Pub Date : 2015-08-27DOI: 10.1109/IVS.2015.7225834
Peter Ritzer, C. Winter, J. Brembeck
This work describes an advanced path following control strategy enabling overactuated robotic vehicles like the ROboMObil (ROMO) [1] to automatically follow predefined paths while all states of the vehicle's planar motion are controlled. This strategy is useful for autonomous vehicles which are guided along online generated paths including severe driving maneuvers caused by e.g. obstacle avoidance. The proposed approach combines path following, i.e. tracking a plane curve without a priori time parameterization of a trajectory, with feedback based vehicle dynamics stabilization. A path interpolation method is introduced which allows to perform the path following task employing a trajectory tracking controller. Furthermore a tracking controller based on I/O linearization and quadratic programming based control allocation is proposed which allows employing the vehicle's overactuation in an optimal manner. The work concludes by a simulative evaluation of the controller performance.
{"title":"Advanced path following control of an overactuated robotic vehicle","authors":"Peter Ritzer, C. Winter, J. Brembeck","doi":"10.1109/IVS.2015.7225834","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225834","url":null,"abstract":"This work describes an advanced path following control strategy enabling overactuated robotic vehicles like the ROboMObil (ROMO) [1] to automatically follow predefined paths while all states of the vehicle's planar motion are controlled. This strategy is useful for autonomous vehicles which are guided along online generated paths including severe driving maneuvers caused by e.g. obstacle avoidance. The proposed approach combines path following, i.e. tracking a plane curve without a priori time parameterization of a trajectory, with feedback based vehicle dynamics stabilization. A path interpolation method is introduced which allows to perform the path following task employing a trajectory tracking controller. Furthermore a tracking controller based on I/O linearization and quadratic programming based control allocation is proposed which allows employing the vehicle's overactuation in an optimal manner. The work concludes by a simulative evaluation of the controller performance.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126557935","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225802
S. Lefèvre, Ashwin Carvalho, F. Borrelli
We propose a learning-based method for the longitudinal control of an autonomous vehicle on the highway. We use a driver model to generate acceleration inputs which are used as a reference by a model predictive controller. The driver model is trained using real driving data, so that it can reproduce the driver's behavior. We show the system's ability to reproduce different driving styles from different drivers. By solving a constrained optimization problem, the model predictive controller ensures that the control inputs applied to the vehicle satisfy some safety criteria. This is demonstrated on a vehicle by artificially creating potentially dangerous situations with virtual obstacles.
{"title":"Autonomous car following: A learning-based approach","authors":"S. Lefèvre, Ashwin Carvalho, F. Borrelli","doi":"10.1109/IVS.2015.7225802","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225802","url":null,"abstract":"We propose a learning-based method for the longitudinal control of an autonomous vehicle on the highway. We use a driver model to generate acceleration inputs which are used as a reference by a model predictive controller. The driver model is trained using real driving data, so that it can reproduce the driver's behavior. We show the system's ability to reproduce different driving styles from different drivers. By solving a constrained optimization problem, the model predictive controller ensures that the control inputs applied to the vehicle satisfy some safety criteria. This is demonstrated on a vehicle by artificially creating potentially dangerous situations with virtual obstacles.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"202 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131513915","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225664
Graeme Morrison, D. Cebon
Active safety systems for Heavy Goods Vehicles (HGVs), like passenger cars, often require an accurate estimate of sideslip angle. However, very little research has been published on HGV sideslip estimation in low friction conditions. This paper proposes three nonlinear Kalman Filters to estimate the tractor sideslip angle of a tractor-semitrailer combination. Performance is compared in simulation to a linear Kalman Filter in both high and low friction conditions. An Unscented Kalman Filter using a yaw-roll vehicle model and nonlinear tire model is found to accurately estimate sideslip in all maneuvers simulated, significantly outperforming the linear Kalman Filter.
{"title":"Sideslip estimation for articulated heavy vehicles in low friction conditions","authors":"Graeme Morrison, D. Cebon","doi":"10.1109/IVS.2015.7225664","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225664","url":null,"abstract":"Active safety systems for Heavy Goods Vehicles (HGVs), like passenger cars, often require an accurate estimate of sideslip angle. However, very little research has been published on HGV sideslip estimation in low friction conditions. This paper proposes three nonlinear Kalman Filters to estimate the tractor sideslip angle of a tractor-semitrailer combination. Performance is compared in simulation to a linear Kalman Filter in both high and low friction conditions. An Unscented Kalman Filter using a yaw-roll vehicle model and nonlinear tire model is found to accurately estimate sideslip in all maneuvers simulated, significantly outperforming the linear Kalman Filter.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131306470","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225670
Niclas Evestedt, Daniel Axehill, M. Trincavelli, F. Gustafsson
In this paper an extension to the sampling based motion planning framework CL-RRT is presented. The framework uses a system model and a stabilizing controller to sample the perceived environment and build a tree of possible trajectories that are evaluated for execution. Complex system models and constraints are easily handled by a forward simulation making the framework widely applicable. To increase operational safety we propose a sampling recovery scheme that performs a deterministic brake profile regeneration using collision information from the forward simulation. This greatly increases the number of safe trajectories and also reduces the number of samples that produce infeasible results. We apply the framework to a Scania G480 mining truck and evaluate the algorithm in a simple yet challenging obstacle course and show that our approach greatly increases the number of feasible paths available for execution.
