Pub Date : 2019-06-01DOI: 10.1109/ICUAS.2019.8798031
Valentim Ernandes Neto, M. Sarcinelli-Filho, A. Brandão
This work discusses a trajectory tracking task accomplished by a formation of two heterogeneous robots, namely a Pioneer 3-DX (UGV) and an AR.Drone 2.0 quadcopter (UAV). The control paradigm considers the formation as a virtual structure, here represented by a 3D straight line linking the two agents. The formation is characterized by the distance between the two robots and two angles, measured between the linking-line and the YZ and XZ planes. In this approach, the formation position coincides with the UGV position. The transformation between the robot to formation variables allows characterizing the motion and shape of the formation starting from the movement of the robots, or vice-versa. Then, a controller guides the formation and manages two conflicting subtasks (reach the desired robot position or keep the formation shape), whose priority is defined by the null space-based approach. Finally, real experiments validate the proposed approach during trajectory-tracking tasks for three scenarios, one assigning the highest priority to the formation position, another to the formation shape, and a last one, without any priority. In addition, we evaluated the formation behavior in failure situations in one of the robots.
{"title":"Trajectory-tracking of a Heterogeneous Formation Using Null Space-Based Control","authors":"Valentim Ernandes Neto, M. Sarcinelli-Filho, A. Brandão","doi":"10.1109/ICUAS.2019.8798031","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8798031","url":null,"abstract":"This work discusses a trajectory tracking task accomplished by a formation of two heterogeneous robots, namely a Pioneer 3-DX (UGV) and an AR.Drone 2.0 quadcopter (UAV). The control paradigm considers the formation as a virtual structure, here represented by a 3D straight line linking the two agents. The formation is characterized by the distance between the two robots and two angles, measured between the linking-line and the YZ and XZ planes. In this approach, the formation position coincides with the UGV position. The transformation between the robot to formation variables allows characterizing the motion and shape of the formation starting from the movement of the robots, or vice-versa. Then, a controller guides the formation and manages two conflicting subtasks (reach the desired robot position or keep the formation shape), whose priority is defined by the null space-based approach. Finally, real experiments validate the proposed approach during trajectory-tracking tasks for three scenarios, one assigning the highest priority to the formation position, another to the formation shape, and a last one, without any priority. In addition, we evaluated the formation behavior in failure situations in one of the robots.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132603808","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8797953
Titouan Verdu, G. Hattenberger, S. Lacroix
Modeling the cloud microphysics processes is essential to improve our understanding in climate changes and reduce the uncertainties in weather predictions. Aircrafts, remote sensing and ground-based infrastructures provide either sparse or coarse spatial measurements that are not sufficient to develop fine cloud models. UAVs have shown their ability to collect relevant cloud in-situ measures, and can be even more efficient when deployed in fleets. However, collecting relevant cloud data call for specific trajectories: this paper introduces a series of flight patterns dedicated to cloud exploration by a fleet of UAVs. The patterns definition comprise both a priori geometric information and real-time reactions to collected data. Results in simulated clouds assess their relevance for cloud in situ data collection.
{"title":"Flight patterns for clouds exploration with a fleet of UAVs","authors":"Titouan Verdu, G. Hattenberger, S. Lacroix","doi":"10.1109/ICUAS.2019.8797953","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8797953","url":null,"abstract":"Modeling the cloud microphysics processes is essential to improve our understanding in climate changes and reduce the uncertainties in weather predictions. Aircrafts, remote sensing and ground-based infrastructures provide either sparse or coarse spatial measurements that are not sufficient to develop fine cloud models. UAVs have shown their ability to collect relevant cloud in-situ measures, and can be even more efficient when deployed in fleets. However, collecting relevant cloud data call for specific trajectories: this paper introduces a series of flight patterns dedicated to cloud exploration by a fleet of UAVs. The patterns definition comprise both a priori geometric information and real-time reactions to collected data. Results in simulated clouds assess their relevance for cloud in situ data collection.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121668646","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 : 2019-06-01DOI: 10.1109/icuas.2019.8797932
{"title":"ICUAS 2019 Final Program","authors":"","doi":"10.1109/icuas.2019.8797932","DOIUrl":"https://doi.org/10.1109/icuas.2019.8797932","url":null,"abstract":"","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"117 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115087816","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8798322
Sunan Huang, R. Teo
Autonomous unmanned aerial vehicles (UAVs) need to dynamically re-plan paths online to avoid newly detected obstacles and no-fly zones. Existing 3D path planning methods are either too computationally intensive for online use or have practical limitations for actual applications. We propose a new method based on visibility graphs that is both computationally efficient for online use and is suitable for actual applications. We consider the 3D space to be composed of many 2D planes that all pass through the current position of the UAV and the destination point. Finding the shortest collision-free path in each plane is a 2D path planning problem which can be solved by using existing visibility graph algorithms. We then collect all the shortest paths generated from each 2D plane and find the shortest path in the whole 3D space. We present the results of the proposed 3D path planning algorithm for two cases to demonstrate that the proposed method is effective.
