Vehicle Collision Avoidance

Accompanying appendix to presentation July 2013 to the National Academy of Forensic Engineers Advanced Accident Reconstruction Special Seminar. Appendix demonstrates methodology presented, calculation methods and statistical evaluation of methodology to actual collision tests.

Forward Collision Avoidance (FCA) systems in automobiles is an essential part of Advanced Driver Assistance System (ADAS) and autonomous vehicles. These devices currently use, radars as the main sensor. The increasing resolution of camera sensors, processing capability of hardware chipsets and advances in image processing algorithms, have been pushing the camera based features recently. Monocular cameras face the challenge of accurate scale estimation which limits it use as a stand-alone sensor for this application. This paper proposes an efficient system which can perform multi scale object detection which is being patent granted and efficient 3D reconstruction using structure from motion (SFM) framework. While the algorithms need to be accurate it also needs to operate real time in low cost embedded hardware. The focus of the paper is to discuss how the proposed algorithms are designed in such a way that it can be provide real time performance on low cost embedded CPU's which makes use of only Digital Signal processors (DSP) and vector processing cores.

Presentation during July 2013 National Academy of Forensic Engineers Advanced Accident Reconstruction Seminar Series. Development of methods for determining the impact speeds and severity levels of minor damage accidents using a specially developed Force-Deflection analysis methodology that incorporates impact restitution and tire force contributions into the collision event. Examples to vehicle-to-vehicle testing are provided.

Collision Avoidance (CA) system is a pivotal part of the autonomous vehicle. Despite the reported advancements, development of CA system in the presence of uncertain obstacles still possesses issues. These include untimely actuators intervention, unreliable planning and tracking strategies, near-miss incidents as well as hazardous lane departure while avoiding obstacles. During the avoidance of sudden appearance of obstacles, the abrupt increment of vehicle longitudinal and lateral forces summation during the maneuver demands a system with the ability to handle the coupled nonlinear dynamics. To address these issues, this study aims to develop an integrated nonlinear control strategy for CA system of uncertain obstacles in multi-risk scenarios. A Nonlinear Model Predictive Control-based (NMPC) path tracking strategy is proposed as the automated motion guidance for the host vehicle which is integrated with the Artificial Potential Field (APF) motion planning strategy. APF measures the collision risks and formulates the desired references for the path replanning. It ensures an optimal replanned trajectory by including the vehicle dynamics into its optimization formulation. NMPC then acts as the coupled path and speed tracking controller to enable vehicle navigation. To accommodate vehicle comfort during the avoidance, NMPC is constrained. To decrease the system's computational burden, Move Blocking strategy is assimilated within the NMPC design. To validate the proposed controller, a research platform vehicle is developed and modeled. The performance of the CA system is compared with four well known integrated control strategies. The consideration of vehicle nonlinearity into NMPC is shown to yield lower mean squared error (MSE) compared to linear MPC, with 97.36 percent decrement for yaw rate tracking and 94.36 percent decrement for deceleration tracking. The inclusion of multi-actuators and vehicle states feedback into the controller prevented lane departure after CA navigation. Analysis of the system’s robustness by computational simulations in multi-risk CA scenario with uncertain obstacles revealed that collisions were avoided in all scenarios. Move Blocking strategies managed to reduce the NMPC computation time with the mean value of 2.281 seconds in six scenarios with 25.33 percent increased performance from the uniform blocking. The architecture of CA system successfully maintained the safe-distance of 2 metres towards the collision points during the minimum safe-time, thus preventing near miss incidents. Hence, this study is important for the development of a multi-actuators controller strategy for CA of the fully autonomous vehicle in varied uncertain risk scenarios.

