The effective management of forces within heavy vehicles is essential for achieving desired performance outcomes. In this study, an auto-generated Model Predictive Control Allocation (MPCA) algorithm is presented. The controller is designed to distribute forces among individual actuators in a vehicle, focusing primarily on longitudinal forces while exploring lateral force dynamics. The approach integrates models of the actuators with vehicle dynamics, encompassing both point mass and dynamic vehicle models, within the controller framework. Through simulation, proof of the MPC's superior performance in reference tracking could be demonstrated, especially in comparison with baseline simulations employing force ratio split (FRS) and equal split (ES) distribution methods. Furthermore, findings show that it was possible to achieve a more energy efficient force distribution using the MPCs.

Some of the results of an extensive track testing program at the Center for Vehicle Systems and Safety (CVeSS) at Virginia Tech for evaluating the roll stability of commercial trucks with 33-ft A-double trailers are evaluated. The study includes straight-rail trailers with heavy and light loading conditions. Commercial trucks are more susceptible to rollovers than passenger cars because of their higher center of gravity relative to their track width. Multi-trailer articulated heavy vehicles, such as A-doubles, are particularly prone to rollovers because of their articulation and rearward amplification. Electronic stability control (ESC) has been mandated by the National Highway Safety Administration (NHSTA) for Class 8 trucks and busses since 2017. When detecting oversteer or understeer, ESC automatically activates the brakes at the correct side of the steer and/or drive axle(s) to regain steering stability. ESC, however, often cannot sense the likelihood of trailer rollover in multi-trailer articulated heavy vehicles because of the articulation between the trailers and tractors. As a result of this, trailers are often equipped with roll stability control (RSC) systems to mitigate speed-induced rollovers. Sensing the trailer lateral acceleration, RSC activates the trailer brakes to reduce speed and lower the likelihood of rollover. However, a limited number of past studies have shown that the trailer roll angle may provide an earlier indication of a pending rollover than the lateral acceleration. This study intends to provide further analysis in this regard in an effort to improve the effectiveness of RSC systems for trailers. An extensive amount of data from track testing with a 33-ft A-double under heavy and light loading is evaluated. Particular attention is given to lateral accelerations and trailer roll angles prior to rollover and relative to RSC activation time. The study's results indicate that the trailer roll angle provides a slightly earlier indication of rollover than lateral acceleration during dynamic driving conditions, potentially resulting in a timelier activation of RSC. Of course, detecting the roll angle is often more challenging than lateral acceleration, which can be detected with an accelerometer. Additionally, the roll angle measurement may be subjected to errors and possibly unwanted RSC engagement. The study's results further indicate that the trailer-based RSC systems effectively mitigate rollovers in both quasi-steady-state and dynamic driving conditions. / Master of Science / Some of the results of an extensive track testing program at the Center for Vehicle Systems and Safety (CVeSS) at Virginia Tech for evaluating the roll stability of commercial trucks with 33-ft A-double trailers are evaluated. "33-ft A-doubles" commonly refer to a commercial truck that has a tractor with two trailers (in this case 33-ft in length) that are connected by an A-dolly. Their modularity and ease of connecting and disconnecting at various drop stations have made such commercial vehicles a common scene on U.S. highways due to the proliferation of e-commerce cargo. Compared to a single-unit or tractor semi-trailer combination, the double- or triple-trailer configurations offer several logistical benefits that make them more advantageous. The multi-trailer vehicles can carry more cargo per driver, lowering driver, fuel, and equipment costs significantly. There are, however, some challenges to operating multi-trailer articulated vehicles. On average, their accidents are more expensive than single-trailer or single-unit trucks. Additionally, they are more susceptible to rolling over and causing property damage, injuries, and at times fatalities. To reduce rollovers, systems with automated braking, called roll stability control (RSC), are often installed on the trailers. RSC applies the trailer brakes if it senses that the vehicle speed — the primary cause of most commercial vehicle accidents — exceeds the safe limit for negotiating a turn. In this study, we intend to evaluate the effectiveness of roll stability control (RSC) systems for reducing the likelihood of speed-induced rollovers. We will also explore ways of improving their performance. Namely, we will evaluate whether sensing the lateral acceleration of the trailer or its roll angle would provide a better means for timely activation of RSC. The study's results indicate that, although more challenging to measure, the trailer roll angle provides a slightly sooner indication of a pending rollover than lateral acceleration. The results also suggest that RSC systems vastly reduce the number of speed-induced rollovers in trucks with 33-ft A-double trailers under different trailer configurations and cargo weights.

