Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10155980
Junhak Yi, Donghee Lee, Wooryong Park, Woohyun Byun, Soobin Huh, Woochul Nam
Recently, various tracking flight techniques of unmanned aerial vehicle (UAVs) have been developed and used in various applications. However, a proximity tracking flight is still challenging because accurate estimation of the position and velocity of a target ground vehicle (GV) is difficult. This paper presents an autonomous UAV system that can fly close to GVs. If the relative position between the UAV and GV was used for flight control and velocity was not used, the tracking can be unsuccessful. To address this issue, the speed of the ground vehicle was also estimated, and it was feedforwarded into the control loop. Real flight experiments showed that this approach greatly improved the tracking performance; the UAV tracked the GV driving at approximately 4 m/s with an average displacement error of less than 1 m.
{"title":"Autonomous Control of UAV for Proximity Tracking of Ground Vehicles with AprilTag and Feedforward Control","authors":"Junhak Yi, Donghee Lee, Wooryong Park, Woohyun Byun, Soobin Huh, Woochul Nam","doi":"10.1109/ICUAS57906.2023.10155980","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10155980","url":null,"abstract":"Recently, various tracking flight techniques of unmanned aerial vehicle (UAVs) have been developed and used in various applications. However, a proximity tracking flight is still challenging because accurate estimation of the position and velocity of a target ground vehicle (GV) is difficult. This paper presents an autonomous UAV system that can fly close to GVs. If the relative position between the UAV and GV was used for flight control and velocity was not used, the tracking can be unsuccessful. To address this issue, the speed of the ground vehicle was also estimated, and it was feedforwarded into the control loop. Real flight experiments showed that this approach greatly improved the tracking performance; the UAV tracked the GV driving at approximately 4 m/s with an average displacement error of less than 1 m.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117265630","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10156448
G. Olivas-Martínez, H. Castañeda
This paper introduces a class of adaptive sliding mode controller for a quad-rotor unmanned aircraft vehicle. The control is based on a non-singular fast terminal surface, and an adaptive law involving only two parameters to be tuned, which produces a smoother gain dynamics. In return, a significantly reduction of undesired behavior such as chattering is achieved, while preserving the properties of robustness against perturbations and finite time convergence. Furthermore, in order to evaluate robustness, the proposed control technique along with a Von Kármán model based wind turbulence generator, are applied in a close-to-real-life scenario. This consists of a 310 meter trajectory inside a city block powered by Unreal Engine. Obtained results support the claim that this control scheme allows the quad-rotor to follow desired trajectories even in presence of wind perturbations. This displays the feasibility and robustness needed for such systems to enable more complex tasks while flying in urban environments.
{"title":"Adaptive Single-Gain Non-Singular Fast Terminal Sliding Mode Control for a Quad-rotor UAV Against Wind Perturbations","authors":"G. Olivas-Martínez, H. Castañeda","doi":"10.1109/ICUAS57906.2023.10156448","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156448","url":null,"abstract":"This paper introduces a class of adaptive sliding mode controller for a quad-rotor unmanned aircraft vehicle. The control is based on a non-singular fast terminal surface, and an adaptive law involving only two parameters to be tuned, which produces a smoother gain dynamics. In return, a significantly reduction of undesired behavior such as chattering is achieved, while preserving the properties of robustness against perturbations and finite time convergence. Furthermore, in order to evaluate robustness, the proposed control technique along with a Von Kármán model based wind turbulence generator, are applied in a close-to-real-life scenario. This consists of a 310 meter trajectory inside a city block powered by Unreal Engine. Obtained results support the claim that this control scheme allows the quad-rotor to follow desired trajectories even in presence of wind perturbations. This displays the feasibility and robustness needed for such systems to enable more complex tasks while flying in urban environments.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117244460","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10156365
Iris David Du Mutel de Pierrepont Franzetti, R. Parin, E. Capello
Multicopters are used for a wide range of applications that often involve approaching buildings or navigating enclosed spaces. Opposed to the freedom found in outdoor flights, indoor UAVs navigating close to surfaces must take into account the airflow variations caused by its rebound and identify them as disturbances to be compensated. A custom made quadcopter has been built for the evaluation of wall effect in climate controlled environments. Two different propeller sizes have been considered for testing.Climate variations consisting in changes of pressure, from 1000 mbar up to the equivalent pressure attained at 5000 m. A fixed 6DOF load cell has been used for the experiments, being able to log forces and moments in three orthogonal axes. The tests simulate a hovering UAV at different wall distances. The influence of the propeller size and air density on the wall effect has been also measured. Experimental data will be used for the definition of a mathematical model, in which the wall effect is considered.