IET indexes its books and journal in SCOPUS and IEEE Xplore.
Computer Vision (CV) and Sensors play a decisive role in the operation of Unmanned Aerial Vehicle (UAV), but there exists a void when it comes to analysing the extent of their impact on the entire UAV system. In general, the fact that a UAV is a Cyber-Physical System (CPS) is not taken into account. In this proposal, we propose to expand on earlier books covering the use of CV and sensing in UAVs. Among other things, an entirely autonomous UAV can help to (i) obtain information about the environment, (ii) work for an extended period of time without human interference, (iii) move either all or part of itself all over its operating location devoid of human help and (iv) stay away from dangerous situations for people and their possessions. A Vision System (VS) entails the way CV data will be utilized, the appropriate architecture for total avionics integration, the control interfaces, and the UAV operation. Since the VS core is its sensors and cameras, multi-sensor fusion, navigation, hazard detection, and ground correlation in real time are important operational aspects that can benefit from CV knowledge and technology. This book will aim to collect and shed some light on the existing information on CV software and hardware for UAVs as well as pinpoint aspects that need additional thinking. It will list standards and a set of prerequisites (or lack of them thereof) when it comes to CV deployment in UAVs. The issue of data fusion takes a centre place when the book explores ways to deal with sensor data and images as well as their integration and display. The best practices to fuse image and sensor information to enhance UAV performance by means of CV can greatly improve all aspects of the corresponding CPS. The CPS viewpoint can improve the way UAVs interact with the Internet of Things (IoT), use cloud computing, meet communications requirements, implement hardware/software paradigms necessary to handle video streaming, incorporate satellite data, and combine CV with Virtual/Augmented Realities.

VOLUME 1-CONTROL AND PERFORMANCE: This tome explores how sensors and computer vision technologies are used in unmanned aerial vehicles for the navigation, control, stability, reliability, guidance, fault detection, self-maintenance, strategic replanning and reconfiguration of the entire system. It helps analyse the manner UAVs interact with the Internet of Things (IoT), use cloud computing, meet communications requirements, implement hardware/software paradigms necessary to handle still imagery, video streaming, incorporate satellite data, and combine computer vision with virtual/augmented realities (VR/AR).NB: This is planned to be the companion volume of Estrela, Hemanth, Saotome (Eds) / Imaging and Sensing for Unmanned Aerial Vehicles: Volume 2-Deployment and Applications

The problem of recursively approximating motion resulting from the Optical Flow (OF) in video thru Total Least Squares (TLS) techniques is addressed. TLS method solves an inconsistent system Gu=z , with G and z in error due to temporal/spatial derivatives, and nonlinearity, while the Ordinary Least Squares (OLS) model has noise only in z. Sources of difficulty involve the non-stationarity of the field, the ill-posedness, and the existence of noise in the data. Three ways of applying the TLS with different noise conjectures to the end problem are observed. First, the classical TLS (cTLS) is introduced, where the entries of the error matrices of each row of the augmented matrix [G;z] have zero mean and the same standard deviation. Next, the Generalized Total Least Squares (GTLS) is defined to provide a more stable solution, but it still has some problems. The Generalized Scaled TLS (GSTLS) has G and z tainted by different sources of additive zero-mean Gaussian noise and scaling [G;z] by nonsingular D and E, that is, D[G;z]E makes the errors iid with zero mean and a diagonal covariance matrix. The scaling is computed from some knowledge on the error distribution to improve the GTLS estimate. For moderate levels of additive noise, GSTLS outperforms the OLS, and the GTLS approaches. Although any TLS variant requires more computations than the OLS, it is still applicable with proper scaling of the data matrix.

