A Modular Energy Aware Framework for Multicopter Modeling in Control and Planning Applications

Sebastian Gasche, Christian Kallies, Andreas Himmel, Rolf Findeisen · 2025

Read A Modular Energy Aware Framework for Multicopter Modeling in Control and Planning Applications by Sebastian Gasche, Christian Kallies, Andreas Himmel, Rolf Findeisen (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-03256.txt.

Introduction

Sebastian Gasche and co-authors built a modular model in 2025 of small multirotor drones that tracks not just flight but energy, so planners can tell whether a mission fits the battery. Drones are now routine in inspection and rescue, but range guesses that ignore motors and sensors mislead.

Their kit models each part: the lithium battery, the speed controllers, and the brushless motors, all checked against real flight logs. Sensor models for laser scanners and cameras add the cost of the payload. The pieces snap together into one energy aware model of the whole craft. Different drones in a mixed swarm can be described by swapping parts. A companion study uses it to plan swarm search paths.

Catalog record

CREATORSebastian Gasche, Christian Kallies, Andreas Himmel, Rolf Findeisen
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE82 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-03256

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A Modular Energy Aware Framework for Multicopter Modeling in Control and Planning Applications Sebastian Gasche, Christian Kallies, Andreas Himmel, Rolf Findeisen Abstract Unmanned aerial vehicles (UAVs), especially multicopters, have recently gained popularity for use in surveillance, monitoring, inspection, and search and rescue missions. Their maneuverability and ability to operate in confined spaces make them particularly useful in cluttered environments. For advanced control and mission planning applications, accurate and resource-efficient modeling of UAVs and their capabilities is essential. This study presents a modular approach to multicopter modeling that considers vehicle dynamics, energy consumption, and sensor integration. The power train model includes detailed descriptions of key components such as the lithium-ion battery, electronic speed controllers, and brushless DC motors. Their models are validated with real test flight data. In addition, sensor models, including LiDAR and cameras, are integrated to describe the equipment often used in surveillance and monitoring missions. The individual models are combined into an energy-aware multicopter model, which provide the basis for a companion study on path planning for unmanned aircaft system (UAS) swarms performing search and rescue missions in cluttered and dynamic environments. The flexible modeling approach enables easy description of different UAVs in a heterogeneous UAS swarm, allowing for energy-efficient operations and autonomous decision making for a reliable mission performance. I Introduction Today, unmanned aerial vehicles are used in various industrial and civilian applications, including search and rescue, surveillance, and inspection missions, due to their flexibility, ease of deployment, and ability to access hard-to-reach areas. They are available in a wide range of designs, such as fixed-wing, rotary-wing, and hybrid configurations, each suited to specific mission profiles [ 36 ] . Among them, multicopters are notable for their high maneuverability, stability, and ability to hover, making them particularly effective in cluttered environments such as urban, forested, or disaster-affected areas. Motivated by their suitability for tasks in confined and complex environments where other types of unmanned aerial vehicles may struggle, this study focuses on multicopters [ 26 ] . Many tasks such as mission planning, path planning, control, and state estimation require accurate models of the unmanned aerial vehicle’s dynamics to ensure that the unmanned aerial vehicle performs as expected. This includes the unmanned aerial vehicle’s vehicle dynamics, such as forces and moments acting on the body, as well as rotational and translational motions. In addition, its interaction with the environment, including disturbances such as wind or turbulence, must also be taken into account. Besides the vehicle dynamics, the dynamics of its energy consumption is often important to model. As the energy consumption has a direct influence on the flight time, an accurate modeling enables to realistically estimate the flight endurance and the mission feasibility. These models are central to several control and optimization methods, such as model predictive control, which require accurate models of the system to predict future states and adjust controls accordingly [ 43 , 24 , 35 ] . For instance, in our companion study [ 14 ] on path planning for a heterogeneous unmanned aircraft system swarm with the goal of planning flight paths for search and rescue missions, models of the unmanned aerial vehicles’ vehicle dynamics and energy consumption are used. The path planning algorithm is based on model predictive control and mixed integer linear programming, which means that a mathematical model of the vehicle, the environment, and the mission goal is needed for path planning and decision making. Based on these models, the future behavior of the unmanned aircraft systems is predicted and optimized in terms of mission success, energy efficiency, and safety, enabling cooperative and sustainable swarm guidance. Furthermore, the estimation of the remaining energy capacity of the unmanned aerial vehicles, which is integrated into the decision making, enables the autonomous return of discharged unmanned aerial vehicles to their landing sites for recharging. Other path planning algorithms, such as rapidly-exploring random trees [ 20 , 19 ] and genetic algorithms [ 33 , 23 ] , also benefit from such models to incorporate motion constraints during sampling. Sensors, such as cameras and LIDARs, which are common payloads for unmanned aircraft system missions, also impose constraints on unmanned aircraft system operations to ensure valid data collection. Therefore, the effects of the sensors on the dynamics and energy consumption of the unmanned aerial vehicle should be considered for accurate planning tasks. In modeling, there are three primary types of models that differ in required knowledge of the inner dynamics of the system, required data on the system response, and descriptiveness. White-box models , also known as physics-based or first-principles models, analytically model system behavior based on physical equations where all parameters must be known. While highly interpretable and providing detailed insight into underlying physics, they require precise knowledge of all system components, challenging for complex systems. They offer high accuracy when the system is well understood but can be computationally expensive, limiting their use in real-time applications. Black-box models rely purely on input-output data without requiring detailed understanding of inner system dynamics. These models use machine learning or system identification methods to model system behavior as a relation between input and output. They can capture complex nonlinear system dynamics and are easy to implement with large datasets but lack physical interpretability. Additionally, their limited extrapolation capabilities result in inaccuracies when encountering unlearned situations, as accuracy depends on training data quantity and quality. Grey-box models combine elements of white-box and black-box models, using partial knowledge of inner dynamics while estimating unknown parameters or component dynamics from data. This offers a balance between accuracy and computational efficiency, particularly useful in practical applications, where some system knowledge is available but precise modeling of every component or process is impractical. However, determining the optimal model structure can be challenging, requiring both physical knowledge and data [ 2 , 41 ] . In this study, a grey-box modeling approach is used for its balance of accuracy, flexibility, and computational efficiency. It allows the incorporation of known unmanned aerial vehicle dynamics [ 13 , 7 , 31 , 40 ] , the approximation of difficult-to-model dynamics, and the consideration of uncertain or unknown parameters. In addition, we are aiming at a modular model for multicopters, which offers significant advantages in terms of flexibility and configurability. Multicopters can vary widely in the number of rotors, payloads, sensors, and battery capacities. A modular approach allows components such as motors, propellers, batteries and sensors to be easily reconfigured or replaced without overhauling the entire model. This is particularly useful when planning missions where the unmanned aerial vehicle may need to be configured to perform different tasks or operate in different environments. For example, by swapping out modules such as batteries or motors, the model can be adapted for missions that require heavier payloads or those that prioritize longer endurance. unmanned aerial vehicles are typically driven by electric motors and propelled by electrical energy stored in a battery. As their size increases, there are also unmanned aerial vehicles that are propelled by fossil fuels in combination with combustion engines or jet engines. However, only electric-propelled unmanned aerial vehicles are considered in this study. Numerous modeling approaches have been proposed to describe the energy consumption of electric-propelled unmanned aerial vehicles, ranging from high-level empirical

Read the complete text