The major purpose of this thesis is to offer a detailed look at the development of two models used to assist in the detailed study of Australian two lane two way highways with particular reference to heavy vehicles. The first model governs the acceleration behaviour of vehicles on upgrades and downgrades. The second model controls overtaking manoeuvres on two lane two way highways where movement into the lane of oncoming traffic is required. Both models are implemented through a suite of transport simulation modelling software called Paramics.

Railway level crossing collisions in Australia are a major cause of concern for both rail and road authorities. Despite the fact that the number of railway crash fatalities in Australia has fallen in recent years, level crossing collisions constitute a significant proportion of the national rail toll. Although rail transport is presently one of the safest forms of land transport, collisions at level crossings are three times more likely to involve fatalities as compared to all other types of road crashes (Afxentis, 1994). With many level crossing fatalities and injuries resulting in coronial inquests, litigation and negative media publicity, the actions of rail and road infrastructure providers and the behaviour of motorists, pedestrians and rail users, come under close scrutiny. Historically, research in this area has been plagued by the rail/road interface and the separation of responsibilities between rail and road authorities reflecting the social and political context in which they are contained. With the recent rail reform in Australia, safety at level crossings has become a key priority area. Accordingly, there is a need to better understand the scope and nature of motorist behaviour at level crossings, in order to develop and implement more effective countermeasures for unsafe driving behaviour. However, a number of obstacles have hindered research into the area of level crossing safety. As with many road crashes, the contributing causes and factors are often difficult to determine, however a recent investigation of fatal collisions at level crossings supports the notion that human fault is a major contributor (Australian Transport Safety Bureau, 2002a). Additionally, there is a lack of reliable data available relating to the behavioural characteristics and perceptions of drivers at level crossings. Studies that do exist have lacked a strong theoretical base to guide the interpretation of results. Due to the lack of financial viability of continuing to approach risk management from an engineering perspective, the merits of human factor research need to be examined for suitability. In Australia, there has been considerable recognition regarding the importance of human factor approaches to level crossing safety (Australian Transport Council, 2003). However, little attempt has been made by authorities to scientifically develop and measure the effectiveness of road safety educational interventions. Therefore, there exists a significant need for developing targeted road safety educational interventions to improve current risk management solutions at level crossings. This research program is the first of its kind in investigating motorist behaviour at level crossings and the measuring the effectiveness of educational interventions for improving driving safety. Although other ‘educational’ campaigns exist in this field, no campaign or intervention has been guided by empirical research or theory. This thesis adopted a multidisciplinary approach to theory, reviewing perspectives from psychology, sociology and public health to explain driver behaviour at level crossings. This array of perspectives is necessary due to the variety of behaviours involved in collisions and near-misses at level crossings. The motivation underlying motorist behaviour determines to a large extent how successful behaviour change strategies (e.g. educational interventions) may be. Fishbein’s Integrated Model of Behaviour Change (IM) based largely on the health belief model, theory of reasoned action and theory of planned behaviour (Fishbein, 2000), assisted in the planning and development of a ‘oneoff’ targeted educational intervention specific for three different road user groups and in questionnaire development to ascertain the present context of motorist behaviour at level crossings. As no known research has been conducted that utilizes any psychosocial model to explain or predict level crossing behavior within different road user groups, this research program used this model as an exploratory tool rather than a tool to asses the model’s capacity in explaining such behaviour. The difference between this model and others is the inclusion of two important constructs in driving: skills (or abilities) and environmental factors. Fishbein (2003) suggests that the model recognises the lack of skills (or abilities) and/or environmental constraints may prevent a person from acting on their intentions, in light of the fact that intention is viewed as the primary determinant of behaviour. While the majority of behaviour change theories are limited by a range of conceptual and contextual factors (Parker, 2004), the IM was used to assist this research program as it appeared to be the most applicable model to examining level crossing safety. A variety of data collection methods were used in this research program as much of what is currently known about level crossing collisions is derived from coroner’s findings and statistics. The first study (Study One) was designed to extend this knowledge by undertaking a more thorough examination of contributing factors to level crossing crashes