Data obtained in teenage cyclist impact against three vehicles (sedan, SUV and Pick Up), tested in previous works, are analyzed to verify the theoretical results by the coupling of momentum conservation principle and energy conservation principle. The speed of vehicle and thorax are compared with theoretical ones, obtaining an excellent agreement. Following the results found in the literature, particular emphasis is done on teenager cyclist thorax and head speeds, indicating that the cyclist speed can be theoretically until 1,41 times the impact speed, while greater values, until 2, are obtained in the simulations. A small slowing of the vehicle is found theoretically and in the simulations, due to kinetic energy transfer. A small component of the vehicle speed exists in the orthogonal plane to the motion, due, for example, to the compression of the rider body on the bonnet. A parameter is determined using all the data in term of thorax speed, in all the impact speeds and in all the relative positions; its value does not appear depending on the vehicle mass. It allows the subsequent determination of the best values of three geometric actual parameters identified in front of the vehicle: bumper height, bonnet height and bonnet inclination angle, by interpolation with a second order curve, by making the conclusion that the frontal part of the vehicle may be designed in order to reduce the injury.

This paper proposes a novel risk management framework, and specifically a methodology to model UAS collision risk. The model inherently accounts for the performance of Communication, Navigation and Surveillance (CNS) systems, as well as aircraft vehicle dynamics in evaluating collision risk. The model is applicable to encounters between two or more aircraft, and to terrain collision scenarios. The methodology is underpinned by modelling of the CNS error characteristics, as well as wind uncertainty, and the translation of those characteristics to the spatial domain to form a virtual risk protection volume around each aircraft. The volume is then inflated in proportion to a Target Level of Safety (TLS). The methodology is demonstrated in a simulation case study representative of aircraft-aircraft collision encounters. A 2σ navigation accuracy bound of 185.2 m was found to incur a maximum collision risk of approximately 1.5×10-4 at a separation of 200m.

We present and validate a framework for visual navigation with obstacle avoidance. The approach was originally designed in [1], but major improvements and real outdoor experiments are added here. Visual navigation consists of following a path, represented as an ordered set of key images, that have been acquired in a preliminary teaching phase. While following such path, the robot is able to avoid new obstacles which were not present during teaching, and which are sensed by a range scanner. We guarantee that collision avoidance and navigation are achieved simultaneously by actuating the camera pan angle, in the presence of obstacles, to maintain scene visibility as the robot circumnavigates the obstacle. The circumnavigation verse and the collision risk are estimated using a potential vector field derived from an occupancy grid. The framework can also deal with unavoidable obstacles, which make the robot decelerate and eventually stop.

Current methods for analyzing motor vehicle deformation utilize a force-deflection analysis for determining deformation work energy, which relies on vehicle-specific structural stiffness coefficients determined from full-scale impact testing. While the current database is quite extensive for frontal stiffness values for passenger cars and many light trucks, vans and SUVs from the 1970’s up to modern day, the database is devoid of specific crash tests needed for deformation analysis of rear and/or side structures of many vehicles. Additionally, there exists very few structural stiffness coefficients for heavy commercial vehicles, buses, recreational vehicles, heavy equipment or motorcycles necessary for application with the current force-deflection analysis methods.
The primary goal of this research is to develop an accurate, reliable and broadly applicable deformation analysis method that requires the structural stiffness coefficients for only one collision involved vehicle. The developed methodology expands the application of deformation analysis to include unconventional vehicles and other objects and surfaces not supported by the current structural stiffness coefficient database. The G-DaTAV™ System of Equations incorporates linear and rotational effects, as well as impact restitution resulting from conservative forces acting during a given collision impulse. Additionally, the G-DaTAV™ System of Equations accounts for tire-ground forces and inter-vehicular friction, non-conservative force contributions acting on the collision system that are commonly present during offset and oblique non-central collision configurations.
Correlation and descriptive statistics, as well as the raw analysis results,  indicate a highly reliable and significantly improved degree of precision and accuracy achieved through the application of the G-DaTADeltaV™ System of Equations when determining vehicular total velocity changes for oblique and offset non-central impacts.

