated warehouses. This scenario can be found in logistics companies like Amazon, DHL, and postal office services. In all cases, a formation controller based on the virtual structure paradigm is used. Such a virtual structure consists of an imaginary line connecting the two robots. In this context, the formation is represented by a set of formation variables derived from the robots’ positions. The formation variables are grouped into two sets, one associated with the shape and the other with the position of the virtual structure. Once a formation control structure is defined, this work can be divided into three phases. The first phase focuses on providing the elements of the formation with obstacle avoidance capability. The second phase proposes an alternative formation control approach, more suitable for scenarios where an aerial robot is positioned above a ground-based differential traction robot. In the third phase, we have a formation consisting of aerial robots, specifically quadrotors. In this context, the focus is on cargo transportation by the formation elements, adopting two approaches: the first considers the cargo as part of the formation, and the second considers the cargo as a disturbance that must be rejected. In the first phase, we address obstacle avoidance using a null-space-based behavioral approach to autonomously guide a formation composed of a differential drive robot and an unmanned aerial vehicle navigating among obstacles modeled as potential fields. The obstacle position is fixed but unknown to the path planner; therefore, the proposed controller must be able to avoid possible collision when an increase in the potential field is perceived. In the second phase, the goal is to position an aerial robot above a ground robot in an inventory task inside an automated warehouse. The ground robot acts as support for the aerial one; therefore, the aerial robot should be positioned above the ground robot while reading binary codes (QR Codes, Barcodes). In the mathematical modeling typically employed and in previous works, a singularity occurs when using Euler angles. In this context, an alternative set of formation variables is proposed, where the singularity is shifted to a point in the XY plane. Finally, in the third phase, the study focuses on cargo transport, with two sub-scenarios: the first where the cargo is part of the formation, and the second where the cargo is treated as a disturbance. In the first sub-scenario, a controller based on the virtual structure paradigm is proposed, considering two quadrotors transporting a cargo suspended by cables. The objective is to control the movement of the cargo by controlling the triangle formed by the elements of the formation, specifically the quadcopter-quadcopter-cargo triangle. A behavioral control
technique based on null space is used, prioritizing the preservation of the formation’s shape and orientation. In the second sub-scenario, a behavioral controller based on the null space is proposed again, to which a sliding mode control law is associated, to reject the disturbance caused by the cargo. The work described here was validated through simulation and experimentation, and the results are detailed throughout this thesis.
