Search

Accidents involving Heavy Vehicles (HV) have serious consequences for road users,and incidents induce major congestion or damage to the environment or the infrastructure. Statistics show that accidents involving such vehicles are more serious than light vehicles (LV). The mortality rate is twice in an accident involving a heavy vehicle (see table 1.1). While they constitute only 3% vehicles in traffic, HV represent10% of the vehicles involved in fatal accidents (13% of HV-related accidents aredeadly against 6% for LV, and 85% of deaths are not HV users). Several types of truck accidents can occur on the road. The most important risks involving HV are rollover (Roll instability), jackknifing (Yaw instability), lane departure (Transversal instability) and collision (Longitudinal instability). All these accidents are characterized by appropriate risks criteria. These are used in accident analysis and designing appropriate prevention systems (warning systems and active control). In this work, we focus on the criteria of rollover and lane departure, but the developed approaches can be generalized to other criteria and other types of accidents.

During the past two decades, significant progress has been made in the field of higher order sliding mode controllers and observers. These observers/controllers ensure the insensibility of the system with respect to matched smooth uncertainties/perturbations by usage of the absolutely continuous controllers. Moreover, higher order sliding mode-based observers ensure the best possible asymptotic accuracy of the state observation and unknown inputs identification with respect to sampling step and deterministic bounded noises. This explains why several researchers have started to use the new techniques for the vehicle dynamics analysis and design. The objective of this Special Issue of IJVD is to present recent research concerning vehicle dynamics, going from the dynamics observation (acceleration, side slip, force estimation, etc.), parametric identification, risk prediction (rollover, jackknifing, etc.) and vehicle control (lane departure avoidance, steering control, etc.) based on modern variable structure techniques (sliding mode observers, sliding mode observers based identification, sliding mode (SM) control). The specific topics of interest within the scope of this special issue include: SM-based vehicles parameters and forces estimation, SM-based control of automotive driver robot, SM control of motor bicycle, SM control of unicycle, SM control of wheel slip, SM-based observer of heavy vehicle, Vehicle disturbance observation.

In this chapter, an original approach to heavy vehicles rollover risk prediction is presented and validated experimentally. It is based on the calculation of the LTR (Load Transfer Ratio) which depends on the estimated vertical forces using high order sliding mode observers. Previously, a tractor model is developed. The validation tests were carried out on an instrumented truck rolling on the road at various speeds and lane-change manoeuvres. Many scenarios have been experienced : driving straight, curved trajectories, zigzag manoeuvre and brake tests to emphasize the rollover phenomenon and its prediction to set off an alarm for the driver. In this study, the vehicle dynamic parameters (masses, inertias, stiffness..) and the static forces infrastructure characteristics (road profile, radius of curvature, longitudinal and lateral slope, skid resistance) are measured or calculated before the tests.

In the present work, an active steering assistance system for heavy vehicles is developed to prevent lane departure and let the vehicle follow the road’s centre line. The controller is based on the super-twisting algorithm. An estimator based on the sliding mode observer is also developed to estimate the vehicle dynamics. The lateral positions and speeds are then controlled to keep the vehicle in the centre of the lane. The lateral offset and the relative yaw angle are supposed to be measured and the road curvature to be known. Some simulation results are given to show the quality of this steering assistance system.

The aim of presented work consists in developing an active steering assistance system to avoid the rollover of heavy vehicle. Single body model of heavy vehicle is developed and presented in this paper. An estimator based on the high order sliding mode observer is developed in order to estimate the vehicle dynamics. The lateral acceleration limit is then estimated and lateral position and lateral speed are controlled using twisting algorithm in order to avoid the rollover of the vehicle. In the same time, an identification of some parameters of the model has been done in order to increase the robustness of the method. Some simulation results are given to show the quality of the proposed approach.

Vehicles are complex mechanical systems with strong nonlinear characteristics and which can present some uncertainties due to their dynamic parameters such as masses, inertias, suspension springs, tires side slip coefficients, etc. A vehicle is composed of many parts, namely the unsprung mass, the sprung mass, the suspension which makes the link between these two masses and therefore ensures passenger comfort, and also the pneumatic which absorbs the energy coming from the road and ensures contact between the vehicle and the road. In addition to its complexity and the presence of many nonlinearities and uncertainties, the presence of some external perturbations, such as the wind and the road inputs with its own characteristics (radius of curvature, longitudinal and lateral slop, road profile and skid resistance) can cause risks not only to the vehicle but also to passengers and other road users. Many methods have been developed in order to understand the behavior of a vehicle, control it and assist the driver in order to avoid possible lane departures, rollover or jackknifing risks, to ensure a better passenger comfort by means of a suspension control and/or to estimate a safety speed and trajectory. The present book is an attempt to show how the sliding mode based observation, uncertainties identification and parameter estimation may be applied in the control of vehicle dynamics as well as for parameter and perturbations estimation.

