Large Crashes Research Articles

Crash severity analysis: A data-enhanced double layer stacking model using semantic understanding

The crash severity analysis is of significant importance in traffic crash prevention and emergency resource allocation. A range of innovations offers potential traffic crash severity prediction models to improve road safety. However, the semantic information inherent in traffic crash data, which is crucial in enabling a deeper understanding of its underlying factors and impacts, has yet to be fully utilized. Moreover, traffic crash data are commonly characterized by a small sample size, which leads to sample imbalance problem resulting in prediction performance decline. To tackle these problems, we propose a semantic understanding-based data-enhanced double-layer stacking model, named EnLKtreeGBDT, for crash severity prediction. Specifically, to fully leverage the inherent semantic information within traffic crash data and analyze the factors influencing crashes, we design a semantic enhancement module for multi-dimensional feature extraction. This module aims to enhance the understanding of crash semantics and improve prediction accuracy. Then we introduce a data enhancement module that utilizes data denoising and migration techniques to address the challenge of data imbalance, reducing the prediction model's dependence on large sample crash data. Furthermore, we construct a two-layer stacking model that combines multiple linear and nonlinear classifiers. This model is designed to augment the capability of learning linear and nonlinear mixed relationships, thereby improving the accuracy of predicting the severity of crashes on complex urban roads. Experiments on historical datasets of UK road safety crashes validate the effectiveness of the proposed model, and superior performance of prediction precision is achieved compared with the state-of-the-arts. The ablation experiments on both semantic and data enhancement modules further confirm the indispensability of each module in the proposed model.

Two-stage crash process in resistive drift ballooning mode driven ELM crash

We report a two-stage crash process in edge localized mode (ELM) driven by resistive drift-ballooning modes (RDBMs) numerically simulated in a full annular torus domain with a scale-separated four-field reduced MHD (RMHD) model using the BOUT++ framework. In the early nonlinear phase, the small first crash is triggered by linearly unstable RDBMs, and m/n=2/1 magnetic islands are nonlinearly excited by nonlinear coupling of RDBMs as well as their higher harmonics. Here, m is the poloidal mode number, n is the toroidal mode number, the q = 2 rational surface exists near the pressure gradient peak, and q is the safety factor. Simultaneously, middle-n RDBM turbulence develops but is poloidally localized around X-points of the magnetic islands, leading to the small energy loss. The second large crash occurs in the late nonlinear phase. Higher harmonics of m/n=2/1 magnetic islands well develop around the q = 2 surface via nonlinear coupling and make the magnetic field stochastic by magnetic island overlapping. Turbulence heat transport develops at X-points of higher harmonics of m/n=2/1 magnetic islands, resulting in the turbulence spreading in the poloidal direction. The large second crash is triggered when the turbulence covers the whole poloidal region so that the magnetic island generation and magnetic field stochastization before the large crash can be interpreted as ELM precursors. It is concluded that the ELM trigger is attributed to the turbulent spreading in the poloidal direction in synchronization with the magnetic field stochastization and the crash is driven by E × B convection rather than the conventional Rechester–Rosenbluth anomalous electron heat transport.

Investigating safety and cost-effectiveness of cable median barriers in Louisiana

Objectives Cable median barriers (CMBs) are installed on freeway medians to prevent cross-median crashes and reduce the severity of median-related crashes. Though CMBs are effective in preventing cross-median crashes, they are also known to increase the number of property damage–only (PDO) crashes. The higher frequency of PDO crashes could result in increased CMB maintenance and repair expenses. The aim of this study is to evaluate the safety impact and economic justification of CMBs in Louisiana. Methods Initially, a flowchart was developed using Louisiana crash data to identify targeted crashes, such as median-related and cross-median crashes. This was followed by a 3-year observational before-and-after crash analysis with an emphasis on head-on collisions and crashes involving large trucks. Using a 4-step improved prediction method, crash modification factors were then developed to quantitatively assess the impact of CMBs on crash outcomes, accounting for and adjusting to changes in the annual average daily traffic (AADT) and relevant crash frequencies before and after CMB implementation. Finally, an exhaustive benefit–cost analysis was conducted to determine the cost-effectiveness of CMBs. Results The results revealed that CMBs significantly reduced cross-median crashes of all severities. However, an increase in PDO crashes was observed in both total and median-related crashes. Large truck cross-median crashes and head-on collisions also decreased significantly after CMB implementation. Testing Level 4 (TL-4) CMBs were found to be more effective in preventing vehicles from crossing the median and in reducing crashes of higher severity levels. The benefit–cost ratios, calculated using economic crash unit costs for both total and targeted crashes, were greater than 1. Notably, the estimated benefit–cost ratios were considerably higher, demonstrating that CMBs are cost-effective countermeasures for enhancing traffic safety. Conclusion This study contributes to the understanding of CMB performance from both traffic safety and economic perspectives. The findings may assist transportation agencies in making decisions regarding the management of CMB systems. Based on the comprehensive analysis of CMBs on Louisiana freeways, this project has revealed that CMBs are an effective and economically justified crash countermeasure. Thus, further implementation of CMBs is recommended until better alternatives are available.