For missions involving multiple Unmanned Aerial Vehicles (UAV), dependable collision avoidance techniques are necessary to ensure mission completion and UAV safety. This paper outlines and implements an algorithm to reduce collisions using Inverse Proportional Navigation (IPN). The algorithm checks for imminent collisions and resolves them by increasing the Line Of Sight (LOS) rate between pairs of UAV. Overall performance is improved by adjusting the algorithm to handle troublesome special cases. A simulator enables the testing of the algorithm in various scenarios for comparison to a basic implementation of IPN and a control. The paper concludes with a discussion of results and opportunities for future research and improvement.

This paper reviews the main hardware/software characteristics and the numerical simulations performed to validate the functionalities of an obstacle warning and avoidance system for various manned and unmanned platform applications. The Laser Obstacle Avoidance and Monitoring (LOAM) system is one of the key non-cooperative sensors adopted for avoiding obstacles/intruders. Robust multi-criteria decision logics are adopted to identify the optimal avoidance manoeuvre considering all currently and previously detected obstacles by employing a so-called history function. Additionally, a new formulation is adopted for defining the overall uncertainty volumes associated with the detected obstacles. The demonstrated detection performances and the robust trajectory generation algorithm ensure a safe avoidance of all classes of obstacles (i.e. ground and aerial) in most weather conditions and flight phases.

Millimeter wave (mmWave) communication systems use large number of antenna elements that can potentially overcome severe channel attenuation by narrow beamforming. Narrow-beam operations in mmWave networks also reduce multiuser interference, introducing the concept of noise-limited wireless networks as opposed to interference-limited ones. The noise-limited or interference-limited regime heavily reflects on the medium access control (MAC) layer throughput and on proper resource allocation and interference management strategies. Yet, these regimes are ignored in current approaches to mmWave MAC layer design, with the potential disastrous consequences on the communication performance. In this paper, these regimes are investigated in terms of collision probability and throughput. Tractable closed-form expressions for collision probability and MAC layer throughput of mmWave ad hoc networks, operating under slotted ALOHA, are derived. The new analysis reveals that mmWave networks may exhibit a non-negligible transitional behavior from a noise-limited regime to an interference-limited one, depending on the density of the transmitters, density and size of obstacles, transmission probability, operating beamwidth, and transmission power. Such transitional behavior necessitates a new framework of adaptive hybrid resource allocation procedure, containing both contention-based and contention-free phases with on-demand realization of the contention-free phase. Moreover, the conventional collision avoidance procedure in the contention-based phase should be revisited, due to the transitional behavior of interference, to maximize throughput/delay performance of mmWave networks. It is concluded that, unless proper hybrid schemes are investigated, the severity of the transitional behavior may significantly reduce throughput/delay performance of mmWave networks.

http://arc.aiaa.org/doi/abs/10.2514/6.2013-4627

Autonomous collision avoidance system (ACAS) for Unmanned Aerial Vehicles (UAVs) is set as a tool to prove that they can achieve the equivalent level of safety, required in context of integrating UAVs flight into the National Airspace System (NAS). This paper focus on the cooperative avoidance part, aiming to define an algorithm that can provide avoidance between cooperative UAVs in general, while still be restricted by some common rules. The algorithm is named the Selective Velocity Obstacle (SVO) method, which is an extension of the Velocity Obstacle method. The algorithm gives guidelines for UAVs to select between three basic modes for avoidance, i.e., to Avoid, Maintain, or Restore. The variation of those three modes gives flexibility for UAVs to choose how will they avoid. By modeling the algorithm as a hybrid system, simulations on various UAVs encounters scenario were conducted and shows satisfying result. Monte Carlo simulations are then conducted to conclude the performance even more. Randomizing the initial parameters, including speed, attitude, positions and avoidance starting point, more than 106 encounter scenario were tested, involving up until five UAVs. A parameter called the Violation Probability is then derived, showing zero violations in the entire encounter samples.