{"title":"Sampling recovery for closed loop rapidly expanding random tree using brake profile regeneration","authors":"Niclas Evestedt, Daniel Axehill, M. Trincavelli, F. Gustafsson","doi":"10.1109/IVS.2015.7225670","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225670","url":null,"abstract":"In this paper an extension to the sampling based motion planning framework CL-RRT is presented. The framework uses a system model and a stabilizing controller to sample the perceived environment and build a tree of possible trajectories that are evaluated for execution. Complex system models and constraints are easily handled by a forward simulation making the framework widely applicable. To increase operational safety we propose a sampling recovery scheme that performs a deterministic brake profile regeneration using collision information from the forward simulation. This greatly increases the number of safe trajectories and also reduces the number of samples that produce infeasible results. We apply the framework to a Scania G480 mining truck and evaluate the algorithm in a simple yet challenging obstacle course and show that our approach greatly increases the number of feasible paths available for execution.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130358760","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225819
M. Islam, L. Laine, B. Jacobson
This paper describes an advance controller designed using the nonlinear inversion technique of a Modelica® based simulation tool, such as Dymola®, for active dolly steering of a high capacity transport vehicle. Actuator dynamics is included in the inverse model controller. Therefore, it can automatically generate required steering angle request for the dolly axles of the vehicle combination. The resultant controller is transfered as a functional mock-up unit (FMU) to Simulink® environment where the actual simulations are conducted. The controller is simulated against a high-fidelity vehicle model of an A-double combination from Virtual Truck Models (VTM) library - developed by Volvo Group Trucks Technology. Effects of variations of the actual actuator dynamics, with respect to the modeled dynamics in the inverse model controller, on overall vehicle performance are investigated.
{"title":"Inverse model control including actuator dynamics for active dolly steering in high capacity transport vehicle","authors":"M. Islam, L. Laine, B. Jacobson","doi":"10.1109/IVS.2015.7225819","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225819","url":null,"abstract":"This paper describes an advance controller designed using the nonlinear inversion technique of a Modelica® based simulation tool, such as Dymola®, for active dolly steering of a high capacity transport vehicle. Actuator dynamics is included in the inverse model controller. Therefore, it can automatically generate required steering angle request for the dolly axles of the vehicle combination. The resultant controller is transfered as a functional mock-up unit (FMU) to Simulink® environment where the actual simulations are conducted. The controller is simulated against a high-fidelity vehicle model of an A-double combination from Virtual Truck Models (VTM) library - developed by Volvo Group Trucks Technology. Effects of variations of the actual actuator dynamics, with respect to the modeled dynamics in the inverse model controller, on overall vehicle performance are investigated.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116591438","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225803
Yonghwan Jeong, Kyuwon Kim, Beomjun Kim, Jihyun Yoon, Hyok-Jin Chong, Bongchul Ko, K. Yi
This paper presents a vehicle sensor and actuator fault detection algorithm for automated vehicles. The diagnostic system is designed to monitor steering wheel angle, yaw-rate, and wheel speed sensors and steering, throttle, and brake actuators used by the lateral and longitudinal controllers of the vehicle. Different combinations of the observer estimates, the sensor measurements, and the control commands are used to construct a bank of residuals. A fault in any of the vehicle sensors and actuators leads to increase of the unique subset of residuals. The adaptive threshold is used to enable exact identification of the abnormal increase of residual. The fault detection performance and its reliability of the proposed algorithm have been investigated via computer simulation studies and real-time vehicle tests. The enhancement of the fault detection allows for realization of autonomous driving vehicle which uses actuation by embedded computer.
