Search Results

Forward collision warning is one of the most challenging concerns in the safety of autonomous vehicles. A cooperation between many sensors such as LIDAR, Radar and camera helps to enhance the safety. Apart from the importance of having a reliable object detector, the safety system should have requisite capabilities to make reasonable decisions in the moment. In this work, we concentrate on detecting front vehicles of autonomous cars using a monocular camera, beyond only a detection method. In fact, we devise a solution based on a cooperation between a deep object detector and a reinforcement learning method to provide forward collision warning signals. The proposed method models the relation between acceleration, distance and collision point using the area of the bounding box related to the front vehicle. An agent of learning automata as a reinforcement learning method interacts with the environment to learn how to behave in eclectic hazardous situations.

A two-stage preview strategy is proposed to characterize steering control properties of commercial vehicle drivers. The strategy includes a near and a far preview points to describe the driver control of lateral path deviation and vehicle orientation. A human driver model comprising path error compensation and dynamic motions of the limb is subsequently formulated and integrated to a yaw-plane model of an articulated vehicle. The coupled driver-vehicle model is analyzed under an evasive steering maneuver to identify limiting values of the driver control parameters through minimization of a generalized performance index comprising driver's steering effort, path deviations and selected vehicle states. The performance index is further analyzed to identify relative contributions of different sensory feedbacks, which may provide important guidance for designs of driver-assist systems (DAS).

One of the relevant applications of this study is related to designing anti-icing surfaces. Supercooled water droplet impact at high Weber number on a wing of airplane is one of the main concerns in aircraft ice accumulation. In order to address this issue, an experimental setup which generates co-flow is designed to mimic the real scenario of droplet impact in practical icing condition. This process is observed using a high-speed camera to capture the correct moment of sliding at 10000 frames per second and 120000 1/s shutter speed. Different air stream velocities are generated by a convergent nozzle with a maximum Mach number of 0.1. Two different cases are considered. First, droplet impact in still air with an impact velocity of 2 m/s is performed as the base case. Then droplet impact accompanied with 10, 18 and 20 m/s air stream velocities having the same impact velocity are conducted. Droplet impact velocity will change either by air flow or gravity.

Compared to the vehicles with conventional steering, the articulated frame steer vehicles (ASV) are known to exhibit lower directional and roll stability limits. Furthermore, the tire interactions with relatively rough terrains could adversely affect the directional and roll stability limits of an ASV due to terrain-induced variations in the vertical and lateral tire forces. It may thus be desirable to assess the dynamic safety of ASVs in terms of their directional control and stability limits while operating on different terrains. The effects of terrain roughness on the directional stability limits of an ASV are investigated through simulations of a comprehensive three-dimensional model of the vehicle with and without a rear axle suspension. The model incorporates a torsio-elastic rear axle suspension, a kineto-dynamic model of the frame steering struts and equivalent random profiles of different undeformable terrains together with coherence between the two tracks profiles.

Integrated modular avionics (IMA) architectures host multiple federated avionics applications on a single platform and provide benefits in terms of size, weight, and power, which, however, leads to increased complexity, especially during the development process. To cope efficiently with the high level of complexity, a novel, structured development methodology is required. This paper presents a model-based systems engineering (MBSE) development approach for the so-called “distributed integrated modular architecture” (DIMA). The proposed methodology adapts the open-source Capella tool, based on the Architecture Analysis & Design Integrated Approach (ARCADIA) methodology, to implement a complete design cycle, starting with requirements captured from the aircraft level to streamline the development, culminating in the integration of an avionics application into an ARINC 653 platform.

This paper proposes a new methodology to conduct thermal analysis in the conceptual phase of the aircraft development process. Traditionally, thermal analysis is conducted after the system architecture has already been defined. The aircraft system thermal environment evaluation may lead to late design changes that can have a significant impact on the development process. To reduce the risk of late design changes, thermal requirements need to be defined and validated in the conceptual design phase. This research paper introduces a novel multi-level modeling strategy based on a bottom-up approach. It proposes an automatic geometrical simplification procedure for Computational Fluid Dynamic (CFD) analysis, a methodology for the generation of analytical correlations based on highly detailed methods, and a thermal risk assessment approach based on dimensionless numbers.

