Autonomous robotics plays a pivotal role to simplify humanmachine interaction while meeting the current industrial demands. In that process, machine intelligence plays a dominant role during the decision making in the operational state-space. Primarily, this decision making and control mechanism relies on sensing and actuation. Simultaneous localization and mapping (SLAM) is the most advanced technique that facilitates both sensing and actuation to achieve autonomy for robots. This work aims to collate multi-dimensional aspects of simultaneous localization and mapping techniques primarily in the purview of both deterministic and probabilistic frameworks. This investigation on SLAM classification is further elaborated into different categories such as Feature-based SLAM and Optimization based SLAM. In this work, the chronological evolution of SLAM technique develops a comprehensive understanding among the concerned research community.
{"title":"Evolution of Simultaneous localization and mapping framework for autonomous robotics - A comprehensive review","authors":"Sabita Pal, Smriti Gupta, Niva Das, Kuntal Ghosh","doi":"10.1115/1.4055161","DOIUrl":"https://doi.org/10.1115/1.4055161","url":null,"abstract":"\u0000 Autonomous robotics plays a pivotal role to simplify humanmachine interaction while meeting the current industrial demands. In that process, machine intelligence plays a dominant role during the decision making in the operational state-space. Primarily, this decision making and control mechanism relies on sensing and actuation. Simultaneous localization and mapping (SLAM) is the most advanced technique that facilitates both sensing and actuation to achieve autonomy for robots. This work aims to collate multi-dimensional aspects of simultaneous localization and mapping techniques primarily in the purview of both deterministic and probabilistic frameworks. This investigation on SLAM classification is further elaborated into different categories such as Feature-based SLAM and Optimization based SLAM. In this work, the chronological evolution of SLAM technique develops a comprehensive understanding among the concerned research community.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124435874","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}
Interactions with marine vegetation can disrupt unmanned underwater vehicle (UUV) mission success. Very little information is publicly available about the mechanisms causing these interactions or the consequences of them. This article compares the interactions between three different style UUVs and two different types of marine vegetation. Similar equipment and procedures were used to allow for the direct comparison between a GhostSwimmer, a REMUS-100, and a BlueROV2. Experimental test runs were conducted at different vegetation densities using either synthetic eelgrass or giant kelp. The resulting interactions depended on the vegetation type, vegetation density, propulsion mechanism of the vehicle, and vehicle geometry. Synthetic giant kelp caused a multitude of interactions including entanglement with the vehicle's body or propeller and blockage of the vehicle depending on the geometry and propulsion mechanism of the UUV. Eelgrass caused propeller entanglement, temporary speed reduction, and even blockage depending on the UUV. The use of an oscillating tail for propulsion coupled with a completely streamlined body appears to successfully mitigate adverse marine vegetation interactions.
{"title":"Effects of Body Geometry and Propulsion Type on Unmanned Underwater Vehicle Interactions with Marine Vegetation","authors":"G. V. Anuat, J. Klamo, A. Pollman","doi":"10.1115/1.4055083","DOIUrl":"https://doi.org/10.1115/1.4055083","url":null,"abstract":"\u0000 Interactions with marine vegetation can disrupt unmanned underwater vehicle (UUV) mission success. Very little information is publicly available about the mechanisms causing these interactions or the consequences of them. This article compares the interactions between three different style UUVs and two different types of marine vegetation. Similar equipment and procedures were used to allow for the direct comparison between a GhostSwimmer, a REMUS-100, and a BlueROV2. Experimental test runs were conducted at different vegetation densities using either synthetic eelgrass or giant kelp. The resulting interactions depended on the vegetation type, vegetation density, propulsion mechanism of the vehicle, and vehicle geometry. Synthetic giant kelp caused a multitude of interactions including entanglement with the vehicle's body or propeller and blockage of the vehicle depending on the geometry and propulsion mechanism of the UUV. Eelgrass caused propeller entanglement, temporary speed reduction, and even blockage depending on the UUV. The use of an oscillating tail for propulsion coupled with a completely streamlined body appears to successfully mitigate adverse marine vegetation interactions.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"93 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131733940","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 study aims to develop and integrate guidance and control functions for applications. The utility of integrated nonlinear optimal control and explicit guidance functions replaces linear PID control laws. This approach leverages UAV flight autonomy, thereby paving the way for creating an autonomous control technology with real-time target-relative guidance and re-targeting capabilities. A 360° roll maneuver combines extremal control and modified explicit guidance. The roll maneuver accurately reaches the desired position and velocity vectors through the proposed integration. Satisfying the first-order necessary optimality conditions demonstrates that the roll maneuver has extremal trajectories. To the best of the authors' knowledge, this is the first time analyzing and testing the Weierstrass condition and the first and second-order conditions of optimality for UAVs. Second-order conditions show that the 360° roll maneuver with explicit rotational attitude guidance does not have an optimal trajectory and yields an extremal trajectory.
