Résumé (MAX 2200 caractères avec les espaces)
Ce projet vise à établir de nouvelles méthodes pour la navigation automobile ainsi que pour la gestion optimale d’une flotte de véhicules en environnement hostile. En outre, le projet vise également à développer des mesures innovantes pour l’analyse temps réel des comportements de conduite dangereuse ainsi que l’analyse temps réel des accidents de voiture afin d’améliorer la sécurité globale des conducteurs Canadiens. De façon générale, cette recherche propose de combiner les mesures d’un récepteur GPS haute sensibilité avec celles provenant d’un système autonome de navigation inertielle ainsi que d’autres capteurs autonomes complémentaires tels que l’odomètre et les magnétomètres. Par ailleurs, afin de fournir une solution abordable, le système cible sera basé exclusivement sur l’utilisation de capteurs à très faible coût. Il est attendu que ce projet permettra une réduction significative de l’empreinte environnementale des véhicules automobiles en plus d’avoir un impact positif sur la sécurité globale des véhicules ciblés. Par exemple, l’amélioration de la précision sur la localisation des véhicules routiers permettrait de réduire considérablement le temps requis afin de trouver un véhicule volé ou égaré, ce qui peut avoir des répercussions importantes sur les finances des entreprises Canadiennes. De plus, l’établissement d’un système de suivi des comportements de conduite des automobilistes pourrait permettre la mise en place d’un nouveau système de taxation basé sur l’utilisation de la voiture ou sur le comportement de conduite, ce qui, selon des études récentes, permettrait de réduire jusqu’à 10% les émissions de gaz à effet de serre des véhicules ciblés. Finalement, la reconstruction précise d’un accident de voiture en temps réel permettrait de prédire les besoins spécifiques sur une scène d’accident, améliorant ainsi le temps de réaction ainsi que la sécurité globale des automobilistes. La preuve de concept sera d’abord réalisée en laboratoire ainsi que sur route à l’aide de matériel de simulation et d’une voiture de test en vue de caractériser les performances du système. Le projet contribuera aux initiatives internationales afin de réduire les émissions de gaz à effet de serre, et de créer de nouveaux emplois pour l’équipe de personnel hautement qualifié.
Responsibilities of the candidate:
According to the schedule, the first Ph.D. student will be in charge of the following tasks:
1) 2-21 Review of the actual analysis metrics
2) 2-22 Modeling of the preliminary analysis metrics
3) 2-23 Metrics implementation using simulation software
4) 2-24 Metrics validation using simulation tools
5) 3-23 A-GNSS integration with navigation algorithms / analysis metrics
6) 3-32 Monitoring data integration with navigation algorithms / analysis metrics
7) 3-41 Implementation of the driving behaviour analysis metrics
8) 3-42 Real car test setups and planning
9) 3-43 Driving behaviour test realisation
10) 3-44 Results analysis/study on possible metrics improvement
11) 3-45 Modeling and testing of improved analysis metrics
12) 3-71 Review of the actual analysis metrics
13) 3-72 Modeling of the preliminary analysis metrics
14) 3-73 Metrics implementation using simulation software
15) 3-74 Metrics validation using simulation tools
16) 3-81 Implementation of the accident analysis metrics
17) 3-82 Real car test setups and planning
18) 3-83 Real car accident test realisation
19) 3-84 Results analysis/study on possible metrics improvement
20) 3-85 Modeling and testing of improved analysis metrics
21) 4-21 Modeling of improved analysis metrics
22) 4-22 Metric validation in controlled environment
23) 4-23 Final driving behaviour test realisation
24) 4-24 Final real car accident tests realisation
25) 4-25 Analysis metrics validation and performance analysis
26) 4-51 Test setup and planning for vehicle fleet testing
27) 4-52 System implementation on different vehicles
28) 4-53 Intensive testing of the fleet management system
29) 4-54 Results analysis and management system improvement
30) 4-55 Implementation of the improved system on vehicle fleet
31) 4-56 Validation of the fleet management system
The objective of this student’s project is to develop innovative metrics for real-time monitoring of driving behaviour and real-time analysis and diagnosis of car accidents. These metrics will be based on measurements from various sensors strategically placed on the vehicle’s frame and results will be carried out using advanced signal processing techniques. He will closely collaborate with the other members of the research team and he will use a part of the research done by the Master’s and trainee students in order to fulfill his project’s goals.
The main goals of the first Ph.D. student’s work are the following:
- During the second year, the PhD student will focus on the study of the dynamics of automotive vehicles during a road accident. He will first consider major types of accidents including, but not limited to, frontal impact, side impact, turnovers and skidding. This work will first involve an in-depth literature review on the state of the art techniques currently used in the accident analysis domain. Following this study, the Ph.D. student will evaluate possible methods for the simulation of road accidents in order to obtain realistic measurements on the dynamics of a car during various accident scenarios. Once the dynamics of accidents studied, the Ph.D. student will conduct a comprehensive analysis of the raw measurements in order to determine appropriate sensors as well as their location on the vehicle’s frame. During the first year, the Ph.D. student will also model preliminary metrics solely based on simulated measurements. These metrics will first be implemented, analyzed and validated in Matlab / Simulink.
- The third year will be mainly dedicated to the real-time implementation and integration of the modeled metrics into the Orchid VTADS prototype. During this phase, the metrics will be intensively tested in order to quantify their performances, accuracy and robustness in various simulated as well as realistic environments. Indeed, real tests will be conducted during this year including dangerous driving behaviour tests executed by a professional driver on a closed circuit as well as car crash testing realized by a specialized company. Following these tests, the Ph.D. student will analyze the results of the preliminary metrics in order to identify its major weaknesses and suggest possible improvements.
- In response to the third year, the last year will be mainly dedicated to the identification of major flaws in the current analysis metrics and its improvement. During this phase, the Ph.D. candidate will return to research in order to explore new approaches for improving these metrics using data recorded during the real environement tests. The student will then model new innovative metrics that are more adapted to realistic environments. He will first validate these metrics in simulation using Matlab / Simulink and then test it on the real data using the previously recorded measurements. Finaly, a second phase of real tests will be conducted in order to validate the real-time analysis and diagnosis metrics.