{"title":"Computationally Efficient Visibility Graph-Based Generation Of 3D Shortest Collision-Free Path Among Polyhedral Obstacles For Unmanned Aerial Vehicles","authors":"Sunan Huang, R. Teo","doi":"10.1109/ICUAS.2019.8798322","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8798322","url":null,"abstract":"Autonomous unmanned aerial vehicles (UAVs) need to dynamically re-plan paths online to avoid newly detected obstacles and no-fly zones. Existing 3D path planning methods are either too computationally intensive for online use or have practical limitations for actual applications. We propose a new method based on visibility graphs that is both computationally efficient for online use and is suitable for actual applications. We consider the 3D space to be composed of many 2D planes that all pass through the current position of the UAV and the destination point. Finding the shortest collision-free path in each plane is a 2D path planning problem which can be solved by using existing visibility graph algorithms. We then collect all the shortest paths generated from each 2D plane and find the shortest path in the whole 3D space. We present the results of the proposed 3D path planning algorithm for two cases to demonstrate that the proposed method is effective.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124067004","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8798236
Kristian Husum Terkildsen, Kjeld Jensen
The Unmanned Aerial Systems (UAS) legislation in the European Union will expectedly become unified by 2020. This legislation divides UAS operations into three categories, one of which is the Specific category, in which operations of higher risk will be placed. Permission for UAS operations in the Specific category is granted based on a Specific Operations Risk Assessment (SORA). The risk assessment categorizes operations into Specific Assurance and Integrity Levels (SAILs), which determines the level of the requirements for the operation such as operational procedures, training of the remote crew, UAS specification and maintenance etc. However, a conceptual gap exists between the defined technology requirements for UAS and the available UAS platforms. In this work, a prototype assessment tool based on the SORA defined technology requirements has been developed. The tool is a web-based questionnaire, which pulls the questions from a data structure and stores the answers in a database for analysis. Upon completion of the questionnaire, the user receives an automatically generated evaluation report. The evaluation report points out the areas of inadequacy and provides suggestions for mitigating these. The tool in its current state has been used to evaluate four commonly used UAS. The four UAS have been tested for SAIL II through IV, and the results show that only one of the selected UAS complies with SAIL II. Future work is to refine the tool including the list of questions and the UAS evaluation report generator, based on feedback from UAS domain experts and to expand it to cover fixed-wing UAS.
{"title":"Towards a Tool for Assessing UAS Compliance with the JARUS SORA Guidelines","authors":"Kristian Husum Terkildsen, Kjeld Jensen","doi":"10.1109/ICUAS.2019.8798236","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8798236","url":null,"abstract":"The Unmanned Aerial Systems (UAS) legislation in the European Union will expectedly become unified by 2020. This legislation divides UAS operations into three categories, one of which is the Specific category, in which operations of higher risk will be placed. Permission for UAS operations in the Specific category is granted based on a Specific Operations Risk Assessment (SORA). The risk assessment categorizes operations into Specific Assurance and Integrity Levels (SAILs), which determines the level of the requirements for the operation such as operational procedures, training of the remote crew, UAS specification and maintenance etc. However, a conceptual gap exists between the defined technology requirements for UAS and the available UAS platforms. In this work, a prototype assessment tool based on the SORA defined technology requirements has been developed. The tool is a web-based questionnaire, which pulls the questions from a data structure and stores the answers in a database for analysis. Upon completion of the questionnaire, the user receives an automatically generated evaluation report. The evaluation report points out the areas of inadequacy and provides suggestions for mitigating these. The tool in its current state has been used to evaluate four commonly used UAS. The four UAS have been tested for SAIL II through IV, and the results show that only one of the selected UAS complies with SAIL II. Future work is to refine the tool including the list of questions and the UAS evaluation report generator, based on feedback from UAS domain experts and to expand it to cover fixed-wing UAS.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124709467","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8797934
Akinori Sakaguchi, T. Takimoto, T. Ushio
This paper is concerned with a novel quadcopter with a tilting capability using a parallel link mechanism. The proposed quadcopter consists of the tilting frame and a main body, and an angle between them is called a tilt angle. A ratio of the current width to the maximum one of the quadcopter is called a folding ratio. Using only one servo motor, it can tilt in the pitch direction and be folded vertically within a user-specified ranges of the tilt angle and the folding ratio, respectively. We show a procedure of determining design parameters of the tilting frame to satisfy specifications given by the user. Due to the tilting mechanism of the frame, the proposed quadcopter is not well-controlled as the tilt angle approaches $pmpi/2$. Therefore, we focus on a stabilization problem of the quadcopter at a desired tilt angle specified by a user around $pmpi/2$ and propose a novel reference tilt angle generator based on the pitch angle. In simulations, we show an effectiveness of the proposed strategy for the reference tilt angle by comparing it with a conventional PID controller under the existence of a disturbance.
