Search Results

This presentation evaluates the contribution of multi-objective programming scheme for the conceptual design of the Hybrid Electric Vehicle frame's structure using topological optimization. The compromise programming method was applied to describe the statically loaded multicompliance topology optimization. Solid Isotropic Material with Penalization (SIMP) was used as the interpolation scheme to indicate the dependence of material modulus upon regularized element densities. The sequential convex programming approach was applied to solve the optimization problem. The application on the chassis frame was used to demonstrate the characteristics of the presented methodologies based on the commercial software package OptiStruct.

How to control and analyze the turning process of a tracked vehicle with electric transmission system is an important issue. In the paper two turning control methods are presented according to its study. The balance relations of its tractive effort and power versus radius are obtained by the calculation with using the mathematical model of basic turning dynamics and constraint conditions. The model of continuously variable radius turning is implemented by a RBF neural network which is of the better identifying ability, and the more turning results can be given from it. These turning analyses are significant for the electric transmission system.

The Opposed Piston Two-Stroke (OPTS) engine has many advantages on power density, fuel tolerance, fuel flexibility and package space. A type of self-balanced opposed-piston folded-crank train two-stroke engine for Unmanned Aerial Vehicle (UAV) was studied in this paper. AVL BOOST was used for the thermodynamic simulation. It was a quasi-steady, filling-and-emptying flow analysis -- no intake or exhaust dynamics were simulated. The results were validated against experimental data. The effects of high altitude environment on engine performance have been investigated. Moreover, the matching between the engine and turbocharger was designed and optimized for different altitude levels. The results indicated that, while the altitude is above 6000m, a multi-stage turbocharged engine system need to be considered and optimized for the UAV.

In this paper, a novel driver model is proposed to track vehicle speed in MIL (Model-in-the-Loop) test system, which has structural consistency with HIL (Hardware-in-the-Loop) test system. First, the MIL test system which contains models of driver, vehicle and test bench is established. Second, according to the connections of the established models in Matlab/Simulink environment, the vehicle speed is calculated in vehicle model. Emphatically, through the deviation between driving cycle speed and calculated vehicle speed, PI controller in driver model adjusts the vehicle speed to ideal point through sending the torque command to drive motor, the ILC (Iterative Learning Control) controller modifies and stores P value of PI controller. Then, in order to obtain the better modification of PI controller, iterative learning control algorithm is deeply researched in term of types and parameters.

Study and modeling of diesel combustion during transient operations is an important scientific objective. This is partially due to the fact that emissions under transient operations have aroused increasing attention by control groups during recent decades. The objective of this paper is to develop a combustion model to predict the peculiarities of transient combustion for developing and testing control strategies. To by-pass the complicated principles of transient combustion, the Neural Networks are applied to link the coefficients in an empirical combustion model with engine operating parameters. Finally, the Neural Networks combustion model would not only reflect the influence of turbocharge lag on combustion process during transient event, which cannot be predicted by its interpolation alternative, but also shown great potential for analyzing combustion characteristics during load increase transient event or other transient operations.

This paper proposes an effective nonlinear bicycle model including longitudinal, lateral, and yaw motions of a vehicle. This bicycle model uses a simplified piece-wise linear tire model and tire force tuning algorithm to produce closely matching vehicle trajectory compared to real vehicle for wide vehicle operation ranges. A simplified piece-wise tire model that well represents nonlinear tire forces was developed. The key parameters of this model can be chosen from measured tire forces. For the effects of dynamic load transfer due to sharp vehicle maneuvers, a tire force tuning algorithm that dynamically adjusts tire forces of the bicycle model based on measured vehicle lateral acceleration is proposed. Responses of the proposed bicycle model have been compared with commercial vehicle dynamics model (CarSim) through simulation in various vehicle maneuvers (ramp steer, sine-with-dwell).

Due to the poor ability of millimeter wave radar in recognizing distant static objects, target loss and incomplete information will occur when it recognizes the static target in front, thus increasing the false alarm rate and missing alarm rate of the radar-dependent driving assistant system, which will reduce the driving safety and the acceptability of the assistant system. Aiming at the radar's poor ability to recognize static targets, this paper uses a model based on machine learning algorithm to recognize and track targets. The radar signals are collected and processed in different conditions, and the results show that the radar has a poor recognition effect when the distance is more than 100 meters and the speed is more than 19m/s.

The fusion of multi-modal perception in autonomous driving plays a pivotal role in vehicle behavior decision-making. However, much of the previous research has predominantly focused on the fusion of Lidar and cameras. Although Lidar offers an ample supply of point cloud data, its high cost and the substantial volume of point cloud data can lead to computational delays. Consequently, investigating perception fusion under the context of 4D millimeter-wave radar is of paramount importance for cost reduction and enhanced safety. Nevertheless, 4D millimeter-wave radar faces challenges including sparse point clouds, limited information content, and a lack of fusion strategies. In this paper, we introduce, for the first time, an approach that leverages Graph Neural Networks to assist in expressing features from 4D millimeter-wave radar point clouds. This approach effectively extracts unstructured point cloud features, addressing the loss of object detection due to sparsity.