A new non-conventional technique based on fuzzy deconvolution for scattering centre detection (F-SCD) is proposed together with its implementation in FPGA for real-time deployment in automotive collision avoidance application. F-SCD emulates the human being interpretation of radar images using fuzzy measurement of features of the radar Point Spread Function (PSF) differently from other classic detection techniques. The first stage of F-SCD detects signal from noise using image oversampling and binary integration technique. The second stage uses a fuzzy description of the radar PSF in order to discriminate among scatterers and side lobes in the radar image. The method/IP have been implemented in FPGA and tested with a FMCW 77 GHz radar prototype in real automotive benchmarks showing high POD, low FAR and high rejection of ambiguities even with poor time-space variant PSF.

We present and validate a framework for visual navigation with obstacle avoidance. The approach was originally designed in [1], but major improvements and real outdoor experiments are added here. Visual navigation consists of following a path, represented as an ordered set of key images, that have been acquired in a preliminary teaching phase. While following such path, the robot is able to avoid new obstacles which were not present during teaching, and which are sensed by a range scanner. We guarantee that collision avoidance and navigation are achieved simultaneously by actuating the camera pan angle, in the presence of obstacles, to maintain scene visibility as the robot circumnavigates the obstacle. The circumnavigation verse and the collision risk are estimated using a potential vector field derived from an occupancy grid. The framework can also deal with unavoidable obstacles, which make the robot decelerate and eventually stop.

History shows that ferry and RoPax collisions with tankers can be devastating for human life. This paper follows up such scenario to study rational increase of safety of tanker structures. Through the coupling of multi-objective structural optimization and crashworthiness analysis, a conventional tanker structure is optimized for higher collision tolerance, accounting also for the hull mass, so that the increase in safety is rational. Two new concepts are proposed for this task: a crash simulations approach utilizing coarse FE mesh to reduce the simulation’s calculation time and a two-stage optimization approach which seeks to minimize the number of collision simulations and thus reduce the overall optimization time. Combining the results with the state-of-the-art knowledge, new reasoning about crashworthiness of tanker structures is realized.

Analysis of vehicle deformation from impacts largely relies upon A and B stiffness coefficients for vehicle structures in order to approximate the velocity change and accelerations produced by an impact. While frontal impact stiffness factors for passenger vehicles, light trucks, vans, and sport utility vehicles are relatively prevalent for modern vehicles, stiffness factors for rear and side structures, as well as heavy vehicles, buses, recreational vehicles, trailers, motorcycles, and even objects, are essentially non-existent.
This paper presents the application of the Generalized Deformation and Total Velocity Change Analysis to real-world collision events (G-DaTADeltaV System of Equations) as developed by this author. The focus of this paper addresses the relative precision and accuracy of the G-DaTADeltaV System of Equations for determining the total velocity change for oblique and/or offset vehicle-to-vehicle collisions involving light trucks and sport utility vehicles, which are largely under-represented with modern vehicle A and B stiffness values for side and rear sur­faces. The previous paper presented by this author to the Academy addressed the relative accuracy and precision of the G-DaTADeltaV System of Equations as they relate to a first validation using the RICSAC-staged collision database1 As a secondary and more comprehensive validation process, the G-DaTADeltaV System of Equations will be applied to real-world collision data obtained through the National Automotive Sampling System (NASS), which provides the National Highway Traffic Safety Administration (NHTSA) with a comprehensive compilation of real-world collision events representing a broad-based collection of collision configurations from across the coun­try. This data represents a reusable source of information that was collected using standardized field techniques implemented by NASS-trained field technicians. Through using a "core set of crash data components," NASS has demonstrated its utility and applicability to a vast array of statistical and analytical studies regarding traffic safety and vehicle collision dynamics.

Transportation of dangerous cargo by ships demands ensuring an appropriate protection level in case of an accident. Two of the most dangerous accident types resulting in an oil spill are grounding and ship to ship collision. The article presents results of research on a new composite construction for a ship hull that increases structural safety during collision. The concept of semi elastic Steel Concrete Polymer structure is presented. The most important results of the research are discussed. The resistance to collision and grounding of the new sandwich solution is investigated by means of two independent large scale experiments. The first collision experiment was conducted on a reference ship section representative for a double bottom or a double side of a ship. The second test was performed on a construction similar to the first one, but with an additional semi elastic protective barrier. Both ship sections were loaded in the same manner by a sphere-shaped intender simulating the assumed collision scenario. On the basis of large scale collision experiments the new solution of semi elastic protective barrier was proven to significantly increase resistance to collision and grounding.

