Alexander Alexopoulos

A Comparative Study of Collision Avoidance Techniques for Unmanned Aerial Vehicles

In this paper three collision avoidance methods for an unmanned aerial vehicle (UAV) are tested and compared to one another. The quadrocopter dynamic model with attitude and velocity controller, a trajectory generator and a selection of collision avoidance approaches were implemented. The first collision avoidance method is based on a geometric approach which computes a direction of avoidance from the flight direction and simple geometric equations. The second technique uses virtual repulsive force fields causing the UAV to be repelled by obstacles. The last method is a grid-based online path re-planning algorithm with A* search that finds a collision free path during flight. Various flight scenarios were defined including static and dynamic obstacles.

Modified particle petri nets for hybrid dynamical systems monitoring under environmental uncertainties

A real-time system in its environment constitutes a complex dynamic system (in general hybrid), which requires the development of new methods of modeling and monitoring. This paper presents the Modified Particle Petri nets (MPPN) approach, which combines Petri nets and particle filtering in order to model and monitor Hybrid Dynamical Systems (HDS) and their interaction with the environment. The approach considers the uncertainty of the process on both discrete and continuous variables as well as the environmental uncertainties. It presents relevant concepts of Petri nets within a probabilistic framework and how uncertainty is introduced (i) in the hybrid marking of the Petri net in order to represent an uncertain knowledge about all next possible hybrid states (ii) as well as in the process noise, which is incorporated into the filter. A marking updating of the Petri nets according to noisy measurements made by the system and a decision making algorithm are then used in order to estimate the hybrid system state, to detect inconsistencies and to attribute these to different kinds of faults. An example of mobile robot application modeled using MPPN is presented to show the usability of the approach.

Probabilistic online-generated monitoring models for mobile robot navigation using modified Petri net

This paper presents a generic hybrid online monitoring approach allowing the detection of inconsistencies in the navigation of autonomous mobile robots. The originalities of this work are (i) the association of classic state estimation based on particle filter to a special class of Petri net in order to deliver an estimation of the next robot state (position) as well as the environment state (graph nodes) and to use both information to distinguish between external noise influences, internals component faults or global behaviour inconsistency (ii) the integration of the geometrical and the logical environment representation into the monitor model in order to generate online the corresponding navigation monitoring model for the present mission trajectory while the system is running. The model takes simultaneously into account the uncertainty of the robot and of the environment through a unified modeling of both. To show the feasibility of the approach we apply it to an Intelligent Wheelchair (IWC) as a special type of autonomous mobile robot.

Cooperative pursue in pursuit-evasion games with unmanned aerial vehicles

This work tackles the problem of pursuit-evasion games between two pursuing and one evading unmanned aerial vehicle. The solution of this problem is derived by introducing a hierarchical decomposition of the game. On a superordinate collaboration level, the pursuers choose their optimal behavioral strategy (i.e. pursue and herd) resulting in a three-player non-cooperative dynamic game which is solved in a subordinate level of the overall game. This structure enables an intelligent behavior change for the pursuers based on game-theoretical solution methods. Depending on the state of the game it has to be evaluated which behavioral strategy yields the best results for the pursuers within the regarded time horizon. It is shown, that the pursuers outcome can be improved by using a superordinate cooperation between them. Furthermore, this application was implemented on a real quad-rotor system. In experiments the quad-rotor teamed-up with a software-emulated pursuer and it was shown that another software-emulated quad-rotor was caught. This framework provides a solution concept for all types of dynamic games with an arbitrary number of players and teams having a non-cooperative and cooperative nature. The hierarchical decomposition gives more flexibility and allows team members to co-ordinate their behavior and thus to maximize their outcome.

Decomposition of multi-player games on the example of pursuit-evasion games with unmanned aerial vehicles

In this work two different decomposition approaches for multi-player dynamic games are proposed. A game decomposition is sought which considerably reduces the time complexity of the solution. The run-time complexity of both the decomposed and the full game are compared. This is done by defining a two-pursuer one-evader pursuit-evasion game (PEG) with unmanned aerial vehicles (UAVs) with three different game structures, which are the Non-Decomposed Game, the Multiple Two-Player Game Decomposition (MTPGD), and the Encapsulated-Team Two-Player Game Decomposition (ETTPGD). In several simulations it is shown that with increasing cardinality of each players strategy space the MTPGD yields a relevant decrease of the run-time.