{"title":"Wall Effect evaluation of small quadcopters in pressure-controlled environments","authors":"Iris David Du Mutel de Pierrepont Franzetti, R. Parin, E. Capello","doi":"10.1109/ICUAS57906.2023.10156365","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156365","url":null,"abstract":"Multicopters are used for a wide range of applications that often involve approaching buildings or navigating enclosed spaces. Opposed to the freedom found in outdoor flights, indoor UAVs navigating close to surfaces must take into account the airflow variations caused by its rebound and identify them as disturbances to be compensated. A custom made quadcopter has been built for the evaluation of wall effect in climate controlled environments. Two different propeller sizes have been considered for testing.Climate variations consisting in changes of pressure, from 1000 mbar up to the equivalent pressure attained at 5000 m. A fixed 6DOF load cell has been used for the experiments, being able to log forces and moments in three orthogonal axes. The tests simulate a hovering UAV at different wall distances. The influence of the propeller size and air density on the wall effect has been also measured. Experimental data will be used for the definition of a mathematical model, in which the wall effect is considered.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116440449","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10156620
M. Nigro, F. Pierri, F. Caccavale, Markus Ryll
This work experimentally validates a novel fully-actuated quadrotor-based unmanned aerial vehicle named ODQuad (OmniDirectional Quadrotor). The ODQuad is composed of three main parts arranged in a gimbal configuration. The internal mechanism is composed of two rotational joints with orthogonal and incident axes which allow to decouple the horizontal motions from the vehicle body rolling and pitching. Firstly, the physical prototype is presented and the motion controller, inherited by [1], has been tailored in such a way to integrate the servo actuators of the internal gimbal mechanism. Three trajectories have been commanded to prove the decoupling between the position and attitude motion. The results confirm the effectiveness of the proposed multirotor design.
{"title":"The ODQuad: Design and Experimental Validation of a Novel Fully-actuated Quadrotor","authors":"M. Nigro, F. Pierri, F. Caccavale, Markus Ryll","doi":"10.1109/ICUAS57906.2023.10156620","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156620","url":null,"abstract":"This work experimentally validates a novel fully-actuated quadrotor-based unmanned aerial vehicle named ODQuad (OmniDirectional Quadrotor). The ODQuad is composed of three main parts arranged in a gimbal configuration. The internal mechanism is composed of two rotational joints with orthogonal and incident axes which allow to decouple the horizontal motions from the vehicle body rolling and pitching. Firstly, the physical prototype is presented and the motion controller, inherited by [1], has been tailored in such a way to integrate the servo actuators of the internal gimbal mechanism. Three trajectories have been commanded to prove the decoupling between the position and attitude motion. The results confirm the effectiveness of the proposed multirotor design.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115077149","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10156596
Mauro Sérgio Mafra Moreira, M. Sarcinelli-Filho
This paper complements a previous study on obstacle avoidance using the null space-based behavioral approach to autonomously guide a formation composed of a differential-drive wheeled platform and an unmanned aerial vehicle, to overtake obstacles modeled as potential fields. The highest priority, regarding the null space behavioral control, is assigned to the task of overcoming an obstacle, with the lowest priority assigned to moving the formation to a destination point. The controller is designed considering the paradigm of virtual structure, which is the three-dimensional straight line linking the robots. This approach allows controlling the robots to move in a coordinate way, leading the formation to reach the desired point while keeping the proposed rigid structure. The obstacle avoidance proposal is adopted for the ground and the aerial robots. When the ground robot maneuver to avoid an obstacle in the ground the position of the point of interest for control also varies, since it is in the ground vehicle, so that the aerial vehicle does not need to break the formation, continuing "attached" to the ground vehicle during the maneuver. However, when the aerial robot faces an obstacle, the formation behaves differently. The formation shape is not guaranteed to be preserved during the maneuver of the aerial robot to avoid the obstacle. This is the behavior this paper proposes to discuss: the effect of the null space-based behavioral control over the navigation of the formation. The scenario for this case study is an automated warehouse, inside which several ground platforms and aerial vehicles are moving to suitably store goods, possibly with boxes in the ground, also obstacles for the ground vehicle.
{"title":"Obstacle Avoidance Based on the Null Space Control Approach for a Formation of an Aerial and a Ground Robot","authors":"Mauro Sérgio Mafra Moreira, M. Sarcinelli-Filho","doi":"10.1109/ICUAS57906.2023.10156596","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156596","url":null,"abstract":"This paper complements a previous study on obstacle avoidance using the null space-based behavioral approach to autonomously guide a formation composed of a differential-drive wheeled platform and an unmanned aerial vehicle, to overtake obstacles modeled as potential fields. The highest priority, regarding the null space behavioral control, is assigned to the task of overcoming an obstacle, with the lowest priority assigned to moving the formation to a destination point. The controller is designed considering the paradigm of virtual structure, which is the three-dimensional straight line linking the robots. This approach allows controlling the robots to move in a coordinate way, leading the formation to reach the desired point while keeping the proposed rigid structure. The obstacle avoidance proposal is adopted for the ground and the aerial robots. When the ground robot maneuver to avoid an obstacle in the ground the position of the point of interest for control also varies, since it is in the ground vehicle, so that the aerial vehicle does not need to break the formation, continuing \"attached\" to the ground vehicle during the maneuver. However, when the aerial robot faces an obstacle, the formation behaves differently. The formation shape is not guaranteed to be preserved during the maneuver of the aerial robot to avoid the obstacle. This is the behavior this paper proposes to discuss: the effect of the null space-based behavioral control over the navigation of the formation. The scenario for this case study is an automated warehouse, inside which several ground platforms and aerial vehicles are moving to suitably store goods, possibly with boxes in the ground, also obstacles for the ground vehicle.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116757794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper presents an online planning algorithm for multiple Unmanned Aerial Vehicles (UAVs) cooperative search tracks based on Distributed Model Predictive Control (DMPC) for dynamic targets. To address the centralized multi-UAV collaboration problem, the proposed approach transforms it into a distributed subsystem MPC problem under the frame-work of DMPC. Firstly, a dynamic target Statistical Probability Map (SPM) update model is established. Next, the system opti-mal solution is obtained by combining Nash optimization and rolling optimization A-star algorithm through iterating the MPC problem of each subsystem. The simulation results demonstrate the efficacy of the proposed dynamic target SPM model in improving search efficiency. Furthermore, the scrolling-optimized A-star algorithm improves the accuracy and speed of subsystem single-step search. In conclusion, the DMPC method significantly reduces the solving scale of cooperative search problems while ensuring high solving accuracy.
{"title":"Multi-UAV Cooperative Search Planning Algorithm Based on Dynamic Target Probability Model","authors":"Zihang Ao, Yulong Zhang, Jing Huang, Yichen Lin, Xiaodeng Zhou, Youmin Zhang","doi":"10.1109/ICUAS57906.2023.10156269","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156269","url":null,"abstract":"This paper presents an online planning algorithm for multiple Unmanned Aerial Vehicles (UAVs) cooperative search tracks based on Distributed Model Predictive Control (DMPC) for dynamic targets. To address the centralized multi-UAV collaboration problem, the proposed approach transforms it into a distributed subsystem MPC problem under the frame-work of DMPC. Firstly, a dynamic target Statistical Probability Map (SPM) update model is established. Next, the system opti-mal solution is obtained by combining Nash optimization and rolling optimization A-star algorithm through iterating the MPC problem of each subsystem. The simulation results demonstrate the efficacy of the proposed dynamic target SPM model in improving search efficiency. Furthermore, the scrolling-optimized A-star algorithm improves the accuracy and speed of subsystem single-step search. In conclusion, the DMPC method significantly reduces the solving scale of cooperative search problems while ensuring high solving accuracy.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121948162","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10155855
M. Rinaldi, Stefano Primatesta, G. Guglieri, A. Rizzo
Market-based task allocation methods represent an effective strategy for scheduling heterogeneous tasks to a heterogeneous multi-agent system, e.g., a fleet of different Unmanned Aerial Vehicles (UAVs). This is mainly due to their computational efficiency, ease of hybridization with optimization techniques and adaptability to different communication architectures. In this paper, a novel hybrid auction-based task allocation architecture with multi-auctioneer agents’ behavior is proposed for an Urban Air Mobility application. The proposed method aims to solve the combined problem of: (i) scheduling parcel pick-up and delivery tasks with time deadlines while minimizing the drones’ energy consumption; (ii) scheduling battery re-charge tasks in order to ensure the service’s persistency; and (iii) evaluating safe aerial routes since the UAVs fly over populated areas. The validity of the approach is demonstrated through Monte Carlo simulations. Moreover, being the proposed architecture distributed among the UAVs, the impact of communication failures on well-defined solution quality parameters is also investigated.