{"title":"Extremal Control and Modified Explicit Guidance for Autonomous Unmanned Aerial Vehicles","authors":"E. Kawamura, D. Azimov","doi":"10.1115/1.4054951","DOIUrl":"https://doi.org/10.1115/1.4054951","url":null,"abstract":"\u0000 This study aims to develop and integrate guidance and control functions for applications. The utility of integrated nonlinear optimal control and explicit guidance functions replaces linear PID control laws. This approach leverages UAV flight autonomy, thereby paving the way for creating an autonomous control technology with real-time target-relative guidance and re-targeting capabilities. A 360° roll maneuver combines extremal control and modified explicit guidance. The roll maneuver accurately reaches the desired position and velocity vectors through the proposed integration. Satisfying the first-order necessary optimality conditions demonstrates that the roll maneuver has extremal trajectories. To the best of the authors' knowledge, this is the first time analyzing and testing the Weierstrass condition and the first and second-order conditions of optimality for UAVs. Second-order conditions show that the 360° roll maneuver with explicit rotational attitude guidance does not have an optimal trajectory and yields an extremal trajectory.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126347992","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}
Jing-Shan Zhao, Song-Tao Wei, Xiao-Cheng Sun, Junjie Ji
This paper proposes a mechanical system for a masonry robot comprising four parts: an omnidirectional mobile platform, a lifting platform, a six-degrees-of-freedom manipulator, and a grasping mechanism. The robot is specifically designed to carry out brick masonry on construction sites and is equipped to move bricks smoothly with slurry to their intended location. To prevent mortar from being shaken off the brick, the grasping mechanism is required to maintain optimal velocity and optimized acceleration. To implement online trajectory planning with velocity and acceleration constraints, the paper suggests an approach based on screw theory for resolving the inverse kinematics of the masonry robot. This method allows the inverse kinematic equations to be used to determine a unique solution for all joints of the redundant driver of the masonry robot. The approach and strategy are validated through numerical simulations of trajectory planning using a fifth-degree polynomial.
{"title":"Kinematics and Trajectory Planning of the Masonry Robot","authors":"Jing-Shan Zhao, Song-Tao Wei, Xiao-Cheng Sun, Junjie Ji","doi":"10.1115/1.4062734","DOIUrl":"https://doi.org/10.1115/1.4062734","url":null,"abstract":"\u0000 This paper proposes a mechanical system for a masonry robot comprising four parts: an omnidirectional mobile platform, a lifting platform, a six-degrees-of-freedom manipulator, and a grasping mechanism. The robot is specifically designed to carry out brick masonry on construction sites and is equipped to move bricks smoothly with slurry to their intended location. To prevent mortar from being shaken off the brick, the grasping mechanism is required to maintain optimal velocity and optimized acceleration. To implement online trajectory planning with velocity and acceleration constraints, the paper suggests an approach based on screw theory for resolving the inverse kinematics of the masonry robot. This method allows the inverse kinematic equations to be used to determine a unique solution for all joints of the redundant driver of the masonry robot. The approach and strategy are validated through numerical simulations of trajectory planning using a fifth-degree polynomial.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131846179","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}
Benjamin Abruzzo, D. Cappelleri, Philippos Mordohai
This paper presents an investigation of collaborative localization for heterogeneous robots, resulting in a scheme for relative localization of a heterogeneous team of lowcost mobile robots. A novel complementary Kalman Filter (CKF) approach is presented to address collaborative localization and mapping by optimally estimating the error states of the team. This indirect filter optimally combines the inertial/visual proprioceptive measurements of each vehicle with stereoscopic measurements made by the UGVs. An analysis is presented for both Complementary Kalman Filter (CKF) and Simultaneous Localization and Mapping (SLAM) approaches on maps containing randomly placed obstacles. In both simulation and experiments, we demonstrate the proposed methodology with a heterogeneous robot team. Six behavioral strategies, specifying the role and behavior of each robot, are simulated and evaluated for both CKF and SLAM approaches on maps containing randomly placed obstacles. Results show that the sources of error, image quantization, asynchronous sensors, and a non-stationary bias, were sufficiently modeled to estimate the pose of the aerial robot. The results demonstrate localization accuracies of 2 cm to 4 cm, on average, while the robots operate at a distance from each-other between 1 m and 4 m. The best performing behavior for the CKF approach maintained an average positional error of 2.2 cm and a relative error of 0.30% of the distance traveled for the entire team at the conclusion of maneuvers. For all multi-UGV strategies, the CKF approach outperformed the best SLAM results by 6.7 cm mean error (0.48% of distance traveled).
{"title":"Comparing Complementary Kalman Filters Against SLAM for Collaborative Localization of Heterogeneous Multi-Robot Teams","authors":"Benjamin Abruzzo, D. Cappelleri, Philippos Mordohai","doi":"10.1115/1.4054817","DOIUrl":"https://doi.org/10.1115/1.4054817","url":null,"abstract":"\u0000 This paper presents an investigation of collaborative localization for heterogeneous robots, resulting in a scheme for relative localization of a heterogeneous team of lowcost mobile robots. A novel complementary Kalman Filter (CKF) approach is presented to address collaborative localization and mapping by optimally estimating the error states of the team. This indirect filter optimally combines the inertial/visual proprioceptive measurements of each vehicle with stereoscopic measurements made by the UGVs. An analysis is presented for both Complementary Kalman Filter (CKF) and Simultaneous Localization and Mapping (SLAM) approaches on maps containing randomly placed obstacles. In both simulation and experiments, we demonstrate the proposed methodology with a heterogeneous robot team. Six behavioral strategies, specifying the role and behavior of each robot, are simulated and evaluated for both CKF and SLAM approaches on maps containing randomly placed obstacles. Results show that the sources of error, image quantization, asynchronous sensors, and a non-stationary bias, were sufficiently modeled to estimate the pose of the aerial robot. The results demonstrate localization accuracies of 2 cm to 4 cm, on average, while the robots operate at a distance from each-other between 1 m and 4 m. The best performing behavior for the CKF approach maintained an average positional error of 2.2 cm and a relative error of 0.30% of the distance traveled for the entire team at the conclusion of maneuvers. For all multi-UGV strategies, the CKF approach outperformed the best SLAM results by 6.7 cm mean error (0.48% of distance traveled).","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"76 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123238919","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 focuses on the development of distributed robust model predictive control (MPC) methods for multiple connected and automated vehicles (CAVs) to ensure their safe operation in the presence of uncertainty. The proposed layered control framework includes reference trajectory generation, distributionally robust obstacle occupancy set computation, distributed state constraint set evaluation, data-driven linear model representation, and robust tube-based MPC design. To enable distributed operation among the CAVs, we present a method, which exploits sampling-based reference trajectory generation and distributed constraint set evaluation methods, that decouples the coupled collision avoidance constraint among the CAVs. This is followed by data-driven linear model representation of the nonlinear system to evaluate the convex equivalent of the nonlinear control problem. Finally, to ensure safe operation in the presence of uncertainty, this paper employs a robust tube-based MPC method. For a multiple CAV lane change problem, simulation results show the efficacy of the proposed controller in terms of computational efficiency and the ability to generate safe and smooth CAV trajectories in a distributed fashion.