Mission-based online generation of probabilistic monitoring models for mobile robot navigation using Petri nets

This paper presents a generic hybrid monitoring approach, which allows the detection of inconsistencies in the navigation of autonomous mobile robots using online-generated models. A mission on the context of the navigation corresponds to an autonomous navigation from a start to an end mission point. The operator defines this mission by selecting a final goal point. Based on this selection the monitoring models for the current mission must be generated online. The originalities of this work are (i) the association of classic state estimation based on a particle filter with a special class of Petri net in order to deliver an estimation of the next robot state (position) as well as the environment state (graph nodes) and to use both pieces of information to distinguish between external noise influences, internal component faults and global behaviour inconsistency (ii) the integration of the geometrical and the logical environment representation into the monitor model (iii) the online generation of the corresponding monitoring model for the present mission trajectory while the system is running. The model takes simultaneously into account the uncertainty of the robot and of the environment through a unified modelling of both. To show the feasibility of the approach we apply it to an intelligent wheelchair (IWC) as a special type of autonomous mobile robot.

Associative memory for modified Petri-net based monitoring of mobile robot navigation

This paper presents an extension of a generic hybrid online monitoring approach for the navigation of autonomous mobile robots through an associative memory. Due to this extension, when a fault occurs the navigation process does not have to stop running in order to keep the system in a safe state and to ensure a high dependability. The associative memory enables the robot to remember the previously traversed paths inside a known environment and to learn new, unknown paths on-line. This long-term memory makes the monitor more dependable and more robust with respect to the fault detection, diagnosis and handling. To show the feasibility of the approach we apply it on an example of a mobile robot application.

Realization of pursuit-evasion games with unmanned aerial vehicles

Multi-agent pursuit-evasion games are a common way to describe problems in topics like police missions, surveillance, or warfare. Since many agents are involved in such problems, a game-theoretical solution seems reasonable. Game-theoretical solution approaches are generally very complex, especially in games with more than two agents, as it is the case in this paper. More and more unmanned aerial vehicles (UAV), in particular multi-rotors, are used in such applications because of their high maneuverability in all three dimensions. Especially in dangerous missions and applications UAVs excel, as they do not risk the life of the pilot. Tele-operated UAVs still require a pilot which has to be trained to do a task and a constant radio link between the pilot and the vehicle has to be maintained. An autonomous agent, able of performing the task of cooperative pursue and evasion can substitute the pilot. In this work, a realization of cooperative pursuit-evasion games with UAVs is presented. A real-time implementation is proposed, enabling a game-theoretical solution of the regarded problem. The applicability was proven during several real outdoors flight experiments with a hex-rotor UAV. Furthermore, the results obtained by the flight experiments were compared to simulations of the same games. Although, the most experimental results met the reference, disturbances like strong wind, wind gusts and sensor uncertainties affected the results significantly.

Non-linear model based control and parameter identification of a hex-rotor UAV

An implementation of a non-linear control scheme for hex rotors will be presented. The cascaded control scheme consisting of an attitude, a velocity and a position control level, empowers the hex rotor to be easily upgraded with more sophisticated behaviors. The attitude control utilizes the backstepping approach, which was introduced by Bouabdallah [1] for quad rotors, later improved and adapted to hex rotors by Arellano-Muro [2]. The implementation of the velocity and position controllers follow Voos [3]. Parameters, necessary for the model based control were identified numerically, with the help of direct measurement and by the use of flight experiment data and parameter estimation techniques. Validation of the controllers is done by analyzing real flight data, which was taken during outdoor flight experiments, testing both manual flight and position hold.

Real-Time Implementation of Pursuit-Evasion Games Between Unmanned Aerial Vehicles

The problem of two-playerpursuit-evasion games with unmanned aerial vehicles (UAVs) in a three-dimensional environment is tackled. A game-theoretical framework is presented, enabling the solution of dynamic games in discrete time. Depending on the cardinality of the action sets, the time complexity of solving such games could rise tremendously. Therefore, a tradeoff between available actions and computational time of the solution has to be found. It was shown that the chosen action space allows manoeuvres with sufficient accuracy, assuring the convergence of the games, while the computational time of the algorithm satisfies the real-time specifications. The UAVs taking part in the pursuit-evasion games are two identical quad-rotor systems with the same dynamical constraints, while the evaders’ absolute velocity is smaller than the pursuers’. The approach was simulated on an embedded computer and successfully tested for real-time applicability. Hence, the implementation and real-time execution on a physical UAV system is feasible.