In this paper, an experimental validation of vertical forces estimation is presented. An original method using High Order Sliding Mode Observer is developed. The test bench is a static vehicle excited vertically by hydraulic jack. It's equipped by different sensors in order to measure its dynamics. Real time tests are done and results are presented in this paper.

Cet article traite de la validation expérimentale de la prévision de risque de renversement des poids lourds. Une méthode originale de prévision de risque de renversement est décrite. Elle est basée sur le calcul de l'indicateur de risque de renversement qui est le LTR (rapport de transfert de charge) qui dépend des forces verticales estimées en utilisant les techniques d'observateurs à modes glissants. La validation expérimentale est faite sur un banc d'essai représenté par un véhicule léger à l'arrêt.

Cet article présente une nouvelle méthode pour évaluer et prédire dynamiquement le risque de renversement de poids lourds par une approche fiabiliste. La méthode est basée sur la prise en compte des incertitudes du système. Les paramètres les plus influents sur le risque sont représentés par des variables aléatoires, alors que les variables dynamiques sont estimées par un filtre de Kalman Unscented, dédié aux systèmes non linaires. L'ensemble des variables aléatoires obtenues est utilisé pour déterminer à chaque instant un indice de fiabilité permettant de caractériser la sécurité du poids lourd vis-à-vis d'un éventuel risque de renversement. Cet indice est ensuite prédit pour une meilleure prévention du risque. Enfin, des simulations sont effectuées pour valider l'approche développée.

In this paper, an experimental validation of rollover risk evaluation of heavy vehicle is presented. This work was developed by the LCPC in collaboration with its industrial partners, Renault Trucks, Michelin and Sodit in the framework of French project VIF (Véhicule Lourd Interactif du Future). An original method about rollover risk evaluation is described. It's based on the calculation of the LTR (Load Transfer Ratio) which depends on the estimated vertical forces using sliding mode observers techniques.

A lot of researches about vehicle-road interaction have been performed without taking into account the infrastructure characteristics (radius of curvature, skid resistance, evenness, super elevation, slope...). These signals coming from the road are very important in the vehicle dynamics, because they contribute to generate the contact tire forces of heavy vehicles. In this paper, we show the influence of the road profile in the heavy vehicle modelling and we propose a method to evaluate the vertical forces based on third order sliding mode observer. This technic is very common in the dynamic systems field and the results of researchs are very satisfactory. First, we develop a heavy vehicle model in interaction with road. Some simulation results are given in order to prove the robustness of the proposed method. Then the vertical forces are estimated to calculate the LT R (Load Transfer Ratio) coefficient, which allow us to evaluate the rollover risk of heavy vehicle.

Many researches on vehicle-road interaction are performed without taking into account the response of the road profile (surface). However, these signals providing from road are very important in the vehicle dynamics, since they contribute to generate the contact tire forces of heavy vehicles. We show in this paper, the influence of the road profile in the heavy vehicle modelling. In this work, we propose a method to evaluate these vertical forces based on modification of the super twisting algorithm. This technic is very much used for the dynamic systems and the results of various research are very satisfactory. Design of such observers requires a dynamic model. Then, we develop a heavy vehicle model in interaction with road (this road profile is considered as inputs of the heavy vehicle model). In a validation procedure, we compare simulation results coming from a developed heavy vehicle estimator and responses done with the PROSPER's simulator which has been validated with numerous experiments on proving ground. When the vertical forces are finally estimated, we can calculate the LTR (Load Transfer Ratio) coefficient to evaluate the rollover risk.

Vehicle motion simulation accuracy, such as in accident reconstruction or vehicle controllability analysis on real roads, can be obtained only if valid road profile and tire-road friction models are available. Regarding road profiles, a new method based on Sliding Mode Observers has been developed and is compared to LPA measure.

This paper deals with a new method for estimation of the road profile using partial measurements of vehicle dynamics. This method is based on a robust observer designed with a nominal dynamic model of the vehicle. This model is a linear half-car model whose inputs have been considered as unknown states to be estimated. The estimation accuracy of our observer has been validated experimentally using a trailer equipped with position sensors and accelerometers.