{"title":"Distributed Model Predictive Control for Connected and Automated Vehicles in the Presence of Uncertainty","authors":"B. Homchaudhuri, Viranjan Bhattacharyya","doi":"10.1115/1.4054696","DOIUrl":"https://doi.org/10.1115/1.4054696","url":null,"abstract":"\u0000 This paper focuses on the development of distributed robust model predictive control (MPC) methods for multiple connected and automated vehicles (CAVs) to ensure their safe operation in the presence of uncertainty. The proposed layered control framework includes reference trajectory generation, distributionally robust obstacle occupancy set computation, distributed state constraint set evaluation, data-driven linear model representation, and robust tube-based MPC design. To enable distributed operation among the CAVs, we present a method, which exploits sampling-based reference trajectory generation and distributed constraint set evaluation methods, that decouples the coupled collision avoidance constraint among the CAVs. This is followed by data-driven linear model representation of the nonlinear system to evaluate the convex equivalent of the nonlinear control problem. Finally, to ensure safe operation in the presence of uncertainty, this paper employs a robust tube-based MPC method. For a multiple CAV lane change problem, simulation results show the efficacy of the proposed controller in terms of computational efficiency and the ability to generate safe and smooth CAV trajectories in a distributed fashion.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116907170","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}
C. Goodin, Lucas Cagle, Greg Henley, Rhett Fereday, Justin Carrillo, Peilin Song, David P. McInnis
Autonomous ground vehicles (AGV) operating collaboratively have several advantages over vehicles operating alone. An AGV team may be more resilient and efficient than a single AGV. Other advantages of AGV teams include increased coverage and multiple viewing angles of terrain features as well as resistance to failure from any single AGV. Additionally, AGV teams can explore large terrains more quickly and thoroughly than a single system. In this work, the feasibility of using a team of high-mobility AGV to explore a navigation corridor, map the terrain, and autonomously flag obstacles for future navigation is evaluated. Focusing on negative obstacles, the value of using multiple vehicles to map a navigation corridor is quantified. This study is the first to evaluate large teams of AGV collaborating in realistic off-road, 3D environments. The feasibility of the large-scale AGV team is demonstrated while avoiding the high cost of purchasing and testing large numbers of vehicles by using the MSU Autonomous Vehicle Simulator (MAVS), a high-fidelity, physics-based simulation tool. The cost and benefits of increasing the AGV team size are evaluated. The simulation results show how factors like fuel use, map coverage, and obstacle detection are influenced by increasing numbers of AGV in the team. The simulation architecture is presented and experiments quantifying the performance of the simulator are shown. Finally, a model for evaluating the tradeoff between mission effectiveness and fuel use is developed and presented to demonstrate the utility of this approach.
{"title":"Evaluating Tradeoffs for Swarm Reconnaissance with Autonomous Ground Vehicles","authors":"C. Goodin, Lucas Cagle, Greg Henley, Rhett Fereday, Justin Carrillo, Peilin Song, David P. McInnis","doi":"10.1115/1.4054581","DOIUrl":"https://doi.org/10.1115/1.4054581","url":null,"abstract":"\u0000 Autonomous ground vehicles (AGV) operating collaboratively have several advantages over vehicles operating alone. An AGV team may be more resilient and efficient than a single AGV. Other advantages of AGV teams include increased coverage and multiple viewing angles of terrain features as well as resistance to failure from any single AGV. Additionally, AGV teams can explore large terrains more quickly and thoroughly than a single system. In this work, the feasibility of using a team of high-mobility AGV to explore a navigation corridor, map the terrain, and autonomously flag obstacles for future navigation is evaluated. Focusing on negative obstacles, the value of using multiple vehicles to map a navigation corridor is quantified. This study is the first to evaluate large teams of AGV collaborating in realistic off-road, 3D environments. The feasibility of the large-scale AGV team is demonstrated while avoiding the high cost of purchasing and testing large numbers of vehicles by using the MSU Autonomous Vehicle Simulator (MAVS), a high-fidelity, physics-based simulation tool. The cost and benefits of increasing the AGV team size are evaluated. The simulation results show how factors like fuel use, map coverage, and obstacle detection are influenced by increasing numbers of AGV in the team. The simulation architecture is presented and experiments quantifying the performance of the simulator are shown. Finally, a model for evaluating the tradeoff between mission effectiveness and fuel use is developed and presented to demonstrate the utility of this approach.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122177927","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}
Velocity obstacle is one of popular reactive navigation algorithms for path planning of autonomous agents. The collision-free property can be guaranteed if the agent is able to choose a velocity outside the velocity obstacle region under the assumption that obstacles maintain a constant velocity within the control cycle time of the agent. To date, selection of the optimal velocity relies on either sampling or optimization approaches. The sampling approach can maintain the same amount of computation cost but may miss feasible solutions under collision risks with insufficient number of samples. The optimization approach such as the linear programming demands for convexity of the constraints in the velocity space which may not be satisfied considering non-holonomic agents. In addition, the algorithm has varying computation demand depending on the navigation situation. This paper proposes an analytic approach for choosing a candidate velocity rather than relying on the sampling or optimization approaches. The analytic approach can significantly reduce computation cost without sacrificing the performance. Agents with both holonomic and non-holonomic constraints are considered to demonstrate the performance and efficiency of the proposed approach. Extensive comparison studies with static, non-reactive, and reactive moving obstacles demonstrate that the analytical velocity obstacle is computationally much more efficient than the optimization based approach and performs better than the sampling based approach.