Observation of tearing mode triggering by sawtooth crash with high-Z impurity accumulation in EAST

The high-Z impurity accumulation in the plasma core could degrade the plasma performance and may affect sawtooth crash and trigger tearing modes (TMs) in experimental advanced superconducting tokamak (EAST). Triggering of m/n = 3/2 TM (m and n are the poloidal and toroidal mode numbers, respectively) and m/n = 2/1 TM by a large sawtooth crash is observed and studied under the condition of high-Z impurity accumulation in core plasma in EAST tokamak. This work mainly focuses on the relationship among impurity, sawtooth crashes, and TMs. When there is a certain amount of impurity in the plasma core, sawtooth crash leads to impurity transport, resulting the changes in impurity concentration, and by inducing the density perturbation near the rational surface, TMs become destabilized. The experimental analyses indicate that the outward impurity transport after a sawtooth crash plays an important role in the destabilization of TMs.

Large Bets and Stock Market Crashes

Abstract Some market crashes occur because of significant imbalances in demand and supply. Conventional models fail to explain the large magnitudes of price declines. We propose a unified structural framework for explaining crashes, based on the insights of market microstructure invariance. A proper adjustment for differences in business time across markets leads to predictions which are different from conventional wisdom and consistent with observed price changes during the 1987 market crash and the 2008 sales by Société Générale. Somewhat larger-than-predicted price drops during 1987 and 2010 flash crashes may have been exacerbated by too rapid selling. Somewhat smaller-than-predicted price decline during the 1929 crash may be due to slower selling and perhaps better resiliency of less integrated markets.

Crash topology optimization for front-end safety parts of battery electric vehicle using an improved equivalent static loads method

Crash topology optimization is a typical nonlinear dynamic response structural topology optimization problem, which is one of the most difficult problem in the structural design field. The equivalent static load method (ESLM) provides a well-defined pattern to solve such difficult problems, which can convert a nonlinear dynamic response optimization into multi-load steps optimization problem with the equivalent static loads (ESLs). However, due to the large deformation in the crash condition and nodal characteristics of the ESLs, it is hard to solve the crash topology optimization directly using the standard ESLM. To expand the application scope of the ESLM, an improved ESLs calculation method is proposed by using the model order reduction method and energy principle, which only acting on some nodes and can be scaled adaptively. Correspondingly, to enrich the connotation of topology optimization, a crash topology optimization method is proposed by using the improved ESLs, which can solve the numerical problems in the design domain by guaranteeing the topology optimization with the improved ESLs perform in the linear rang. First, the principle of the standard ESLM is introduced, and corresponding problems and deficiencies in solving the crash topology optimization are summarized. Then, to solve the above problems, the improved ESLs calculation method is proposed. Meanwhile, the corresponding crash topology strategy is proposed based on the improved ESLs. Finally, to verify the effectiveness and engineering application value of the proposed crash topology optimization method, a test-verified frontal crash simulation model of the BEV front-end is established, and their safety parts are redesigned by using the proposed method. The results show that, the proposed method can effectively solve the crash topology optimization of thin-walled structures under large deformation crash condition. This method also provides a new idea and practical method for the crashworthiness and lightweight design of automobile structures.