{"title":"Multi-Auctioneer Market-based Task Scheduling for Persistent Drone Delivery","authors":"M. Rinaldi, Stefano Primatesta, G. Guglieri, A. Rizzo","doi":"10.1109/ICUAS57906.2023.10155855","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10155855","url":null,"abstract":"Market-based task allocation methods represent an effective strategy for scheduling heterogeneous tasks to a heterogeneous multi-agent system, e.g., a fleet of different Unmanned Aerial Vehicles (UAVs). This is mainly due to their computational efficiency, ease of hybridization with optimization techniques and adaptability to different communication architectures. In this paper, a novel hybrid auction-based task allocation architecture with multi-auctioneer agents’ behavior is proposed for an Urban Air Mobility application. The proposed method aims to solve the combined problem of: (i) scheduling parcel pick-up and delivery tasks with time deadlines while minimizing the drones’ energy consumption; (ii) scheduling battery re-charge tasks in order to ensure the service’s persistency; and (iii) evaluating safe aerial routes since the UAVs fly over populated areas. The validity of the approach is demonstrated through Monte Carlo simulations. Moreover, being the proposed architecture distributed among the UAVs, the impact of communication failures on well-defined solution quality parameters is also investigated.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130728358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10156272
Gervase H.L.H. Lovell-Prescod, Ziqing Ma, E. Smeur
Tailsitter Micro Air Vehicles with two rotors are promising due to their simplicity and efficient forward flight, but actuator saturation due to ineffective pitch control at a high angle of attack flight is a challenge limiting the flight envelope. This paper proposes a novel tilt-rotor tailsitter design which features two tilting rotors as the only means for control moment generation. Incremental Nonlinear Dynamic Inversion (INDI) is applied to the attitude control problem of the tiltrotor tailsitter, whose attitude angle tracking performance is validated by indoor and outdoor flight tests. It is found that actuator saturation is largely avoided by using thrust vectoring which provides sufficient capability of pitch moment generation. However, it is also found that the proposed design with only leading-edge tilting motors excluding any aerodynamic control surfaces has limited roll control effectiveness in forward flight.
{"title":"Attitude Control of a Tilt-rotor Tailsitter Micro Air Vehicle Using Incremental Control","authors":"Gervase H.L.H. Lovell-Prescod, Ziqing Ma, E. Smeur","doi":"10.1109/ICUAS57906.2023.10156272","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156272","url":null,"abstract":"Tailsitter Micro Air Vehicles with two rotors are promising due to their simplicity and efficient forward flight, but actuator saturation due to ineffective pitch control at a high angle of attack flight is a challenge limiting the flight envelope. This paper proposes a novel tilt-rotor tailsitter design which features two tilting rotors as the only means for control moment generation. Incremental Nonlinear Dynamic Inversion (INDI) is applied to the attitude control problem of the tiltrotor tailsitter, whose attitude angle tracking performance is validated by indoor and outdoor flight tests. It is found that actuator saturation is largely avoided by using thrust vectoring which provides sufficient capability of pitch moment generation. However, it is also found that the proposed design with only leading-edge tilting motors excluding any aerodynamic control surfaces has limited roll control effectiveness in forward flight.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131167490","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10156047
P. Valianti, Kleanthis Malialis, P. Kolios, G. Ellinas
Unmanned aerial vehicles (UAVs) are increasingly being utilized for a wide variety of applications. However, malicious or illegal UAV (drone) activity poses great challenges for public safety. To address such challenges, this work proposes a framework based on reinforcement learning (RL) in which multiple UAVs cooperatively jam multiple rogue drones in flight in order to safely disable their operation. The main objective is to select mobility and power level control actions for each UAV to best jam the rogue drones, while also accounting for the interference power received by surrounding communication systems. Simulation experiments are conducted to evaluate the performance of the proposed approach, demonstrating its effectiveness and advantages as compared to a centralized solution.
