Methodology of hierarchical collision avoidance for high‐speed self‐driving vehicle based on motion‐decoupled extraction of scenarios

Liu, Zhaolin; Chen, Jiqing; Lan, Fengchong; Xia, Hongyang

doi:10.1049/iet-its.2019.0334

Cited by 10 publications

(4 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Comparing the motion control of aerial vehicles with surface transport allows concluding that of surface autonomous vehicles have a higher probability of collision because there is no ability to change the altitude component. A hierarchical collision-avoidance strategy is proposed as a theoretically approach for high-speed self-driving vehicle collision-avoiding task solving [7]. Fuzzy logic and FMCW radar (Frequency-Modulated Continuous Wave) usability for motion control task solving are described in research [8], but this system also is not provided with surrounding environment visual monitoring elements or sensors.…”

Section: Review Of the Scientific Research And Literature About Auton...mentioning

confidence: 99%

Immune neural network machine learning of autonomous drones for energy efficiency and collision prevention

Gorobetz,

Ribickis,

Beinarovica

et al. 2023

Drones - Various Applications

View full text Add to dashboard Cite

The chapter is related to the safe and energy-efficient motion of autonomous drones and describes the developed novel immune neural network-based machine learning technology for its control. The technology is inspired by two biological systems – immune system and neural networks and their artificial analogs and evolutionary theory. The developed novel mathematical models and algorithms for this technology allow skipping the preliminary supervised training step and adapted for real-time continuous unsupervised self-learning of the drone to recognize the dangerous situation, to prevent the collision by making control decisions autonomously and continuously learning to keep optimal energy consumption during the motion. The chapter includes the study of existing neural network-based solutions for the recognition and prevention of dangerous situations and energy efficiency of drones, describes the developed target function and algorithm for immune neural network machine learning technology and simulation and experimental results proving the efficiency of this technology.

show abstract

Section: Review Of the Scientific Research And Literature About Auton...mentioning

confidence: 99%

Immune neural network machine learning of autonomous drones for energy efficiency and collision prevention

Gorobetz,

Ribickis,

Beinarovica

et al. 2023

Drones - Various Applications

View full text Add to dashboard Cite

show abstract

“…Therefore, when various vehicles are under high-speed working conditions, the control stability of the vehicle may not meet requirements. Therefore, a vehicle tracking control model that integrates dynamic constraints is designed here [16]. This is also the fundamental condition for designing and improving vehicle collision avoidance models in the future.…”

Section: Design Of Vehicle Collision Avoidance Model Combining Mpc An...mentioning

confidence: 99%

Collision Avoidance Path Planning for Vehicles Combining MPC and CACC Controllers

Zeng

Liu²

2023

IEEE Access

View full text Add to dashboard Cite

The vehicle collision avoidance system has problems such as poor emergency collision avoidance ability and susceptibility to noise interference. Therefore, this study adopts a vehicle tracking model that integrates dynamic constraints and a two car following model to construct a Collaborative Adaptive Cruise Control (CACC) model that integrates Multi Constraints Model Predictive Control (MPC). And conducted simulation experimental research. The results show that in emergency braking collision avoidance conditions, the collision avoidance model designed in this study can quickly respond to the distance control requirements, and the control stability is higher than that of common collision avoidance systems. The standard deviation of the rear car acceleration of the MPC+CACC, reinforcement learning (RL), and proportional-integral-differential (PID) models are 0.18m/s2, 0.25m/s2, and 0.46m/s2, respectively. The car spacing error of the MPC+CACC, RL, and PID models converge to 0m when the time is about 13 seconds, 33 seconds, and 40 seconds, respectively. This study has certain reference significance for improving the collision avoidance ability of China's intelligent vehicle collision avoidance system.

show abstract

“…Displacement, lateral velocity, lateral acceleration and lane-changing time were key optimisation variables of lane change trajectory [20][21][22]. Most lane-changing trajectory planning methods based on numerical optimisation relied on other methods to find the optimal trajectory [23,24]. Zhang et al [25] used the time-dependent cubic polynomial equation and presented a cost function considering the driving comfort and efficiency to optimise the lane-changing trajectory.…”

Section: Lane-changing Trajectory Modelmentioning

confidence: 99%

“…Displacement, lateral velocity, lateral acceleration and lane‐changing time were key optimisation variables of lane change trajectory [20–22]. Most lane‐changing trajectory planning methods based on numerical optimisation relied on other methods to find the optimal trajectory [23, 24]. Zhang et al.…”

Section: Introductionmentioning

confidence: 99%

Time‐dependent lane change trajectory optimisation considering comfort and efficiency for lateral collision avoidance

Guo

et al. 2021

IET Intelligent Trans Sys

View full text Add to dashboard Cite

Lateral collision is one of the top two accidents in the world and often occurs in the lane change. Trajectory optimisation is an effective method to solve traffic conflicts in the lane-changing process. However, current trajectory optimisation methods are not friendly to human-computer-based driving assistance. This study proposes a timedependent lanechange trajectory optimisation considering comfort and efficiency. First, spacing constraints between the lane-changing vehicle and surrounding vehicles are determined and quantified. Second, lane-changing trajectory data are obtained by driving simulation experiments and extracted by data features. Third, a quintic multinomial model of lane-changing trajectory is proposed. The results reveal that the predicted trajectory is close to the observed value. Then, the obtained trajectory is optimised by the objective function considering lane-changing efficiency and comfort. Finally, a case study is used to demonstrate the application of the model. When a lane-changing vehicle changes a lane at an initial speed of 90 km/h to the faster lane at the terminal speed of 110 km/h, the optimal lane-changing time is 3.4 s, with the lateral acceleration 1.79 m/s 2 and the maximum yaw angle 0.081 rad. 1 INTRODUCTION Lateral and rear-end collisions are the two most collision manoeuvres in the world. According to the national highway traffic safety administration (NHTSA) traffic accident data from 2014 to 2018, rear-end collisions accounted for the highest proportion, reaching 40%. The second was angle collision, accounting for 24%, of which the proportion of lateral collisions was 13% [1]. In China, lateral collisions have a higher percentage. The 2018 traffic accident data revealed that sideswipe accounted for the highest proportion, reaching 42.7%, followed by rearend collision, accounting for 7.95%. Car following is a process of following but not exceeding obstacles, during which there is only one potential traffic conflict, while lane-changing is a process of surpassing obstacles to reach the ideal position, and there are multiple potential traffic conflicts. This is also the reason why side-impact accidents occur frequently in China. The surrounding traffic environment influences the lanechanging process of the lane-changing vehicle. In the lanechanging process, drivers need to analyse and make a This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Methodology of hierarchical collision avoidance for high‐speed self‐driving vehicle based on motion‐decoupled extraction of scenarios

Cited by 10 publications

References 32 publications

Immune neural network machine learning of autonomous drones for energy efficiency and collision prevention

Immune neural network machine learning of autonomous drones for energy efficiency and collision prevention

Collision Avoidance Path Planning for Vehicles Combining MPC and CACC Controllers

Time‐dependent lane change trajectory optimisation considering comfort and efficiency for lateral collision avoidance

Contact Info

Product

Resources

About