and the road user groups at risk. This study used the method of ‘triangulation’ (i.e. combining research methods to give a range of perspectives) whereby both qualitative (focus groups) and quantitative (modified Delphi technique) research designs were utilised (Barbour, 1999, Bryman, 1992). With the discipline of road safety research requiring methodological strategies that will enhance efforts to conceptualise the multi-faceted nature of motorist behaviour at level crossings, this application provided the robustness required. Results from the Delphi technique indicated that older, younger and heavy vehicle drivers are considered to be three of the highest risk road user groups by experts in the field. For the older driver group, experts agreed that errors in judgment were the most important issue for this group when driving at level crossings. Risk taking by younger drivers, such as trying to beat the train across the crossing, was viewed as the central issue for the younger driver group. Like the younger driver group, a concern by experts with the heavy vehicle group was intentional risk taking at level crossings. However, experts also rated the length of heavy vehicles a major concern due to the possibility of a truck over-hanging a crossing. Results from focus groups with train drivers in Study One indicated that there are unique problems associated with crossings in rural/regional areas compared to urban areas. The metropolitan train drivers generally experienced motorist behaviour at active crossings with flashing lights and boom gates while the regional train drivers experienced behaviours at active crossings with boom gates, crossings with lights only and passive crossings with stationary signs. In the metropolitan train driver group, experiences of motorist behaviour at level crossings included: motorists driving around boom gates, getting stuck under boom gates, queuing over congested crossings and driving through the crossing after the red lights commence flashing. The behaviour of motorists driving around boom gates was noted to occur quite regularly. The majority of metropolitan train drivers reported that it was a common occurrence for motorists to drive through a crossing when the lights are flashing both before and after the booms were activated and some crossings were named as ‘black spots’ (locations where motorists repeatedly violate the road rules). Vehicles protruding into the path of the train and motorists entering congested crossings and then panicking and driving backwards into the boom gates were also mentioned. Regional train drivers indicated that motorists not stopping or giving way to trains is a continual problem at passively controlled crossings (i.e. no boom gates or flashing lights). Regional train drivers generally agreed that the majority of motorists obey protection systems; however some motorists drive through flashing lights or drive around boom gates. Other high risk behaviours included motorists attempting to beat the train across the crossing, speeding up to go through flashing lights, and general risk taking by younger drivers in particular. Motorists not allowing enough time to cross in front of the train or hesitating (stopstarting) at crossings were also noted to be at high risk. There was a general perception by regional train drivers that motorists are unable to judge the speed and distance of an approaching train to determine a safe gap during which to cross. Local motorists were also reported to be a problem at level crossings for regional train drivers. A theme common to regional and metropolitan train drivers was the risk of catastrophic consequence associated with level crossing collisions. The reasons given for this were the threat of derailment, serious property damage, the high risk of a fatality, personal injury and, most earnestly, the potential for enduring psychological consequences. Drivers uniformly spoke about the continual fear they had of being involved in a collision with a heavy vehicle, and many spoke of the effects that such collisions had on train drivers involved. For this reason, train drivers were said to consider any near-miss incident involving trucks particularly serious. The second study undertaken as part of this research program (Study Two), involved formative research as part of the planning, development and delivery of behavioural interventions for each of the three road user groups identified in Study One. This study also used both qualitative and quantitative data collection methods to provide methodological triangulation and ensure reliability of the data. The overall objective of the qualitative data collection was to obtain rich data using a qualitative mode of inquiry, based on the key variables of attitudes, norms, self-efficacy (perceived behavioural control), perceived risk, environmental constraints and the skills/abilities of drivers. The overall objective of the quantitative data collection was to prioritise the issues identified in order to direct and allocate project resources for intervention planning, development and delivery. This combined recruitment strategy was adopted as it was an appropriate and practical data collection strategy within the qualitative and exploration methodology. Information obtained from each of the groups was critical in assisting, guiding, and identifying priority areas for message and material development. The use of focus groups and one-on-one interviews provided insights into why drivers think or do what they do at level crossings. The qualitative component of this study found that for the older driver group, regional drivers hold a greater perception of risk at