{"title":"Vehicle sensor and actuator fault detection algorithm for automated vehicles","authors":"Yonghwan Jeong, Kyuwon Kim, Beomjun Kim, Jihyun Yoon, Hyok-Jin Chong, Bongchul Ko, K. Yi","doi":"10.1109/IVS.2015.7225803","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225803","url":null,"abstract":"This paper presents a vehicle sensor and actuator fault detection algorithm for automated vehicles. The diagnostic system is designed to monitor steering wheel angle, yaw-rate, and wheel speed sensors and steering, throttle, and brake actuators used by the lateral and longitudinal controllers of the vehicle. Different combinations of the observer estimates, the sensor measurements, and the control commands are used to construct a bank of residuals. A fault in any of the vehicle sensors and actuators leads to increase of the unique subset of residuals. The adaptive threshold is used to enable exact identification of the abnormal increase of residual. The fault detection performance and its reliability of the proposed algorithm have been investigated via computer simulation studies and real-time vehicle tests. The enhancement of the fault detection allows for realization of autonomous driving vehicle which uses actuation by embedded computer.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114962780","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225836
J. Funke, Matthew Brown, Stephen M. Erlien, J. C. Gerdes
One approach to autonomous vehicle control is to generate and then track a desired trajectory without explicit consideration of vehicle stability. Stabilization is then entrusted to the vehicle's built-in production systems, such as electronic stability control, which constantly augment driving inputs to ensure stability. Other approaches explicitly consider stabilization criteria and implement permanently active constraints on the vehicle's actions. Situations exist, however, where enforcing stability constraints could lead to an otherwise avoidable collision. This paper presents an alternative paradigm for autonomous vehicle control that explicitly considers vehicle stability and environmental boundaries as it attempts to track a trajectory; such a mediator can choose to violate short term stability constraints in order to avoid a collision. Model predictive control provides an implementation framework, and an autonomous vehicle demonstrates the viability of the controller as it performs aggressive maneuvers. Driving around a turn at the vehicle's limits exhibits the importance of vehicle stability for autonomous vehicle control. Performing an emergency double lane change, however, highlights a situation where stability criteria must be temporarily violated to avoid a collision.
{"title":"Prioritizing collision avoidance and vehicle stabilization for autonomous vehicles","authors":"J. Funke, Matthew Brown, Stephen M. Erlien, J. C. Gerdes","doi":"10.1109/IVS.2015.7225836","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225836","url":null,"abstract":"One approach to autonomous vehicle control is to generate and then track a desired trajectory without explicit consideration of vehicle stability. Stabilization is then entrusted to the vehicle's built-in production systems, such as electronic stability control, which constantly augment driving inputs to ensure stability. Other approaches explicitly consider stabilization criteria and implement permanently active constraints on the vehicle's actions. Situations exist, however, where enforcing stability constraints could lead to an otherwise avoidable collision. This paper presents an alternative paradigm for autonomous vehicle control that explicitly considers vehicle stability and environmental boundaries as it attempts to track a trajectory; such a mediator can choose to violate short term stability constraints in order to avoid a collision. Model predictive control provides an implementation framework, and an autonomous vehicle demonstrates the viability of the controller as it performs aggressive maneuvers. Driving around a turn at the vehicle's limits exhibits the importance of vehicle stability for autonomous vehicle control. Performing an emergency double lane change, however, highlights a situation where stability criteria must be temporarily violated to avoid a collision.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"16 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129503279","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225697
M. Stuebler, J. Wiest, K. Dietmayer
This paper introduces a precise self-localization method for road vehicles. The presented approach is based on a single grayscale camera in addition with a conventional estimation of the ego motion and a map of the environment. This map is built in advance and independently from the localization process utilizing the same techniques. The proposed algorithm is based on Maximally Stable Extremal Regions which are robust features that are extracted from grayscale images. These features are matched in consecutive images using moment invariants. Together with an estimation of the ego motion, a 3D reconstruction of corresponding landmarks is obtained by applying multiple view geometry. For the unsupervised mapping process, landmarks are tracked and their corresponding global coordinates are stored in a geospatial database using a high-precision real-time kinematic system. The localization process itself is based on a particle filter to estimate the pose of the vehicle by making use of the previously generated map and currently observed landmarks. A standard GPS receiver is used to initialize the pose estimate. The evaluation with real world data shows that this approach achieves very good results despite the marginal sensor setup.
{"title":"Feature-based mapping and self-localization for road vehicles using a single grayscale camera","authors":"M. Stuebler, J. Wiest, K. Dietmayer","doi":"10.1109/IVS.2015.7225697","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225697","url":null,"abstract":"This paper introduces a precise self-localization method for road vehicles. The presented approach is based on a single grayscale camera in addition with a conventional estimation of the ego motion and a map of the environment. This map is built in advance and independently from the localization process utilizing the same techniques. The proposed algorithm is based on Maximally Stable Extremal Regions which are robust features that are extracted from grayscale images. These features are matched in consecutive images using moment invariants. Together with an estimation of the ego motion, a 3D reconstruction of corresponding landmarks is obtained by applying multiple view geometry. For the unsupervised mapping process, landmarks are tracked and their corresponding global coordinates are stored in a geospatial database using a high-precision real-time kinematic system. The localization process itself is based on a particle filter to estimate the pose of the vehicle by making use of the previously generated map and currently observed landmarks. A standard GPS receiver is used to initialize the pose estimate. The evaluation with real world data shows that this approach achieves very good results despite the marginal sensor setup.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129964701","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225654
Fumito Shinmura, Yasutomo Kawanishi, Daisuke Deguchi, I. Ide, H. Murase, H. Fujiyoshi
This paper proposes a method for pedestrian orientation classification. In image recognition, the accuracy is often degraded by the influence of background. In addition, it is also difficult to remove the background and extract only the human body from an image. To overcome these problems, we utilize a single-chip RGB-ToF camera. This camera can acquire RGB and depth images along the same optical axis at the same moment, and thus segmentation of the RGB image becomes easier by using the coaxial depth image. Our proposed method segmented a human body from its background accurately, which lead to the improvement of the accuracy of pedestrian orientation classification.