{"title":"Multi-Agent Reinforcement Learning for Multiple Rogue Drone Interception","authors":"P. Valianti, Kleanthis Malialis, P. Kolios, G. Ellinas","doi":"10.1109/ICUAS57906.2023.10156047","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156047","url":null,"abstract":"Unmanned aerial vehicles (UAVs) are increasingly being utilized for a wide variety of applications. However, malicious or illegal UAV (drone) activity poses great challenges for public safety. To address such challenges, this work proposes a framework based on reinforcement learning (RL) in which multiple UAVs cooperatively jam multiple rogue drones in flight in order to safely disable their operation. The main objective is to select mobility and power level control actions for each UAV to best jam the rogue drones, while also accounting for the interference power received by surrounding communication systems. Simulation experiments are conducted to evaluate the performance of the proposed approach, demonstrating its effectiveness and advantages as compared to a centralized solution.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129713268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-06DOI: 10.1109/ICUAS57906.2023.10156315
M. Hamanaka
We propose a method to improve the resolution of drone position and direction estimation on the basis of deep learning using three-dimensional (3D) topographic maps in nonglobal positioning system (GPS) environments. GPS is typically used to estimate the position of drones flying outdoors. However, it becomes difficult to estimate the position if the signal from GPS satellites is blocked by tall mountains or buildings, or if there are interference signals. To avoid this loss of GPS, we previously developed a learning-based flight area estimation method using 3D topographic maps. With this method, the flight area could be estimated with an accuracy of 98.4% in experiments conducted in 25 areas, each 40 meters square. However, a resolution of 40 meters square is difficult to use for drone control. Therefore, in this study, we will verify whether it is possible to improve the resolution by multiplexing the area division and the data acquisition direction. We also investigated whether the flight direction of the drone can be detected using a 3D map. Experimental results show that the position estimation was 96.8% accurate at a resolution of 25 meters square, and the direction estimation was 92.6% accurate for 12-direction estimation.
{"title":"Improving resolution in deep learning-based estimation of drone position and direction using 3D maps","authors":"M. Hamanaka","doi":"10.1109/ICUAS57906.2023.10156315","DOIUrl":"https://doi.org/10.1109/ICUAS57906.2023.10156315","url":null,"abstract":"We propose a method to improve the resolution of drone position and direction estimation on the basis of deep learning using three-dimensional (3D) topographic maps in nonglobal positioning system (GPS) environments. GPS is typically used to estimate the position of drones flying outdoors. However, it becomes difficult to estimate the position if the signal from GPS satellites is blocked by tall mountains or buildings, or if there are interference signals. To avoid this loss of GPS, we previously developed a learning-based flight area estimation method using 3D topographic maps. With this method, the flight area could be estimated with an accuracy of 98.4% in experiments conducted in 25 areas, each 40 meters square. However, a resolution of 40 meters square is difficult to use for drone control. Therefore, in this study, we will verify whether it is possible to improve the resolution by multiplexing the area division and the data acquisition direction. We also investigated whether the flight direction of the drone can be detected using a 3D map. Experimental results show that the position estimation was 96.8% accurate at a resolution of 25 meters square, and the direction estimation was 92.6% accurate for 12-direction estimation.","PeriodicalId":379073,"journal":{"name":"2023 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"2081 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129842661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：