This paper presents natural frequency analysis of the individual components of the railway vehicle and entire track system obtained through eigenvalue analysis. Natural frequencies of the vehicle components have been identified considering rigid body mass. In order to identify the natural frequencies of the track, the three-dimensional railway track model has been simplified to a generalized track element with equivalent stiffness representing the railpad and ballast stiffness and sleeper mass. The nonlinear Hertzian contact stiffness has been converted in linear contact stiffness in order to facilitate the eigenvalue analysis of the vehicle. Linear parameters of the railpad and ballast stiffness have also been considered for eigenvalue analysis of the track model while damping properties of the railpads and ballasts are ignored.

To improve the power and economy performance of pure electric vehicles, chaos particle swarm optimization (CPSO) algorithm is adopted in this study to optimize the parameters of the power system. The optimized parameter is then imported into CRUISE. The whole vehicle performance simulation in power system optimization for pure electric vehicle is carried out in CRUISE. Simulation results show that optimized vehicles can meet the expected dynamic performance and the driving range has been greatly improved. Meanwhile, it is also viable that the parameters of the optimal objective function can achieve the purpose of balancing the power performance and economic performance, which provides a reference for the development of vehicle power performance.

An Active Independent Front Steering (AIFS) offers attractive potential for realizing improved directional control performance compared to the conventional Active Front Steering (AFS) system, particularly under more severe steering maneuvers. The AIFS control strategy adjusts the wheel steer angles in an independent manner so as to utilize the maximum available adhesion at each wheel/road contact and thereby compensate for cornering loss caused by the lateral load transfer. In this study, the performance potentials of AIFS are explored for vehicles experiencing greater lateral load transfers during steering maneuvers such as partly-filled tank trucks. A nonlinear yaw plane model of a two-axle truck with limited roll degree-of-freedom is developed to study the performance potentials of AIFS under different cargo fill conditions.

A range of axle suspensions, comprising hydro-pneumatic struts and diverse linkage configurations, have evolved in recent years for large size mining trucks to achieve improved ride and higher operating speeds. This paper presents a comprehensive analysis of different independent front suspension linkages that have been implemented in various off-road vehicles, including a composite linkage (CL), a candle (CA), a trailing arm (TA), and a double Wishbone (DW) suspension applied to a 190 tons mining truck. Four different suspension linkages are modeled in MapleSim platform to evaluate their kinematic properties. The relative kinematic properties of the suspensions are evaluated in terms of variations in the kingpin inclination, caster, camber, toe-in and horizontal wheel center displacements considering the motion of a hydro-pneumatic strut. The results revealed the CL and DW suspensions yield superior kinematic response characteristics compared to the CA and TA suspensions.

Rivulet dynamics is involved in different scientific problems and industrial applications from production of microchips, to rain flow on structural systems such as wind turbine blades and aircraft wings. In the latter case for example, rivulet flow leads to accumulation of ice on airfoil surface that may affect the aircraft performance. Hence, understanding the dynamics of rivulets is important in solving various scientific problems. In this paper the results of a numerical simulation based on smoothed particle hydrodynamic (SPH) method will be reported on dynamics of rivulets under the effect of various air shear speeds and different surface morphologies. The simulation results will be compared and validated with experiments to shed more lights in understanding the mechanism of shear driven rivulet flow on solid substrates.

Experimental study is performed to analyze the shedding behavior of droplets with different shear flow speeds typical of those in the flight conditions. Droplet shedding phenomena has significant effect on ice accumulation on critical components such as airfoil and nacelle. In order to mimic this scenario experimental set up is designed to generate shear flow as high as 90m/s. The high shear effect is combined to the surface wettability impact by using hydrophilic and superhydrophobic surfaces. It is shown that the wetting length of the droplet on hydrophilic surface increases by shear speed while on the superhydrophobic surface a drastic reduction on wetting length is detected. Furthermore, it is observed that the droplet is detached from the superhydrophobic surface with moderate shear speeds.

In this paper, a gain-scheduling optimal control approach is proposed to enhance yaw stability of articulated commercial vehicles through active braking of the proper wheel(s). For this purpose, an optimal feedback control is used to design a family of yaw moment controllers considering a broad range of vehicle velocities. The yaw moment controller is designed such that the instantaneous tractor yaw rate and articulation angle responses are forced to track the target values at each specific vehicle velocity. A gain scheduling mechanism is subsequently constructed via interpolations among the controllers. Furthermore, yaw moments derived from the proposed controller are realized by braking torque distribution among the appropriate wheels. The effectiveness of the proposed yaw stability control scheme is evaluated through software-in-the-loop (SIL) co-simulations involving Matlab/Simulink and TruckSim under lane change maneuvers.