{"title":"Pedestrian orientation classification utilizing single-chip coaxial RGB-ToF camera","authors":"Fumito Shinmura, Yasutomo Kawanishi, Daisuke Deguchi, I. Ide, H. Murase, H. Fujiyoshi","doi":"10.1109/IVS.2015.7225654","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225654","url":null,"abstract":"This paper proposes a method for pedestrian orientation classification. In image recognition, the accuracy is often degraded by the influence of background. In addition, it is also difficult to remove the background and extract only the human body from an image. To overcome these problems, we utilize a single-chip RGB-ToF camera. This camera can acquire RGB and depth images along the same optical axis at the same moment, and thus segmentation of the RGB image becomes easier by using the coaxial depth image. Our proposed method segmented a human body from its background accurately, which lead to the improvement of the accuracy of pedestrian orientation classification.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132862145","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 : 2015-08-27DOI: 10.1109/IVS.2015.7225824
Hailong Liu, T. Taniguchi, Yusuke Tanaka, Kazuhito Takenaka, T. Bando
Driving behavior can be represented by many different types of measured sensor information obtained through a control area network. We assume that the measured sensor information is generated from several hidden time-series data through multiple nonlinear transformations. These hidden time-series data are statistically independent of each other and capture essential driving behavior. Driving behavior information is usually generated by multiple nonlinear transformations that fuse essential features, e.g., "Yaw rate" is generated by fusing the velocity of the vehicle and the change of driving direction. However, driving behavior data is often redundant because such data includes multivariate information and involves duplicated essential features. In this paper, we propose a feature extraction method to extract essential features from redundant driving behavior data using a deep sparse autoencoder (DSAE), which is a deep learning method. Two-dimensional features are extracted from seven-dimensional artificial data using a DSAE and are determined experimentally to be highly correlated with the prepared essential features. DSAEs are also used to extract features from an actual driving behavior data set. To verify a DSAE's ability to extract essential driving behavior features and filter out redundant information, we prepare twelve data sets that include some or all of the driving behavior information. Twelve DSAEs are used to independently extract features from the twelve prepared data sets, and canonical correlation analysis is used to analyze the canonical correlation coefficients between extracted features. Furthermore, we verify DSAEs' ability to extract essential driving behavior features from the redundant driving behavior data sets.
{"title":"Essential feature extraction of driving behavior using a deep learning method","authors":"Hailong Liu, T. Taniguchi, Yusuke Tanaka, Kazuhito Takenaka, T. Bando","doi":"10.1109/IVS.2015.7225824","DOIUrl":"https://doi.org/10.1109/IVS.2015.7225824","url":null,"abstract":"Driving behavior can be represented by many different types of measured sensor information obtained through a control area network. We assume that the measured sensor information is generated from several hidden time-series data through multiple nonlinear transformations. These hidden time-series data are statistically independent of each other and capture essential driving behavior. Driving behavior information is usually generated by multiple nonlinear transformations that fuse essential features, e.g., \"Yaw rate\" is generated by fusing the velocity of the vehicle and the change of driving direction. However, driving behavior data is often redundant because such data includes multivariate information and involves duplicated essential features. In this paper, we propose a feature extraction method to extract essential features from redundant driving behavior data using a deep sparse autoencoder (DSAE), which is a deep learning method. Two-dimensional features are extracted from seven-dimensional artificial data using a DSAE and are determined experimentally to be highly correlated with the prepared essential features. DSAEs are also used to extract features from an actual driving behavior data set. To verify a DSAE's ability to extract essential driving behavior features and filter out redundant information, we prepare twelve data sets that include some or all of the driving behavior information. Twelve DSAEs are used to independently extract features from the twelve prepared data sets, and canonical correlation analysis is used to analyze the canonical correlation coefficients between extracted features. Furthermore, we verify DSAEs' ability to extract essential driving behavior features from the redundant driving behavior data sets.","PeriodicalId":294701,"journal":{"name":"2015 IEEE Intelligent Vehicles Symposium (IV)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130749553","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：