Predicting Vehicle Behavior Using Automotive Radar and Recurrent Neural Networks

This study reviews the Artificial Intelligence and Machine Learning approaches developed thus far for driver profile and driving pattern recognition, representing a set of macroscopic and microscopic behaviors respectively, to enhance the understanding of human factors in road safety, and therefore reduce the number of crashes. It provides a definition of the two scientific fields in terms of safety, and identifies the most efficient approaches used regarding methodology, data collection and driving metrics. Results show that K-means and Neural Networks are the most commonly used methodologies for driver profile identification, and Dynamic Time Warping for driving pattern detection. Most studies discovered driver profiles related to aggressiveness, considering mainly speed and acceleration as driving metrics. Based on the gaps and challenges identified, this paper provides a new framework for combining microscopic and macroscopic driving behavior analysis, instead of examining them separately as is the state-of-theart. Such combined results can potentially improve the development of traffic risk models, which could be exploited in applications that monitor drivers in real-time and provide feedback. These models will represent human behavior more accurately, which can eventually lead to the recognition of "optimal" human driving patterns that Automated Vehicles (AV) could 'mimic' to become safer.

Section: A Main Findingsmentioning

confidence: 99%

Driver Profile and Driving Pattern Recognition for Road Safety Assessment: Main Challenges and Future Directions

Tselentis

Papadimitriou

2023

“…A novel sideslip angle estimation method integrating deep neural networks and nonlinear Kalman filters is presented in a research. A recurrent neural network [44] with large shortterm memory, which is useful for analysing sequential sensor data, is part of the deep neural network, which is used to increase the robustness of the estimation and to capture its uncertainty. The sideslip angle estimate and associated uncertainty are provided by the deep neural network, which is trained utilising input sets made up of on-board sensor measurements (yaw rate, velocity, steering wheel angle, and lateral acceleration).…”

Section: Introductionmentioning

confidence: 99%

On the Prediction of the Sideslip Angle Using Dynamic Neural Networks

Marotta,

Strano,

Terzo

et al. 2024

With the growing interest in self-driving vehicles, safety in vehicle driving is becoming an increasingly important aspect. The sideslip angle is a key quantity for modern control systems that aim to improve passenger safety. It directly affects the lateral motion and stability of a vehicle. In particular, a large sideslip angle can cause the vehicle to experience oversteer or understeer, which can lead to loss of control and potentially result in an accident. For this reason, it is necessary to constantly monitor this quantity while driving in order to implement appropriate action if necessary. Sensors that directly measure this quantity are expensive and difficult to implement. In this paper, two neural networks to estimate the sideslip angle are proposed. The quantities that most influence the vehicle's sideslip angle were assessed. Furthermore, the neural networks can exploit data from previous instants of time for estimation purposes. In particular, the first uses lateral acceleration and steering wheel angle as input, the second uses longitudinal acceleration, lateral acceleration and yaw rate. Experimental tests carried out on manoeuvres that stimulate the sideslip angle have shown that, although the estimators use few measures, they are able to accurately estimate the quantity of interest.

“…Object tracking uses continuous multi-frame detection results to filter inaccurate measurements and estimate high-order states (velocity, acceleration, etc.). For vehicles in a structured road environment, their motion is regular [6], and object tracking can predict the objects' states accurately by using the historical motion state and motion model. These prediction results can be used as candidate proposals for object detection to improve the number of proposals.…”

Section: Introductionmentioning

confidence: 99%

3-D Vehicle Detection Enhancement Using Tracking Feedback in Sparse Point Clouds Environments

Qian

Wang

Zhuang

et al. 2023

In recent years, vehicle detection in intelligent transportation systems using 3D LIDAR point clouds based on deep neural networks has made substantial progress. However, when the point clouds are very sparse, the detection model cannot generate proposals efficiently, resulting in false negative results. Considering that the object tracking technology accurately predicts vehicles based on historical measurements and motion models, and these prediction results can become proposals for object detection. Therefore, this paper proposes a novel object detection paradigm based on tracking feedback to address the false negative problem based on sparse point clouds. According to the distribution of the state vector from the Kalman prediction, multiple proposals are sampled and fed back to the second stage of twostage detection models. After regression and non-maximum suppression, the false negative results can be effectively reduced. This method enhances the vehicle detection capability of classical neural networks. Comparing the recall metric of multiple detection models at different distances in the public KITTI and nuSences datasets, the proposed method can promote up to 5.31% compared to the previous method, which reflects the effectiveness and versatility of the proposed method. INDEX TERMS Vehicle detection, tracking feedback, sparse point clouds. YEQIANG QIAN (Member, IEEE) received the Ph.D. degree in control science and engineering from