level crossings than urban older drivers, with many recalling near-misses. Participants from the urban older driver group indicated that level crossings are not as dangerous as other aspects of driving, with many participants being doubtful that motorists are killed while driving at level crossings. Both urban and regional younger drivers tended to hold a low perception of risk for driving at level crossings, however many participants reported having great difficulty in judging the distance a train is from a crossing. Impatience for waiting at level crossings was reported to be the major reason for any risk taking at level crossings in the younger driver group. Complacency and distraction were viewed by heavy vehicle participants as two of the major driver factors that put them at risk at level crossings, while short-stacking (when the trailer of the truck extends onto the crossing), angle of approach (acute or obtuse angle) and lack of advance warning systems were seen as the major engineering problems for driving a truck at level crossings. The quantitative component of this study involving research with train drivers found that at the aggregate train driver level, it is apparent that train drivers consider motorists’ deliberate violations of the road rules and negligently lax approach to hazard detection as the predominant causes of dangerous driving at level crossings. Experts were observed to rank risk taking behaviours slightly lower than train drivers, although they agreed with train drivers that ‘trying to beat the train’ is the single most critical risk taking behaviour observed by motorists. The third study (Study Three) involved three parts. The aim of Part One of this study was to develop targeted interventions specific to each of the three road user groups by using Fishbein’s theoretical model (Integrated Model of Behaviour Change) as a guide. The development of interventions was originally seen as being outside of the scope of this project, however it became intertwined in questionnaire development and thus deemed to be within the realms of the current mode of inquiry. The interventions were designed in the format of a pilot radio road safety advertisement, as this medium was found to be one of the most acceptable to each of the road user groups as identified in the formative research undertaken in Study Two. The interventions were used as a ‘one-off’ awareness raising intervention for each road user group. Part Two involved the investigation of the present context of unsafe driving behaviour at level crossings. This second part involved the examination of the present context of motorist behaviour at level crossings using key constructs from Fishbein’s Integrated Model of Behaviour Change (IM). Part Three involved trialing a pilot road safety radio advertisement using an intervention and control methodology. This part investigated the changes in pre and post-test constructs including intentions, self-reported behaviour, attitudes, norms, selfefficacy/ perceived behaviour control, perceived risks, environment constraints and skills/ability. Results from this third study indicated that younger drivers recognise that level crossings are potentially a highly dangerous intersection yet are still likely to engage in risk taking behaviours. Additionally, their low levels of self-efficacy in driving at level crossings pose challenges for developing interventions with this age group. For the older driver sample, this research confirms the high prevalence of functional impairments such as increasing trouble adjusting to glare and night-time driving, restricted range of motion to their neck and substantial declines in their hearing. While factors contributing to the over-representation of older drivers in collisions at level crossings are likely to be complex and multi-faceted, such functional impairments are expected to play a critical role. The majority of heavy vehicle drivers reported driving safely and intending to drive safely in the future, however, there is a sub-set of drivers that indicate they have in the past and will in the future take risks when traversing crossings. Although this sub-set is relatively small, if generalised to the larger trucking industry it could be problematic for the rail sector and greater public alike. Familiarity was a common factor that was found to play a role in driving intention at level crossings for all three road user groups. This finding supports previous research conducted by Wigglesworth during the 1970’s in Australia (Wigglesworth, 1979). Taken together, the results of the three studies in this research program have a number of implications for level crossing safety in Australia. Although the ultimate goal to improve level crossing safety for all motorists would be to have a combination of engineering, education and enforcement countermeasures, the small number of fatalities in comparison to the national road toll limits this. It must be noted though that the likelihood of creating behavioural change would be increased if risk taking at level crossings by all motorists was detected and penalised, or alternatively, if perceptions of such detection were increased. The instilling of fear in drivers with the threat of punishment via some form of sanction can only be achieved through a combination of a mass media campaign and increasing police presence. Ideally, the aim would be to combine fear of punishment with the guilt associated with the social non-acceptability of disobeying road rules at level crossings. Such findings have direct implications for improving the present context of motorist behaviour at level crossings throughout Australia.