The use of Power Electric wheelchairs for physically disabled people is becoming increasingly common; however the systems used in order to manage it are reduced to the use of specialized drivers, only using joysticks to manipulate the wheelchair, and these are generally suitable only for patients with motor disabilities in their lower limbs. This paper proposes the development of a speech control system and a magnetic control system to drive a wheelchair as an alternative for patients with severe disabilities. Both are applied to a smart wheelchair; the proposed systems include the development of new communication protocols for the wheelchair through a microcontroller, an H-bridge, a microphone and a magnetic control.

http://arc.aiaa.org/doi/abs/10.2514/6.2015-0592

Autonomous systems are required in order to enable UAVs to conduct self-separation and collision avoidance, especially for flights within the civil airspace system. A method called the Velocity Obstacle Method can provide the necessary situational awareness for UAVs in a dynamic environment, and can help to generate a deconflicting maneuver. This paper focuses on the assessment of the Velocity Obstacle Method application and its ability to resolve various conflict situations in three dimensional space. This assessment results in a redefinition of the criteria of avoidance. A novel technique is introduced to support the avoidance decision, by representing the conflict situation in various avoidance-planes. Several new definitions to support the method are introduced. This method is then implemented in three-dimensional simulations for UAVs in cases of conflict, in which more than one option of resolution is provided.

A Vehicular Ad hoc Network is a model of mobile Ad hoc network for running vehicles in which vehicles can communicate with wireless communication devices within a confined geographical region in the absence of fixed infrastructure. Also, it works as an intelligent system for road safety and traffic management by disseminating information among moving vehicles with the wireless sensor Ad hoc network, which has so many challenges and prospects since routing protocols are semi-organized for vehicular movements and constraints of road topology with rules within the limited physical connectivity. A successful broadcast is a challenging task within limited resources. The availability of the number of nodes in a range is an obstacle for rebroadcasting, can be resolved by reducing the figure with a fuzzy logic-based forwarder selection method. Signal strength can be improved in a transmission range by employing an optimum number of broadcasting points with less delay, fewer hop counts, and a high packet dissemination ratio. This optimal fuzzy broadcast algorithm is proposed for the selection of the best forwarder nodes with the consideration of inter-vehicle: distance, mobility, and density into account. Also, the performance of inter-vehicular geographic routing in high speed dynamic vehicular environment is proved with the MATLAB software, OMNeT++ tool, and SUMO environment.

Abstract We propose an autonomous vehicle guidance framework which combines visual navigation with simultaneous obstacle avoidance. The method was originally designed in [1], but real outdoor experiments and major improvements have been added in this paper. ...

This paper analyzes and studies the effect of varied vehicle loads to the collision avoidance (CA) system performance. The design comprises of Artificial Potential Field as the risk assessment and motion planning strategy and Nonlinear Model Predictive Control (NMPC) as the automated motion guidance. The study is important to determine robust NMPC weighting parameters for the vehicle states in varied loads vehicle collision avoidance situations. Simulation of the proposed system was done and evaluated. The results showed that the varied loads of the host vehicle affect the vehicle states error penalization. The findings will be helpful for a real-time implementation of a multi-scenario highway CA system to provide a well-tuned avoidance actuation by NMPC. It is done by identifying the most mercurial vehicle dynamics states in all variations of vehicle loads during CA navigations.