{"title":"A Novel Quadcopter with A Tilting Frame using Parallel Link Mechanism","authors":"Akinori Sakaguchi, T. Takimoto, T. Ushio","doi":"10.1109/ICUAS.2019.8797934","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8797934","url":null,"abstract":"This paper is concerned with a novel quadcopter with a tilting capability using a parallel link mechanism. The proposed quadcopter consists of the tilting frame and a main body, and an angle between them is called a tilt angle. A ratio of the current width to the maximum one of the quadcopter is called a folding ratio. Using only one servo motor, it can tilt in the pitch direction and be folded vertically within a user-specified ranges of the tilt angle and the folding ratio, respectively. We show a procedure of determining design parameters of the tilting frame to satisfy specifications given by the user. Due to the tilting mechanism of the frame, the proposed quadcopter is not well-controlled as the tilt angle approaches $pmpi/2$. Therefore, we focus on a stabilization problem of the quadcopter at a desired tilt angle specified by a user around $pmpi/2$ and propose a novel reference tilt angle generator based on the pitch angle. In simulations, we show an effectiveness of the proposed strategy for the reference tilt angle by comparing it with a conventional PID controller under the existence of a disturbance.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129420824","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8798048
Jorge M. Arizaga, H. Castañeda, P. Castillo
This manuscript presents the modeling and control for a tilted-motors hexacopter unmanned aircraft system, which is subject to external perturbations. A tilt angle for each rotor on the rotorcraft is considered allowing a fully-actuated system, capable to exert horizontal forces. Given the robustness properties and the management of control effort, the controller design is based on adaptive sliding mode technique providing stabilization and tracking of desired trajectories under a perturbed environment. Finally, numerical simulations show the feasibility and advantages of the proposed scheme.
{"title":"Adaptive Control for a Tilted-Motors Hexacopter UAS Flying on a Perturbed Environment","authors":"Jorge M. Arizaga, H. Castañeda, P. Castillo","doi":"10.1109/ICUAS.2019.8798048","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8798048","url":null,"abstract":"This manuscript presents the modeling and control for a tilted-motors hexacopter unmanned aircraft system, which is subject to external perturbations. A tilt angle for each rotor on the rotorcraft is considered allowing a fully-actuated system, capable to exert horizontal forces. Given the robustness properties and the management of control effort, the controller design is based on adaptive sliding mode technique providing stabilization and tracking of desired trajectories under a perturbed environment. Finally, numerical simulations show the feasibility and advantages of the proposed scheme.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130815789","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8797863
Minjun Kim, J. R. Morrison
We develop a task allocation method for persistent UAV security presence (PUSP). UAVs accompany customers and thereby provide security services to them. Key features incorporated are randomness in the arrival of customers and travel durations. We formalize our system as a general network consisting of nodes, arcs, UAVs and routes. From the network, we automatically generate a Markov decision process (MDP) model and simulator. The MDP formulation can be solved exactly only for small problems. In such cases, we employ classic value iteration to obtain optimal polices. To address larger systems consisting of more resources, we develop a greedy task assignment heuristic (GTAH) and simplified MDP heuristics (SMH). Numerical studies demonstrate that the GTAH is approximately 10% suboptimal and that the SMH is about 4% suboptimal with regard to small-scale problems. For larger problems $(sim 10^{90}$ states), the performance of the SMH is approximately 3% better than that of the GTAH
{"title":"On systems of UAVs for persistent security presence: A generic network representation, MDP formulation and heuristics for task allocation","authors":"Minjun Kim, J. R. Morrison","doi":"10.1109/ICUAS.2019.8797863","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8797863","url":null,"abstract":"We develop a task allocation method for persistent UAV security presence (PUSP). UAVs accompany customers and thereby provide security services to them. Key features incorporated are randomness in the arrival of customers and travel durations. We formalize our system as a general network consisting of nodes, arcs, UAVs and routes. From the network, we automatically generate a Markov decision process (MDP) model and simulator. The MDP formulation can be solved exactly only for small problems. In such cases, we employ classic value iteration to obtain optimal polices. To address larger systems consisting of more resources, we develop a greedy task assignment heuristic (GTAH) and simplified MDP heuristics (SMH). Numerical studies demonstrate that the GTAH is approximately 10% suboptimal and that the SMH is about 4% suboptimal with regard to small-scale problems. For larger problems $(sim 10^{90}$ states), the performance of the SMH is approximately 3% better than that of the GTAH","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130411800","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8797940
Dylan Cawthorne, Alessandra Cenci
Value Sensitive Design (VSD) is an interdisciplinary approach to technological development that systematically incorporates ethical considerations and social impacts as design inputs. Here, the VSD methodology is described, and elements of VSD are applied with a technological focus to analyze an existing prototype humanitarian cargo drone. Then, a new proposed drone design that better supports the values of human welfare (physical, psychological, and material welfare), and environmental sustainability is developed. The new drone is a high-speed fixed-wing drone which uses internal combustion engines and drops its payload via parachute to minimize transportation time and maximize patient physical welfare. It uses lower levels of automation such as manual flight monitoring to increase reliability and safety (physical welfare), and support the local workforce (material welfare). The drone uses much less energy than the technology it replaces, and is therefore much more environmentally friendly, supporting environmental sustainability. This work contributes by being the first to apply VSD methods to the technological development of a specific drone platform, and by demonstrating how drone engineers can use VSD to develop ”ethical” drones.