The major purpose of this thesis is to offer a detailed look at the development of two models used to assist in the detailed study of Australian two lane two way highways with particular reference to heavy vehicles. The first model governs the acceleration behaviour of vehicles on upgrades and downgrades. The second model controls overtaking manoeuvres on two lane two way highways where movement into the lane of oncoming traffic is required. Both models are implemented through a suite of transport simulation modelling software called Paramics.

Näringslivet efterfrågar ett införande av 74 tons lastbilar på det statliga vägnätet. Det finns ett flertal fördelar med att höja maximala bruttovikten från nuvarande 64 till 74 ton. Däremot är höga bruttovikter direkt kopplade till bärighetsrelaterade skador. Syftet med denna studie är att ligga till grund för vidare arbete av Trafikverket. Detta för att uppnå bättre kontroll på hela det statliga vägnätet och den tunga trafikens inverkan. Studien har utgått från observationer i PMSv3, ett webbaserat system med grafiska tvärprofiler baserade på vägytemätning samt information om det statliga belagda vägnätet. Observationerna innefattar mönsteridentifiering av återkommande tvärprofilstyper tillskrivna skador av den tunga trafiken. Resultatet av tvärprofilstyperna har sedan analyserats och beskrivits som representativa spårtyper för det statliga vägnätet. Slutsatsen är att det finns indikation på mönster av spårtyper. Utifrån observationerna i PMSv3 har spårtyperna formulerats som hypoteser för vidare forskning. / The wood industry asks for an imposition of 74-ton trucks on the national road system. There are several advantages of increasing the maximum gross weight from 64 to 74 tons. On the other hand, high gross weights are directly connected to structural road damage. The purpose of this study is to serve as foundation for further research by the Swedish transport administration. The ulterior goal is to reach better control on the national road system. The basis of the study is a web-based system - PMSv3, which contain road information and transverse road profiles. The observations in PMSv3 have led to rut-types ascribed due to heavy gross weights. The conclusion is that rut-patterns indications exist. Based on the observations in PMSv3, the rut-types have been formulated as hypothesis for further research.

In this thesis project, brake performance of heavy vehicles is improved by the development of new wheel-based functions for a longitudinal slip control braking system using novel Fast Acting Braking Valves (FABVs). To achieve this goal, Volvo Trucks' vehicle dynamics model has been extended to incorporate the FABV system. After validating the updated model with experimental data, a slip-slope based recursive least squares friction estimation algorithm has been implemented. Using information about the tire-road friction coefifcient, the sliding mode slip controller has been made adaptive to different road surfaces by implementing a friction dependent reference slip signal and switching gain for the sliding mode controller. This switching gain is further optimized by means of a novel on-line optimization algorithm. Simulations show that the on-line friction estimation converges close to the reference friction level within one second for hard braking. Furthermore, using this information for the optimized controller has resulted in reduction of braking distance on most road surfaces of up to 20 percent, as well as in most cases a reduction in air usage.

O objetivo desta pesquisa foi desenvolver um método para estabelecer o efeito das características e da operação dos veículos pesados sobre a infra-estrutura rodoviária através do uso de resultados de simulação microscópica de tráfego. Para tanto, foram definidos dois objetivos secundários. O primeiro deles envolveu a obtenção de um banco de dados detalhado sobre as características de 6.253 veículos pesados. Esses dados foram coletados em sete balanças localizadas em rodovias de pista dupla do estado de São Paulo. O segundo objetivo secundário consistiu na calibração e validação do simulador de tráfego CORSIM com o auxílio de um algoritmo genético, que adaptou simultaneamente 19 parâmetros do simulador para que ele reproduzisse adequadamente as características e o comportamento do tráfego observado em um trecho de rodovia de pista dupla paulista. A metodologia desenvolvida é demonstrada através da realização de um estudo de caso em que foram simulados dois cenários de tráfego e em que foi considerada uma ponte hipotética com 100 m de extensão. Nele, foram determinadas as probabilidades de ocorrências simultâneas dos veículos pesados sobre a ponte. As conclusões da pesquisa indicam que o método de calibração de simuladores de tráfego com uso de um algoritmo genético mostrou-se eficaz, reduzindo o erro médio de 9,11% para 6,32%. Além disso, as conclusões revelam que a obtenção de um carregamento móvel a partir de dados extraídos de um simulador de tráfego é possível e que a associação dos dados simulados a um banco de dados detalhado permite o cálculo do efeito do tráfego dos veículos pesados sobre a infra-estrutura rodoviária. Do estudo de caso, pode-se concluir que: (1) a probabilidade de ocorrência simultânea de veículos pesados sobre a ponte é muito freqüente e está diretamente relacionada às características do fluxo de tráfego e; (2) a distribuição das cargas e seus respectivos valores são fatores determinantes para o cálculo dos esforços. Os resultados do estudo de caso demonstram a viabilidade do procedimento proposto. / The objective of this research was to develop a method for establishing the effect of heavy vehicles characteristics and operation on highway infra-structure by using the results of a microscopic traffic simulation model. Therefore, two secondary objectives were defined. The first of them involved the attainment of a detailed database containing the characteristics of 6,253 heavy vehicles. This data was collected at seven weight stations located on the state of São Paulo multilane highways. The other secondary objective was to calibrate and validate CORSIM by using a genetic algorithm, which simultaneously adapted 19 model parameters in order to appropriately reproduce the characteristics and the behavior of the observed traffic flow. The developed methodology is demonstrated through a case study in which two traffic scenarios were simulated and in which a hypothetic 100 m bridge was considered. In the case study, the probabilities of simultaneous presence of heavy vehicles on the bridge were determined. The conclusions of this research indicate that the method of calibrating a traffic simulation model utilizing a genetic algorithm was efficient, reducing the mean error from 9.11% to 6.32%. Additionally, the conclusions reveal that the attainment of a live load from extracted data through a traffic simulation model is possible and that the association of simulated data with a detailed database allows the estimation of heavy vehicle traffic effect on highway infra-structure. From the case study, it can be concluded that: (1) the probability of simultaneous presence of heavy vehicles on the bridge is very frequent and; (2) the loads distribution and its respective values are determinant in calculating efforts. The case study results demonstrate the proposed procedure viability.

Autonomous driving is increasingly popular among people and automotive industries in realizing their presence both in passenger and goods transportation. Safer autonomous navigation might be very challenging if there is a failure in sensing system. Among several sensing systems, image classification plays a major role in understanding the road signs and to regulate the vehicle control based on urban road rules. Hence, a robust classifier algorithm irrespective of camera position, view angles, environmental condition, different vehicle size & type (Car, Bus, Truck, etc.,) of an autonomous platform is of prime importance. In this study, Convolutional Neural Network (CNN) based classifier algorithm has been implemented to ensure improved robustness for recognizing traffic signs. As training data play a crucial role in supervised learning algorithms, there come an effective dataset requirement which can handle dynamic environmental conditions and other variations caused due to the vehicle motion (will be referred as challenges). Since the collected training data might not contain all the dynamic variations, the model weakness can be identified by exposing it to variations (Blur, Darkness, Shadow, etc.,) faced by the vehicles in real-time as a initial testing sequence. To overcome the weakness caused due to the training data itself, an effective augmentation technique enriching the training data in order to increase the model capacity for withstanding the variations prevalent in urban environment has been proposed. As a major contribution, a framework has been developed to identify model weakness and successively introduce a targeted augmentation methodology for classification improvement. Targeted augmentation is based on estimated weakness caused due to the challenges with difficulty levels, only those necessary for better classification were then augmented further. Predictive Augmentation (PA) and Predictive Multiple Augmentation (PMA) are the two proposed methods to adapt the model based on targeted challenges by delivering with high numerical value of confidence. We validated our framework on two different training datasets (German Traffic Sign Recognition Benchmark (GTSRB) and Heavy Vehicle data collected from bus) and with 5 generated test groups containing varying levels of challenge (simple to extreme). The results show impressive improvement by ≈ 5-20% in overall classification accuracy thereby keeping their high confidence.