Technological progress in the nineteenth century carried mankind’s experience of locomotion to dazzling new heights. Where before, riding horses, skating on ice, sailing the winds, or the simple act of running were the utmost one could achieve with regards to attaining high speeds, the advent of the Industrial Age provided a wide range of new possibilities. The invention of the steam engine, especially, enabled new developments in the relation between man and motion. So how did these developments alter this relation? How is this altered relation reflected in nineteenth century literature? And what are the philosophical, psychological and conceptual foundations for this relation?

Forward Collision Avoidance (FCA) systems in automobiles is an essential part of Advanced Driver Assistance System (ADAS) and autonomous vehicles. These devices currently use, radars as the main sensor. The increasing resolution of camera sensors, processing capability of hardware chipsets and advances in image processing algorithms, have been pushing the camera based features recently. Monocular cameras face the challenge of accurate scale estimation which limits it use as a stand-alone sensor for this application. This paper proposes an efficient system which can perform multi scale object detection which is being patent granted and efficient 3D reconstruction using structure from motion (SFM) framework. While the algorithms need to be accurate it also needs to operate real time in low cost embedded hardware. The focus of the paper is to discuss how the proposed algorithms are designed in such a way that it can be provide real time performance on low cost embedded CPU's which makes use of only Digital Signal processors (DSP) and vector processing cores.

Recent technological advances in cognitive radio based on software defined radio platforms have extended the capabilities of wireless communication systems. The unique ability to alter communication protocols to meet changing system demands makes cognitive radio suitable for wireless applications otherwise difficult to implement using conventional wireless terminals. One such wireless application is the rapid deployment of a heterogeneous wireless network for the purpose of establishing a communication infrastructure between an incompatible set of wireless consumer devices. This paper presents the design and implementation details of cognitive radio terminals used to build a rapidly deployable, frequency-agile, heterogeneous wireless network. Research findings assisted researchers in the development of a wireless transceiver capable of connecting a variety of wireless transceivers, detecting the presence of interference, and accordingly switching to available channels as allocated by its spectrum sensing capability. The focus of this paper is to propose an architecture for cognitive radio terminals that can be used to build rapidly deployable wireless network applications 1 .

Rear-end collision often leads to serious casualties and traffic congestion. The consequences are even worse for multiple-vehicle collision. Many previous works focused on collision warning and avoidance strategies of two consecutive vehicles based on onboard sensor detection only. This paper proposes a centralized control strategy for multiple vehicles to minimize the impact of multiple-vehicle collision based on vehicle-to-vehicle communication technique. The system is defined as a coupled group of vehicles with wireless communication capability and short following distances. The safety relationship can be represented as lower bound limit on deceleration of the first vehicle and upper bound on maximum deceleration of the last vehicle. The objective is to determine the desired deceleration for each vehicle such that the total impact energy is minimized at each time step. The impact energy is defined as the relative kinetic energy between a consecutive pair of vehicles (approaching ...

Near-miss accidents at intersections are still reported globally. This is due to the driver inattention during the scenario. Though many collision warning systems have been introduced, such systems are ineffective in alerting the drivers during the sudden emergence of unknown moving vehicles at intersections. This paper proposes an intersection collision avoidance (CA) architecture, composed of Artificial Potential Field (APF) as the motion planning and Nonlinear Model Predictive Control (NMPC) as the path tracking strategy. To aid the driver inattention, APF considers the risk of intersections and collision with the moving obstacles. NMPC is adopted into the architecture to guide the CA navigation. When the host vehicle approaches the intersection, the desired deceleration value is yielded. Once the occluded moving obstacle appears, APF provides the desired vehicle states for path replanning based on the collision risks. NMPC acts as automated motion guidance for the host vehicle to handle the abrupt increment of vehicle summation forces during the navigation. To evaluate the system efficacy, computational simulation is done and the performance is compared with the Time-to-Collision and Safe-Distance to ensure its reliability in preventing near-miss accidents. Results show that the consideration of intersection risk together with the risk of collision with the moving obstacle aids in preventing near-miss intersection collisions.