Integrated deep learning framework for unstable event identification and disruption prediction of tokamak plasmas

The ability to identify underlying disruption precursors is key to disruption avoidance. In this paper, we present an integrated deep learning (DL) based model that combines disruption prediction with the identification of several disruption precursors like rotating modes, locked modes, H-to-L back transitions and radiative collapses. The first part of our study demonstrates that the DL-based unstable event identifier trained on 160 manually labeled DIII-D shots can achieve, on average, 84% event identification rate of various frequent unstable events (like H-L back transition, locked mode, radiative collapse, rotating MHD mode, large sawtooth crash), and the trained identifier can be adapted to label unseen discharges, thus expanding the original manually labeled database. Based on these results, the integrated DL-based framework is developed using a combined database of manually labeled and automatically labeled DIII-D data, and it shows state-of-the-art (AUC = 0.940) disruption prediction and event identification abilities on DIII-D. Through cross-machine numerical disruption prediction studies using this new integrated model and leveraging the C-Mod, DIII-D, and EAST disruption warning databases, we demonstrate the improved cross-machine disruption prediction ability and extended warning time of the new model compared with a baseline predictor. In addition, the trained integrated model shows qualitatively good cross-machine event identification ability. Given a labeled dataset, the strategy presented in this paper, i.e. one that combines a disruption predictor with an event identifier module, can be applied to upgrade any neural network based disruption predictor. The results presented here inform possible development strategies of machine learning based disruption avoidance algorithms for future tokamaks and highlight the importance of building comprehensive databases with unstable event information on current machines.

Factors influencing fatal vehicle-involved crash consequence metrics at spatio-temporal hotspots in South Korea: application of GIS and machine learning techniques

ABSTRACT Studies have employed several techniques to identify the effect of individual risk factors influencing crashes at hotspot locations. Nevertheless, as crashes are sometimes influenced by a combination of risk factors, identifying the chains of factors collectively contributing to fatal crashes at hotspot locations could provide added insights for improving traffic safety. By employing fatal crash data from Korea, this study identifies hotspots with increasing (critical) and decreasing (diminishing) temporal trends using a spatio-temporal hotspot analysis tool in GIS. Further, a machine learning technique is employed to explore the chains of factors influencing the number of vehicles and the number of casualties involved in fatal crashes at intersections and midblocks in each hotspot type identified. In general, results showed that minibuses/vans and construction vehicles were mainly at fault for fatal single-vehicle pedestrian-involved crashes. While many casualties and vehicles are likely to be involved in crashes at midblocks during the daytime regardless of the hotspot type, the nighttime variable was particularly associated with large casualty-size crashes at critical intersection hotspots. Further, while reckless driving was mostly associated with single-vehicle crashes at intersections in diminishing hotspots, pedestrian protection, and improper centreline crossing violations were more pronounced at midblocks in diminishing hotspots. This analysis identified groups of factors that could be collectively controlled to improve road safety and proposed countermeasures to mitigate fatal crashes on roadways.

Large truck fatal crash severity segmentation and analysis incorporating all parties involved: A Bayesian network approach

Studies on fatal crashes involving large trucks are typically focused on the examination of injury severity at the personal level rather than the crash level. In this study, instead of adopting the predefined, commonly used rule of transformation of individual injury severity to crash level severity, we develop a data specific transformation approach. Crash-level severity for fatal large-truck involved crashes is first identified applying a two-step clustering method. A Bayesian network is then constructed to investigate the impacts of influential factors on the crash severity. Data obtained from the Fatality Analysis Reporting System in 2019, are used to calibrate the model. The results indicate that the number of vehicles involved in the crash, collision manner, truck-airbag deployment, overturn of vehicles, driver’s speeding offense, and gross vehicle weight are the key determinants of crash injury severity. Uni- and two-dimensional inference analyses were conducted to determine the relations among these factors and their association with crash severity.

Injury Severity Analysis for Large Truck-Involved Crashes: Accounting for Heterogeneity

This study explores the crash injury severity of large truck-involved crashes, where the truck driver was identified as the at-fault driver. The paper focuses on vehicle-in-motion crashes that occurred on Florida’s state highways between 2007 and 2016. A random parameter ordered logit (RPOL) model was developed to identify random parameters and interaction effects. Results indicated that not using restraint systems, running a red light, wrong-way driving, failing to yield the right of way, tire or brake defects, and dark conditions had positive associations with higher levels of crash injury severity. The random variables—straight alignment, paved shoulders, and unpaved shoulders—showed significant random effects among the observations. For straight alignment, running red lights, following too closely, vision obstruction caused by fixed objects, and vision obstruction caused by fog were the sources of heterogeneity. Unpaved shoulders, running red lights, wrong-way driving, and the presence of parked or stopped vehicles were found as interaction effects. Results showed that accounting for heterogeneity and interaction effects significantly improved the goodness of fit of the model. This study provides more comprehensive knowledge of the influencing factors of large truck crashes by considering the role of heterogeneity and its potential sources in crash injury severity.

A proposed approach for the evaluation of consequences of a large aircraft crash accident at an RBMK type reactor site during decommissioning

Nuclear power plants with RBMK reactors were designed in late sixties - early seventies and were built entirely in the European part of the Russian Federation, the Lithuanian Republic and the Ukrainian Republic. The plants in the Lithuanian and the Ukrainian Republics are in various stages of decommissioning. As currently planned, the most of RBMK reactors still operating in the Russian Federation will end their operation in this decade.The reactor building, spent nuclear fuel and radioactive waste storage facilities, built at the time of construction of RBMK plants are not specially designed to withstand the direct impact of an aircraft crash. There were no such regulatory requirements at the time the plants were designed and, no such requirements were introduced by regulatory bodies later on. But for now, in the aftermath of the September 11, 2001 terrorist attack on the United States, the event of an aircraft crash on a nuclear facility can no longer be considered as extremely unlikely.This paper considers the state of a nuclear power plant with the RBMK type reactor(s) during certain decommissioning and dismantling stages and discusses the approach for the evaluation of radiological consequences to the public in the event of a large aircraft crash on potentially hazardous objects of the plant. The damage to spent nuclear storage facilities, the reactor graphite stack and storage facilities for the combustible waste may cause a significant release of radionuclides to the environment. Key components of the approach include the evaluation of a structural damage to the facilities, the evaluation of a fire propagation and impact, the evaluation of a spent nuclear fuel cooling capability in storage pools, the evaluation of an airborne radionuclide release, the evaluation of a radionuclide dispersion in the environment and the evaluation of the public exposure. A set of widely used software and methods for the performance of the specific tasks is discussed. The paper also describes how the proposed approach was applied for the evaluation of radiological consequences from a Boeing 767-400 type aircraft crash at the Ignalina NPP site where a plant with two RBMK-1500 type reactors is decommissioned.The paper does not address modern radioactive waste and spent nuclear fuel management and storage facilities that might be built at RBMK sites as well. As a rule, assessment and mitigation of an aircraft crash accident consequences is an integral part of design of these new facilities.

Liquidity Shocks, Price Volatilities, and Risk-managed Strategy: Evidence from Bitcoin and Beyond

This study examines the prediction power of market liquidity (the ease with which asset are traded) and funding liquidity (the ease with which traders can obtain funding) on the price volatilities of Bitcoin. We find that both market and funding liquidity shocks forecast future volatility. More importantly, liquidity shocks have a stronger and persistent effect on the long-term trend component of volatility. Exploiting the predictability of liquidity shocks, we propose a risk-managed strategy to manage extremely high volatility and avoid occasional large crashes in cryptocurrency markets. This novel strategy virtually eliminates crashes and improves the Sharpe ratio substantially against the benchmark buy-and-hold strategy. The outperformance is much stronger during the turbulent periods of cryptocurrencies. Hence, this paper provides important insights into cryptocurrency investment and portfolio management combining traditional assets and cryptocurrencies.

Asset price volatility and investment horizons: An experimental investigation

We study the effects of the investment horizon on asset price volatility using a Learning to Forecast laboratory experiment. We find that, for short investment horizons, participants coordinate on self-fulfilling trend-extrapolating predictions. Price deviations are then reinforced and amplified, possibly leading to large bubbles and crashes in asset prices. For longer investment horizons such bubbles do not emerge and price volatility tends to be lower. This is due to the fact that, for longer horizons, there is more dispersion in participants’ forecasts, and participants extrapolate trends in past prices to a lesser extent. We also show that, independent of the investment horizon, if the initial history of asset prices is already relatively stable before participants start their prediction task, price volatility remains small, with prices close to their fundamental values for the duration of the experiment.