O objetivo desta pesquisa foi desenvolver um método para estabelecer o efeito das características e da operação dos veículos pesados sobre a infra-estrutura rodoviária através do uso de resultados de simulação microscópica de tráfego. Para tanto, foram definidos dois objetivos secundários. O primeiro deles envolveu a obtenção de um banco de dados detalhado sobre as características de 6.253 veículos pesados. Esses dados foram coletados em sete balanças localizadas em rodovias de pista dupla do estado de São Paulo. O segundo objetivo secundário consistiu na calibração e validação do simulador de tráfego CORSIM com o auxílio de um algoritmo genético, que adaptou simultaneamente 19 parâmetros do simulador para que ele reproduzisse adequadamente as características e o comportamento do tráfego observado em um trecho de rodovia de pista dupla paulista. A metodologia desenvolvida é demonstrada através da realização de um estudo de caso em que foram simulados dois cenários de tráfego e em que foi considerada uma ponte hipotética com 100 m de extensão. Nele, foram determinadas as probabilidades de ocorrências simultâneas dos veículos pesados sobre a ponte. As conclusões da pesquisa indicam que o método de calibração de simuladores de tráfego com uso de um algoritmo genético mostrou-se eficaz, reduzindo o erro médio de 9,11% para 6,32%. Além disso, as conclusões revelam que a obtenção de um carregamento móvel a partir de dados extraídos de um simulador de tráfego é possível e que a associação dos dados simulados a um banco de dados detalhado permite o cálculo do efeito do tráfego dos veículos pesados sobre a infra-estrutura rodoviária. Do estudo de caso, pode-se concluir que: (1) a probabilidade de ocorrência simultânea de veículos pesados sobre a ponte é muito freqüente e está diretamente relacionada às características do fluxo de tráfego e; (2) a distribuição das cargas e seus respectivos valores são fatores determinantes para o cálculo dos esforços. Os resultados do estudo de caso demonstram a viabilidade do procedimento proposto. / The objective of this research was to develop a method for establishing the effect of heavy vehicles characteristics and operation on highway infra-structure by using the results of a microscopic traffic simulation model. Therefore, two secondary objectives were defined. The first of them involved the attainment of a detailed database containing the characteristics of 6,253 heavy vehicles. This data was collected at seven weight stations located on the state of São Paulo multilane highways. The other secondary objective was to calibrate and validate CORSIM by using a genetic algorithm, which simultaneously adapted 19 model parameters in order to appropriately reproduce the characteristics and the behavior of the observed traffic flow. The developed methodology is demonstrated through a case study in which two traffic scenarios were simulated and in which a hypothetic 100 m bridge was considered. In the case study, the probabilities of simultaneous presence of heavy vehicles on the bridge were determined. The conclusions of this research indicate that the method of calibrating a traffic simulation model utilizing a genetic algorithm was efficient, reducing the mean error from 9.11% to 6.32%. Additionally, the conclusions reveal that the attainment of a live load from extracted data through a traffic simulation model is possible and that the association of simulated data with a detailed database allows the estimation of heavy vehicle traffic effect on highway infra-structure. From the case study, it can be concluded that: (1) the probability of simultaneous presence of heavy vehicles on the bridge is very frequent and; (2) the loads distribution and its respective values are determinant in calculating efforts. The case study results demonstrate the proposed procedure viability.