Search Results

As an off-road racing car, driving conditions for a Baja racing car are particularly complex. Extreme working conditions such as deep pits and rocky roads have put higher demand on structural strength and frame safety. To solve this problem, extreme working conditions are first studied to check the safety of the steel tube frame of Baja racing cars. Secondly, based on Noise, Vibration, and Harshness (NVH) to explore the frame's characteristics, analyze the frame's six-order mode, make the corresponding optimization, and solve the resonance problem caused by engine excitation and other factors. Finally, the natural frequency of the frame is measured to verify the effectiveness of the NVH characteristic optimization results, and it is found that the experimental results match the theoretical values. The theoretical analysis results are mainly based on ANSYS software's static and modal analysis.

Intelligent sweeper vehicle is gradually applied to human life, in which the accuracy of garbage identification and classification can improve cleaning efficiency and save labor cost. Although Deep Learning has made significant progress in computer vision and the application of semantic network segmentation can improve waste identification rate and classification accuracy. Due to the loss of some spatial information during the convolution process, coupled with the lack of specific datasets for garbage identification, the training of the network and the improvement of recognition and classification accuracy are affected. Based on the Unet algorithm, in this paper we adjust the number of input and output channels in the convolutional layer to improve the speed during the feature extraction part. In addition, manually generated datasets are used to greatly improve the robustness of the model.

The road feature has an important influence on the safe speed of the unmanned vehicle and the safe space between two vehicles. Real-time access to the features of the road ahead of time and timely adjustment of engine torque are significant to unmanned driving. Most of the researches nowadays make full use of vehicle sensor technology and environment perception technology. Vehicle sensor is widely used to collect the features of the road. While in this paper, a new type of road feature extraction is proposed based on vehicle speed change. Under the premise of less sensor installed, vehicle speed-time data series is collected. The pavement parameters can be estimated with vehicle speed. Based on the vehicle dynamics, this paper studies the relationship between vehicle speed and rolling resistance. Different road features have different influences on road friction resistance.

Li-ion power battery is the core component of the electric vehicle power system, and the battery temperature will increase because of the electrochemical reaction of the Li-ion battery. The heat accumulates inside of the battery, which can degrade the working performance of the power battery and shorten the battery cycle life. At present, the wind cooling technology is relatively mature. However, it cannot achieve ideal heat dissipation effect under the working conditions of the high-power or high ambient temperature. In this research, the battery thermal management is carried out by the characteristics of the working fluid’s flowing phase change heat transfer. The phase change working fluid is R-1233zd(E) which is a kind of environmentally friendly liquid with nonconductive and nonflammable. It can achieve the purpose of controlling the battery’s temperature using the characteristics of isothermal heat absorption under different gas phase rate of phase change working fluid.

The automobile exhaust energy can be recovered by the thermoelectric module generator(TEG). Owing to the complex urban traffic, the exhaust gas’s temperature fluctuations are resulted, which means the unstable hot-end temperature of the TEG. By installing solid heat capacity material(SHCM) to the area between the outer wall of the exhaust pipe and the TEG, it is possible to appropriately reduce the temperature fluctuation, but there is still a fluctuation of the TEG’s power output. Then by adding voltage filter circuit (VFC) after the TEG, the power output stability can be improved. This research uses SHCM and VFC to improve the stability of the exhaust gas generation. Firstly, the three-dimensional heat transfer model of the exhaust pipe thermoelectric power generation system is established. The heat capacity materials with low thermal resistance and high heat capacity were selected as the research object based on previous research.

Air disc brake (ADB) is the execution unit of automobile brake, which plays a vital role in traffic safety. Lever is a very important driving as well as stress part of ADB. However, seldom periodical or thesis has given specifications about the lever. In this paper, working principle of the lever is illustrated in detail. Finite element analysis (FEA) of the lever is conducted and some structural problems of the lever are found out. Finally, for the condition that the lever doesn't meet the strength requirement, topography optimization is introduced to improve the shape of lever. Simulation results show that the strength of the lever improved obviously with the optimization, and analysis method as well as optimization method in this paper is feasible.

According to the resonant pavement crusher's work principle, its front frame mounted with the resonance system must meet the needs of the structural requirements. To satisfy the strength and stiffness requirement and avoid the resonance, the natural frequency of the front frame should be designed away from the crusher's working frequency. In this paper, the author builds a finite element model of the front frame and analyses its modal. According to the modal analysis results, the fourth modal frequency is close to the working frequency of the crusher. So the front frame should be optimized. In the finite element model, the front frame has been divided into a number of components of shell elements. Through optimal Latin hypercube experimental design, the author analyses the different component thickness's relationship of the frequencies of the front frame. The components with higher correlation coefficient have been chosen as the variables of optimization.

Based on the fuzzy inference system, it constructs a discretionary lane-changing decision model for different types of preceding vehicles and compares and analyzes the parameter differences of their input membership functions. According to the driver questionnaire survey, the model uses three parameters that drivers can easily percept as the model input—preceding vehicle distance in the current lane, preceding vehicle distance in the target lane, and following-vehicle distance in the target lane—uses Next-Generation Simulation (NGSIM) vehicle trajectory data to optimize the input membership functions of models based on genetic algorithm according to different vehicle lane-changing trajectory data to analyze the impact of the preceding vehicle type before lane change to the intelligent lane-changing decision.

A key problem of designing a light off-road vehicle with separate frame construction is to improve its torsional characteristic, which has a significant influence on the performance of the vehicle. Inevitably, a certain distortion of the body would be produced by the vibration and impact passing from the road. In present research, an analysis model of light off-road vehicle is established based on the theories and methods of finite element (FEM). The static stiffness of the body is simulated and the deformation of openings on the body, mainly the windows and the doors of the vehicle is studied. On the working conditions of torsion and braking combination, torsion and cornering combination, diagonal dangling, ultimate torsion of unilateral wheels and diagonal wheels, the static strength of separate frame construction is studied as well. The stress concentration regions are obtained according to the results of simulation.

Accurate surrounding vehicle motion prediction is critical for enabling safe, high quality autonomous driving decision-making and motion planning. