Design and Implementation of an SAE Level-2 Lane Keeping System for Class 8 Trucks Using Nonlinear Model Predictive Control. Ward, J. W., Pierce, J. D., Brown, L., & Bevly, D. M. In 2023 IEEE Conference on Control Technology and Applications (CCTA), pages 841–846, August, 2023. ISSN: 2768-0770![Design and Implementation of an SAE Level-2 Lane Keeping System for Class 8 Trucks Using Nonlinear Model Predictive Control [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex This paper focuses on the design and evaluation of a lateral Nonlinear Model Predictive Control (NMPC) path following algorithm for Class 8 vehicles. While NMPC allows for the inclusion of constraints such as obstacle avoidance or rollover prevention, this work is primarily focused on testing the path following capabilities of the NMPC controller without the consideration of constraints. Although many lateral controllers for tractor-trailer systems use either fully linear or fully nonlinear Equations of Motion (EOMs), this paper presents a hybrid model. This hybrid model uses a linear model to describe the dynamics of the tractor-trailer while using nonlinear equations to describe the global position of the system. This model is then implemented in a NMPC architecture. The controller is first designed and tested in MATLAB to yield a controller that can follow a lane change maneuver. This controller is then implemented in a real-time format using ROS and C++. The objective of the real-time controller was to replicate the path of a Kia Optima which was driving in front of the test vehicle. To accomplish this, an estimator was developed to calculate the relative path between the vehicles with enough accuracy to stay within the standard road lane. This estimation algorithm was able to produce lateral position errors with a standard deviation of 1.96 cm. Finally, the lateral controller was shown to track the generated reference path with a mean error of 1.47 cm and a RMS error of under 27 cm.
@inproceedings{ward_design_2023,
title = {Design and {Implementation} of an {SAE} {Level}-2 {Lane} {Keeping} {System} for {Class} 8 {Trucks} {Using} {Nonlinear} {Model} {Predictive} {Control}},
url = {https://ieeexplore.ieee.org/abstract/document/10253308},
doi = {10.1109/CCTA54093.2023.10253308},
abstract = {This paper focuses on the design and evaluation of a lateral Nonlinear Model Predictive Control (NMPC) path following algorithm for Class 8 vehicles. While NMPC allows for the inclusion of constraints such as obstacle avoidance or rollover prevention, this work is primarily focused on testing the path following capabilities of the NMPC controller without the consideration of constraints. Although many lateral controllers for tractor-trailer systems use either fully linear or fully nonlinear Equations of Motion (EOMs), this paper presents a hybrid model. This hybrid model uses a linear model to describe the dynamics of the tractor-trailer while using nonlinear equations to describe the global position of the system. This model is then implemented in a NMPC architecture. The controller is first designed and tested in MATLAB to yield a controller that can follow a lane change maneuver. This controller is then implemented in a real-time format using ROS and C++. The objective of the real-time controller was to replicate the path of a Kia Optima which was driving in front of the test vehicle. To accomplish this, an estimator was developed to calculate the relative path between the vehicles with enough accuracy to stay within the standard road lane. This estimation algorithm was able to produce lateral position errors with a standard deviation of 1.96 cm. Finally, the lateral controller was shown to track the generated reference path with a mean error of 1.47 cm and a RMS error of under 27 cm.},
urldate = {2024-06-20},
booktitle = {2023 {IEEE} {Conference} on {Control} {Technology} and {Applications} ({CCTA})},
author = {Ward, Jacob W. and Pierce, J. Daniel and Brown, Lowell and Bevly, David M.},
month = aug,
year = {2023},
note = {ISSN: 2768-0770},
keywords = {Collision avoidance, GNSS, Lateral Control, Mathematical models, NMPC, Nonlinear dynamical systems, Nonlinear equations, Path Duplication, Real-time systems, Rollover, Transient analysis},
pages = {841--846},
}
Downloads: 0
{"_id":"Bp8yeDmju467Ndyb4","bibbaseid":"ward-pierce-brown-bevly-designandimplementationofansaelevel2lanekeepingsystemforclass8trucksusingnonlinearmodelpredictivecontrol-2023","author_short":["Ward, J. W.","Pierce, J. D.","Brown, L.","Bevly, D. M."