A correlated random parameters approach to investigate large truck rollover crashes on mountainous interstates

Rollover risk on mountainous interstates is a major concern for transportation agencies due to the combined mixed effects of adverse weather conditions and complex topography. Such crashes incur hazardous consequences on road users’ lives. Therefore, a correlated random parameters logit modeling framework was employed to investigate the influences of crash precursors on rollover risk to identify effective safety countermeasures. This approach was selected to account for both the crash contributing factors’ unobserved heterogeneity effects and the correlations among those effects. The data, used in this study, were those of single-truck crashes on Wyoming’s interstate curved sections. The traditional logit and uncorrelated random parameters, or mixed, logit models were attempted as well. With that, the analysis results indicated that the fit of the correlated random parameters logit model was superior to those of the others. It also revealed insights regarding correlations among random parameters that were obscure in the other models. According to its results, on average, veering off the road, overcorrections and severe winds raised the risk of single-truck rollover crashes. On the other hand, median barriers, roadside guardrails, tight horizontal curves, icy road surfaces, wet surfaces and surfaces covered by loose material, in general, reduced the likelihood of rollovers. Correlations, such as those between severe winds and icy surfaces and those between roadside guardrails and icy surfaces, were inferred as well. This study’s results will assist transportation officials in efficiently identifying appropriate countermeasures to mitigate the impact of truck rollovers particularly during adverse weather conditions.

Analysis of Severe Injuries in Crashes Involving Large Trucks Using K-Prototypes Clustering-Based GBDT Model

The unobserved heterogeneity in traffic crash data hides certain relationships between the contributory factors and injury severity. The literature has been limited in exploring different types of clustering methods for the analysis of the injury severity in crashes involving large trucks. Additionally, the variability of data type in traffic crash data has rarely been addressed. This study explored the application of the k-prototypes clustering method to countermeasure the unobserved heterogeneity in large truck-involved crashes that had occurred in the United States between the period of 2016 to 2019. The study segmented the entire dataset (EDS) into three homogeneous clusters. Four gradient boosted decision trees (GBDT) models were developed on the EDS and individual clusters to predict the injury severity in crashes involving large trucks. The list of input features included crash characteristics, truck characteristics, roadway attributes, time and location of the crash, and environmental factors. Each cluster-based GBDT model was compared with the EDS-based model. Two of the three cluster-based models showed significant improvement in their predicting performances. Additionally, feature analysis using the SHAP (Shapley additive explanations) method identified few new important features in each cluster and showed that some features have a different degree of effects on severe injuries in the individual clusters. The current study concluded that the k-prototypes clustering-based GBDT model is a promising approach to reveal hidden insights, which can be used to improve safety measures, roadway conditions and policies for the prevention of severe injuries in crashes involving large trucks.

Numerical simulations of base-isolated LNG storage tanks subjected to large commercial aircraft crash

LNG storage tanks are specified to be designed to withstand the impact of large commercial aircraft, while the existing studies refer to the safety assessment of the base-isolated LNG storage tank are quite limited. At present firstly, the detailed finite element (FE) models of LNG storage tanks are established. Based on the scaled tests of RC slabs under the impact of deformable and non-deformable projectiles, the numerical algorithm, constitutive model and the corresponding parameters of concrete are validated, which can both reproduce the behaviors of bending and punching shear failure potential induced by the aircraft impact. Then, the energy dissipation capacities and relative constitutive model of isolation systems are evaluated and validated. The reliable FE models of base-isolated LNG storage tanks with three typical isolation systems are established Furthermore, by adopting the missile–target coupled method, the dynamic responses and damage failure of tanks under aircraft impact are assessed, it derives that, (i) the fuselage and engines impact regions of tank wall are locally perforated and penetrated, the connection between the inner steel wall and deck is torn, and the safety of tanks is hard to be guaranteed; (ii) the selection isolation systems have sufficient lateral stiffness to resist the excessive global deformation of the base-isolated LNG storage tanks under aircraft impact, different types of isolation systems and boundary conditions of tanks have slight influence on the local damage of tank wall. The present work could provide beneficial guidances for the design and safety assessment of the base-isolated large-scale LNG storage tanks under aircraft collisions.