{"title":"Analytic velocity obstacle for efficient collision avoidance computation and a comparison study with sampling and optimization based approaches","authors":"Zhimin Xi, E. Torkamani","doi":"10.1115/1.4054527","DOIUrl":"https://doi.org/10.1115/1.4054527","url":null,"abstract":"\u0000 Velocity obstacle is one of popular reactive navigation algorithms for path planning of autonomous agents. The collision-free property can be guaranteed if the agent is able to choose a velocity outside the velocity obstacle region under the assumption that obstacles maintain a constant velocity within the control cycle time of the agent. To date, selection of the optimal velocity relies on either sampling or optimization approaches. The sampling approach can maintain the same amount of computation cost but may miss feasible solutions under collision risks with insufficient number of samples. The optimization approach such as the linear programming demands for convexity of the constraints in the velocity space which may not be satisfied considering non-holonomic agents. In addition, the algorithm has varying computation demand depending on the navigation situation. This paper proposes an analytic approach for choosing a candidate velocity rather than relying on the sampling or optimization approaches. The analytic approach can significantly reduce computation cost without sacrificing the performance. Agents with both holonomic and non-holonomic constraints are considered to demonstrate the performance and efficiency of the proposed approach. Extensive comparison studies with static, non-reactive, and reactive moving obstacles demonstrate that the analytical velocity obstacle is computationally much more efficient than the optimization based approach and performs better than the sampling based approach.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"55 14","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133203955","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}
Autonomous vehicle control approaches are rapidly being developed for everyday street-driving scenarios. This paper considers autonomous vehicle control in a less common, albeit important, situation – a vehicle stuck in a ditch. In this scenario, a solution is typically obtained by either using a tow- truck or by humans rocking the vehicle to build momentum and push the vehicle out. However, it would be much more safe and convenient if a vehicle was able to exit the ditch autonomously – without human intervention. In exploration of this idea, this paper derives the governing equations for a vehicle moving along an arbitrary ditch profile with torques applied to front and rear wheels and the consideration of four regions of wheel-slip. A reward function was designed to minimize wheel-slip and the model was used to train control agents using Probabilistic Inference for Learning COntrol (PILCO) and Deep Deterministic Policy Gradient (DDPG) Reinforcement Learning (RL) algorithms. Both Rear-Wheel-Drive (RWD) and All-Wheel-Drive (AWD) results were compared, showing the capability of the agents to achieve escape from a ditch while minimizing wheel-slip for several ditch profiles. The policy results from applying RL to this problem intuitively increased the momentum of the vehicle and applied “braking” to the wheels when slip was detected so as to achieve a safe exit from the ditch. The conclusions show a pathway to apply aspects of this paper to specific vehicles.