{"title":"Value Sensitive Design of a Humanitarian Cargo Drone","authors":"Dylan Cawthorne, Alessandra Cenci","doi":"10.1109/ICUAS.2019.8797940","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8797940","url":null,"abstract":"Value Sensitive Design (VSD) is an interdisciplinary approach to technological development that systematically incorporates ethical considerations and social impacts as design inputs. Here, the VSD methodology is described, and elements of VSD are applied with a technological focus to analyze an existing prototype humanitarian cargo drone. Then, a new proposed drone design that better supports the values of human welfare (physical, psychological, and material welfare), and environmental sustainability is developed. The new drone is a high-speed fixed-wing drone which uses internal combustion engines and drops its payload via parachute to minimize transportation time and maximize patient physical welfare. It uses lower levels of automation such as manual flight monitoring to increase reliability and safety (physical welfare), and support the local workforce (material welfare). The drone uses much less energy than the technology it replaces, and is therefore much more environmentally friendly, supporting environmental sustainability. This work contributes by being the first to apply VSD methods to the technological development of a specific drone platform, and by demonstrating how drone engineers can use VSD to develop ”ethical” drones.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131018216","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 : 2019-06-01DOI: 10.1109/ICUAS.2019.8798034
Rahul Peddi, N. Bezzo
Autonomous flight in unmanned aerial vehicles (UAVs) generally requires platform-specific knowledge of the dynamical parameters and control architecture. Recently, UAVs have become more accessible with off-the-shelf options that are well-tuned and stable for user teleoperation but due to unknown model parameters, they are typically not ready for autonomous operations. In this paper, we develop a method to enable autonomous flight on vehicles that are designed for teleoperation with minimal knowledge of the dynamical and controller parameters. The proposed method uses a basic knowledge of the control and dynamic architecture along with human teleoperated trajectories as demonstrations to train a thin-plate spline (TPS) regression model, which is then used to manipulate the pre-trained commands to generate new autonomous input commands for autonomous navigation over new trajectories. A statistical approach is also presented together with a satisfiability modulo theories (SMT) solver to assess the learned prediction error and correct to minimize errors in the input generation. A robust control-based strategy is also proposed to adjust autonomous input commands during run-time for closed loop trajectory tracking. Finally, we validate the proposed approach with trajectory-following experiments on a quadrotor UAV.
{"title":"Parameter-free Regression-based Autonomous Control of Off-the-shelf Quadrotor UAVs","authors":"Rahul Peddi, N. Bezzo","doi":"10.1109/ICUAS.2019.8798034","DOIUrl":"https://doi.org/10.1109/ICUAS.2019.8798034","url":null,"abstract":"Autonomous flight in unmanned aerial vehicles (UAVs) generally requires platform-specific knowledge of the dynamical parameters and control architecture. Recently, UAVs have become more accessible with off-the-shelf options that are well-tuned and stable for user teleoperation but due to unknown model parameters, they are typically not ready for autonomous operations. In this paper, we develop a method to enable autonomous flight on vehicles that are designed for teleoperation with minimal knowledge of the dynamical and controller parameters. The proposed method uses a basic knowledge of the control and dynamic architecture along with human teleoperated trajectories as demonstrations to train a thin-plate spline (TPS) regression model, which is then used to manipulate the pre-trained commands to generate new autonomous input commands for autonomous navigation over new trajectories. A statistical approach is also presented together with a satisfiability modulo theories (SMT) solver to assess the learned prediction error and correct to minimize errors in the input generation. A robust control-based strategy is also proposed to adjust autonomous input commands during run-time for closed loop trajectory tracking. Finally, we validate the proposed approach with trajectory-following experiments on a quadrotor UAV.","PeriodicalId":426616,"journal":{"name":"2019 International Conference on Unmanned Aircraft Systems (ICUAS)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127898982","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：