Investigating factors affecting severity of large truck-involved crashes: Comparison of the SVM and random parameter logit model

Introduction: Reducing the severity of crashes is a top priority for safety researchers due to its impact on saving human lives. Because of safety concerns posed by large trucks and the high rate of fatal large truck-involved crashes, an exploration into large truck-involved crashes could help determine factors that are influential in crash severity. The current study focuses on large truck-involved crashes to predict influencing factors on crash injury severity. Method: Two techniques have been utilized: Random Parameter Binary Logit (RPBL) and Support Vector Machine (SVM). Models have been developed to estimate: (1) multivehicle (MV) truck-involved crashes, in which large truck drivers are at fault, (2) MV track-involved crashes, in which large truck drivers are not at fault and (3) and single-vehicle (SV) large truck crashes. Results: Fatigue and deviation to the left were found as the most important contributing factors that lead to fatal crashes when the large truck-driver is at fault. Outcomes show that there are differences among significant factors between RPBL and SVM. For instance, unsafe lane-changing was significant in all three categories in RPBL, but only SV large truck crashes in SVM. Conclusions: The outcomes showed the importance of the complementary approaches to incorporate both parametric RPBL and non-parametric SVM to identify the main contributing factors affecting the severity of large truck-involved crashes. Also, the results highlighted the importance of categorization based on the at-fault party. Practical Applications: Unrealistic schedules and expectations of trucking companies can cause excessive stress for the large truck drivers, which could leads to further neglect of their fatigue. Enacting and enforcing comprehensive regulations regarding large truck drivers’ working schedules and direct and constant surveillance by authorities would significantly decrease large truck-involved crashes.

Recursive Bivariate Probit Analysis of Injury Severity and Non-Truck Improper Actions in Large Truck-Related Crashes on Florida Suburban Roads

This study investigated the hierarchical connection among injury severity, non-truck improper actions, and contributing factors in large-truck-involved crashes. Data for 4 years (2011–2014) of crashes that involved a large truck (≥ 10,000 lb) and a non-truck vehicle were collected from suburban roads in Florida, U.S. A recursive bivariate probit model was fitted with collected data to identify the cause-effect chain, including contributing factors influenced by improper actions, the effects of improper actions on injury severity, and contributing factors indirectly affecting injury severity in large-truck-related crashes. Study results indicate that non-truck vehicle improper actions, such as excessive speed, careless driving, failure to yield right-of-way, and others, significantly increase the likelihood of fatal and severe injury in large-truck crashes, and factors such as crash month, darkness, intersection-related, surface and shoulder width, truck parking, truck driver age, non-truck driver age, and non-truck alcohol/drug impaired indirectly influence injury severity through their impacts on non-truck improper actions. Two factors—truck right-turn and non-truck driver physical defects—affect injury severity and non-truck improper actions simultaneously. Other factors, including crash year, annual average daily traffic, speed limit, crash type, truck type, truck speed, truck alcohol/drug-impaired, and motorcycle involvement, directly contribute to injury severity in large-truck crashes and have no influence on non-truck improper actions.

Observations on the Relationship between Crash Frequency and Traffic Flow

This work analyzes the relationship between crash frequency N (crashes per hour) and exposure Q (cars per hour) on the macroscopic level of a whole city. As exposure, the traffic flow is used here. Therefore, it analyzes a large crash database of the city of Berlin, Germany, together with a novel traffic flow database. Both data display a strong weekly pattern, and, if taken together, show that the relationship N(Q) is not a linear one. When Q is small, N grows like a second-order polynomial, while at large Q there is a tendency towards saturation, leading to an S-shaped relationship. Although visible in all data from all crashes, the data for the severe crashes display a less prominent saturation. As a by-product, the analysis performed here also demonstrates that the crash frequencies follow a negative binomial distribution, where both parameters of the distribution depend on the hour of the week, and, presumably, on the traffic state in this hour. The work presented in this paper aims at giving the reader a better understanding on how crash rates depend on exposure.

Analyzing injury severity of rear-end crashes involving large trucks using a mixed logit model: A case study in North Carolina

As one of the most frequently occurring crashes, rear-end crashes often result in injuries and property damage, especially when large trucks are involved. To investigate the contributing factors and the unobserved heterogeneity in such factors, a mixed logit model is developed to analyze rear-end crashes involving large trucks. A dataset containing 7,976 rear-end crashes involving large trucks is collected from Highway Safety Information System (HSIS) in North Carolina between 2005 and 2013. Driver, roadway, and environmental related characteristics are considered in the analysis. Speed limit over 50 mph is found to be better modeled as a random-parameter at specific injury severity levels. Results also indicate that driving under the influence of alcohol or drugs, rural roadways, dark light condition, grade roadway configuration, speed limit over 50 mph will significantly increase the injury severity of large truck involved rear-end crashes. Roadway with traffic control will significantly decrease the injury severity of such crashes. The findings in this study can greatly help traffic agencies and truck companies develop better large truck-involved rear-end crash prevention strategies.