{"title":"Modeling and Reinforcement Learning Control of an Autonomous Vehicle to Get Unstuck From a Ditch","authors":"Levi H. Manring, B. Mann","doi":"10.1115/1.4054499","DOIUrl":"https://doi.org/10.1115/1.4054499","url":null,"abstract":"\u0000 Autonomous vehicle control approaches are rapidly being developed for everyday street-driving scenarios. This paper considers autonomous vehicle control in a less common, albeit important, situation – a vehicle stuck in a ditch. In this scenario, a solution is typically obtained by either using a tow- truck or by humans rocking the vehicle to build momentum and push the vehicle out. However, it would be much more safe and convenient if a vehicle was able to exit the ditch autonomously – without human intervention. In exploration of this idea, this paper derives the governing equations for a vehicle moving along an arbitrary ditch profile with torques applied to front and rear wheels and the consideration of four regions of wheel-slip. A reward function was designed to minimize wheel-slip and the model was used to train control agents using Probabilistic Inference for Learning COntrol (PILCO) and Deep Deterministic Policy Gradient (DDPG) Reinforcement Learning (RL) algorithms. Both Rear-Wheel-Drive (RWD) and All-Wheel-Drive (AWD) results were compared, showing the capability of the agents to achieve escape from a ditch while minimizing wheel-slip for several ditch profiles. The policy results from applying RL to this problem intuitively increased the momentum of the vehicle and applied “braking” to the wheels when slip was detected so as to achieve a safe exit from the ditch. The conclusions show a pathway to apply aspects of this paper to specific vehicles.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126399785","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}
Sam Kassoumeh, Vijitashwa Pandey, D. Gorsich, P. Jayakumar
This work presents some results in the value of information calculations for multi-attribute decision making under uncertainty. Almost all engineering activities are undertaken in the face of uncertainty and a decision that maximizes a suitably chosen metric is generally selected. It becomes essential sometimes to collect information regarding these uncertainties so that better informed decisions can be made. Calculation of the worth of this information (VoI) is a difficult task, particularly when multiple attributes are present and there exists dependence between the random attributes in the same alternative or across different alternatives. In this paper, closed-form expressions and numerical models for the calculation of VoI are presented. Particularly, we derive methods for the general scenario where we have to decide over two or more alternatives, each involving two or more continuous random attributes exhibiting some level of dependence with the others. These reduce or completely eliminate the need for conducting simulations or approximations, both of which tend to be either computationally expensive (such as Monte Carlo), limited in accuracy, or both. We also introduce “attribute-wise VoI”, which shows that collecting information on one or more of the attribute(s) makes sense only in specific dependence scenarios and tradeoff relationships between attributes. Calculation methods for value of such information are also provided. We illustrate our models on mobile autonomous system selection decisions. We conclude with a discussion on the avenues for future research into the optimal mix of a system's intelligence (autonomy), communication and information gathering.
{"title":"VALUE OF INFORMATION IN MULTIATTRIBUTE DECISIONS WITH APPLICATIONS IN GROUND VEHICLE AUTONOMY","authors":"Sam Kassoumeh, Vijitashwa Pandey, D. Gorsich, P. Jayakumar","doi":"10.1115/1.4054125","DOIUrl":"https://doi.org/10.1115/1.4054125","url":null,"abstract":"\u0000 This work presents some results in the value of information calculations for multi-attribute decision making under uncertainty. Almost all engineering activities are undertaken in the face of uncertainty and a decision that maximizes a suitably chosen metric is generally selected. It becomes essential sometimes to collect information regarding these uncertainties so that better informed decisions can be made. Calculation of the worth of this information (VoI) is a difficult task, particularly when multiple attributes are present and there exists dependence between the random attributes in the same alternative or across different alternatives. In this paper, closed-form expressions and numerical models for the calculation of VoI are presented. Particularly, we derive methods for the general scenario where we have to decide over two or more alternatives, each involving two or more continuous random attributes exhibiting some level of dependence with the others. These reduce or completely eliminate the need for conducting simulations or approximations, both of which tend to be either computationally expensive (such as Monte Carlo), limited in accuracy, or both. We also introduce “attribute-wise VoI”, which shows that collecting information on one or more of the attribute(s) makes sense only in specific dependence scenarios and tradeoff relationships between attributes. Calculation methods for value of such information are also provided. We illustrate our models on mobile autonomous system selection decisions. We conclude with a discussion on the avenues for future research into the optimal mix of a system's intelligence (autonomy), communication and information gathering.","PeriodicalId":164923,"journal":{"name":"Journal of Autonomous Vehicles and Systems","volume":"05 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116692052","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：