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"Design and Implementation of an SAE Level-2 Lane Keeping System for Class 8 Trucks Using Nonlinear Model Predictive Control","url":"https://ieeexplore.ieee.org/abstract/document/10253308","doi":"10.1109/CCTA54093.2023.10253308","abstract":"This paper focuses on the design and evaluation of a lateral Nonlinear Model Predictive Control (NMPC) path following algorithm for Class 8 vehicles. While NMPC allows for the inclusion of constraints such as obstacle avoidance or rollover prevention, this work is primarily focused on testing the path following capabilities of the NMPC controller without the consideration of constraints. Although many lateral controllers for tractor-trailer systems use either fully linear or fully nonlinear Equations of Motion (EOMs), this paper presents a hybrid model. This hybrid model uses a linear model to describe the dynamics of the tractor-trailer while using nonlinear equations to describe the global position of the system. This model is then implemented in a NMPC architecture. The controller is first designed and tested in MATLAB to yield a controller that can follow a lane change maneuver. This controller is then implemented in a real-time format using ROS and C++. The objective of the real-time controller was to replicate the path of a Kia Optima which was driving in front of the test vehicle. To accomplish this, an estimator was developed to calculate the relative path between the vehicles with enough accuracy to stay within the standard road lane. This estimation algorithm was able to produce lateral position errors with a standard deviation of 1.96 cm. Finally, the lateral controller was shown to track the generated reference path with a mean error of 1.47 cm and a RMS error of under 27 cm.","urldate":"2024-06-20","booktitle":"2023 IEEE Conference on Control Technology and Applications (CCTA)","author":[{"propositions":[],"lastnames":["Ward"],"firstnames":["Jacob","W."],"suffixes":[]},{"propositions":[],"lastnames":["Pierce"],"firstnames":["J.","Daniel"],"suffixes":[]},{"propositions":[],"lastnames":["Brown"],"firstnames":["Lowell"],"suffixes":[]},{"propositions":[],"lastnames":["Bevly"],"firstnames":["David","M."],"suffixes":[]}],"month":"August","year":"2023","note":"ISSN: 2768-0770","keywords":"Collision avoidance, GNSS, Lateral Control, Mathematical models, NMPC, Nonlinear dynamical systems, Nonlinear equations, Path Duplication, Real-time systems, Rollover, Transient analysis","pages":"841–846","bibtex":"@inproceedings{ward_design_2023,\n\ttitle = {Design and {Implementation} of an {SAE} {Level}-2 {Lane} {Keeping} {System} for {Class} 8 {Trucks} {Using} {Nonlinear} {Model} {Predictive} {Control}},\n\turl = {https://ieeexplore.ieee.org/abstract/document/10253308},\n\tdoi = {10.1109/CCTA54093.2023.10253308},\n\tabstract = {This paper focuses on the design and evaluation of a lateral Nonlinear Model Predictive Control (NMPC) path following algorithm for Class 8 vehicles. While NMPC allows for the inclusion of constraints such as obstacle avoidance or rollover prevention, this work is primarily focused on testing the path following capabilities of the NMPC controller without the consideration of constraints. Although many lateral controllers for tractor-trailer systems use either fully linear or fully nonlinear Equations of Motion (EOMs), this paper presents a hybrid model. This hybrid model uses a linear model to describe the dynamics of the tractor-trailer while using nonlinear equations to describe the global position of the system. This model is then implemented in a NMPC architecture. The controller is first designed and tested in MATLAB to yield a controller that can follow a lane change maneuver. This controller is then implemented in a real-time format using ROS and C++. The objective of the real-time controller was to replicate the path of a Kia Optima which was driving in front of the test vehicle. To accomplish this, an estimator was developed to calculate the relative path between the vehicles with enough accuracy to stay within the standard road lane. This estimation algorithm was able to produce lateral position errors with a standard deviation of 1.96 cm. Finally, the lateral controller was shown to track the generated reference path with a mean error of 1.47 cm and a RMS error of under 27 cm.},\n\turldate = {2024-06-20},\n\tbooktitle = {2023 {IEEE} {Conference} on {Control} {Technology} and {Applications} ({CCTA})},\n\tauthor = {Ward, Jacob W. and Pierce, J. Daniel and Brown, Lowell and Bevly, David M.},\n\tmonth = aug,\n\tyear = {2023},\n\tnote = {ISSN: 2768-0770},\n\tkeywords = {Collision avoidance, GNSS, Lateral Control, Mathematical models, NMPC, Nonlinear dynamical systems, Nonlinear equations, Path Duplication, Real-time systems, Rollover, Transient analysis},\n\tpages = {841--846},\n}\n\n\n\n","author_short":["Ward, J. W.","Pierce, J. D.","Brown, L.","Bevly, D. M."],"key":"ward_design_2023","id":"ward_design_2023","bibbaseid":"ward-pierce-brown-bevly-designandimplementationofansaelevel2lanekeepingsystemforclass8trucksusingnonlinearmodelpredictivecontrol-2023","role":"author","urls":{"Paper":"https://ieeexplore.ieee.org/abstract/document/10253308"},"keyword":["Collision avoidance","GNSS","Lateral Control","Mathematical models","NMPC","Nonlinear dynamical systems","Nonlinear equations","Path Duplication","Real-time systems","Rollover","Transient analysis"],"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://bibbase.org/zotero-group/keb0115/5574615","dataSources":["kDK6fZ4EDThxNKDCP"],"keywords":["collision avoidance","gnss","lateral control","mathematical models","nmpc","nonlinear dynamical systems","nonlinear equations","path duplication","real-time systems","rollover","transient analysis"],"search_terms":["design","implementation","sae","level","lane","keeping","system","class","trucks","using","nonlinear","model","predictive","control","ward","pierce","brown","bevly"],"title":"Design and Implementation of an SAE Level-2 Lane Keeping System for Class 8 Trucks Using Nonlinear Model Predictive Control","year":2023}