Because motor vehicle crashes have decreased during the last decade in many countries in the world and are more diffuse, local authorities have difficulties to define road safety policies. An experiment with 51 cars of public fleets equipped with a specific Event Data Recorder was carried out in France during one year. The purposes of this research were to evaluate if incident data (critical driving situations) help to understand crashes, and to explore a new way for road infrastructure safety diagnosis. The analysis of 339 genuine incidents and 1237 simple events recorded illustrates the potentiality of such an experiment and provides: some insights about conditions in which incidents occur, a general overview of their distribution according to different road layouts, as well as information on the different levels of accelerations reached. It can be noticed that there is an overrepresentation of incidents in right curves compared to left curves. The simple events involving mostly the infrastructure could be used to detect road defects. Genuine incidents where the vehicle is subjected to important dynamic demands, related to potentially unsafe driving situations, can be used to improve knowledge of the motor vehicle crashes thanks to incident mechanisms analysis.

Road profile acts as a disturbance input to the vehicle dynamics and results in undesirable vibrations affecting the vehicle stability. An accurate knowledge of this data is a key for a better understanding of the vehicle dynamics behavior and active vehicle control systems design. However, direct measurements of the road profile are not trivial for technical and economical reasons, and thus alternative solutions are needed. This paper develops a novel observer, known as a virtual sensor, suitable for real-time estimation of the road profile. The developed approach is built on a quarter-car model and uses measurements of the vehicle body. The road roughness is modeled as a sinusoidal disturbance signal acting on the vehicle system. Since this signal has unknown and time-varying characteristics, the proposed estimation method implements an adaptive control scheme based on the internal model principle and on the use of Youla–Kučera (YK) parameterization technique (also known as Q-parameterization). For performances assessment, estimations are comparatively evaluated with respect to measurements issued from Longitudinal Profile Analyzer (LPA) and Inertial Profiler (IP) instruments during experimental trials. The proposed method is also compared to the approach provided in Doumiati, Victorino, Charara, and Lechner (2011), where a stochastic Kalman filter is applied assuming a linear road model. Results show the effectiveness and pertinence of the present observation scheme.

This paper deals with the diagnosis of the critical driving situations. This work is divided into three parts: the first one presents two steering controllers using the sliding mode and the switched H∞ controllers, the second one describes an unknown input sliding mode observer, while the last one presents an approach to the diagnosis of critical driving situations. All existence conditions are established using the Lyapunov approach. Simulations are conducted to highlight the efficiency of the proposed approach using the experimental data recorded by an instrumented Peugeot 307 laboratory car. The diagnosis of the critical driving situations is achieved by the speed extrapolation concept using simultaneously a non–linear model, a steering control and road bank observer. The speed extrapolation concept is used in order to evaluate situations under high dynamic loads on a bend, using stability of the sideslip motion.

This paper proposes a switching steering vehicle control designed using the linear quadratic regulator (LQR) problem, the Linear Matrix Inequality (LMI) framework and the H∞ norm. The proposed switched control law comprises two levels: the first level is a switched Proportional–Integral-Derivative controller of lateral deviation (PIDy) and the second is a switched Proportional-Derivative controller of yaw angle (PDψ). These two levels are used to ensure an accurate tracking of the vehicle's lateral deviation y and yaw angle ψ. This control strategy makes use of a common Lyapunov function design method used for the stability analysis of switched continuous-time systems. Sufficient conditions for global convergence of the switched control law are presented and proved under arbitrary switching signals. All these conditions are expressed in terms of LMIs. The switched steering control was developed for an application seeking to identify approximately the maximum achievable speed in a bend. This application requires a steering control for simulating a realistic nonlinear four-wheel vehicle model and for performing a speed extrapolation test to evaluate the physical limits of a vehicle in a bend. This study includes the performance tests using experimental data from the Peugeot 307 prototype vehicle developed by IFSTTAR Laboratory.

One of the most important tasks of the vehicle dynamics estimation problem is to obtain the longitudinal, lateral, and vertical forces acting upon the wheels. These variables are fundamental for studying vehicle controllability and stability. However, they are difficult to obtain in terms of measurement techniques. Regarding lateral/tire forces, two observers based on the state-observer Extended Kalman Filter (EKF) theory and on the dynamic responses of a vehicle instrumented with standard sensors are proposed, discussed and compared in this study. The first observer estimates the lateral forces per axle using a random walk model, and then calculates the individual lateral tire forces on the basis of the vertical tire forces distribution. The second observer directly estimates the lateral forces per tire using a transient relaxation model. The developed observers are able to work in real-time in normal and in critical driving situations. Performances are tested using a laboratory car running on a wet track.

Improving road safety requires the assessment of risk indicators. Predicting these indicators is of major importance to in efforts to enhance safety systems and to warn drivers of dangerous situations. This article presents a prediction algorithm for risk indicators that combines assumptions about a vehicle's trajectory, velocity, and acceleration with existing road information to calculate risk indicators and detect possibly risky conditions in imminent driving situations. The algorithm is validated through experiments with a laboratory vehicle.

This article leads to the challenging problem of increasing vehicle driving security by applying on boarded intelligent diagnosis systems; it presents a methodology of evaluating, in an anticipated way, the risk of having an accident (skid and rollover). The methodology consists in adopting assumptions about the trajectory, the longitudinal velocity and the longitudinal acceleration in future instants and use these assumptions, allied to previous road information to calculate the future vehicle dynamics parameters. Once calculated, the risk indicators based on these parameters could be predicted in order to expect and avoid possible dangerous situations. These indicators are the lateral load transfer (LTR) based on vertical forces, and the lateral skid indicator (LSI) based ont the maximum friction coefficient and the used friction coefficient. A sliding window system is used to apply the method on the whole trajectory to take into account the vehicle dynamics updates by the driver.

Chapter 4 Estimation of the Lateral Tire Forces Moustapha Doumiati, Moustapha DoumiatiSearch for more papers by this authorAli Charara, Ali ChararaSearch for more papers by this authorAlessandro Victorino, Alessandro VictorinoSearch for more papers by this authorDaniel Lechner, Daniel LechnerSearch for more papers by this authorBernard Dubuisson, Bernard DubuissonSearch for more papers by this author Moustapha Doumiati, Moustapha DoumiatiSearch for more papers by this authorAli Charara, Ali ChararaSearch for more papers by this authorAlessandro Victorino, Alessandro VictorinoSearch for more papers by this authorDaniel Lechner, Daniel LechnerSearch for more papers by this authorBernard Dubuisson, Bernard DubuissonSearch for more papers by this author Book Editor(s):Moustapha Doumiati, Moustapha DoumiatiSearch for more papers by this authorAli Charara, Ali ChararaSearch for more papers by this authorAlessandro Victorino, Alessandro VictorinoSearch for more papers by this authorDaniel Lechner, Daniel LechnerSearch for more papers by this authorBernard Dubuisson, Bernard DubuissonSearch for more papers by this author First published: 17 December 2012 https://doi.org/10.1002/9781118578988.ch4Citations: 1 AboutPDFPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShareShare a linkShare onEmailFacebookTwitterLinkedInRedditWechat Summary This chapter outlines some of the fundamental concepts of the lateral tire force. First, a survey of existing methods dealing with parameters that describes the lateral tire force is presented. Then, the chapter presents the original method for estimating individual lateral tire force and sideslip angle. The developed observers are derived from a simplified four-wheel vehicle model, linear/nonlinear tire models and potentially available measurements. The estimation techniques are based on extended and unscented Kalman filtering techniques. Tire-road interaction is represented by a relaxation model involving a linear and a quasi-static Dugoff's tire model. Experimental data acquired on both dry and wet pavements confirm that the proposed algorithm performs reliably in a number of different maneuvers. In the future, the improvement of the vehicle/road model will be valuable to widen validity domains of the observer, and make it adaptive to road conditions. Controlled Vocabulary Terms Kalman filters; tyres; vehicle dynamics Citing Literature Vehicle Dynamics Estimation using Kalman Filtering RelatedInformation

This paper proposes the design of three control laws dedicated to vehicle steering control, two based on robust linear control strategies and one based on nonlinear control strategies, and presents a comparison between them. The two robust linear control laws (indirect and direct methods) are built around M linear bicycle models, each of these control laws is composed of two M proportional integral derivative (PID) controllers: one M PID controller to control the lateral deviation and the other M PID controller to control the vehicle yaw angle. The indirect control law method is designed using an oscillation method and a nonlinear optimisation subject to H-infinity constraint. The direct control law method is designed using a linear matrix inequality optimisation in order to achieve H-infinity performances. The nonlinear control method used for the correction of the lateral deviation is based on a continuous first-order sliding-mode controller. The different methods are designed using a linear bicycle vehicle model with variant parameters, but the aim is to simulate the nonlinear vehicle behaviour under high dynamic demands with a four-wheel vehicle model. These steering vehicle controls are validated experimentally using the data acquired using a laboratory vehicle, Peugeot 307, developed by National Institute for Transport and Safety Research - Department of Accident Mechanism Analysis Laboratory's (INRETS-MA) and their performance results are compared. Moreover, an unknown input sliding-mode observer is introduced to estimate the road bank angle.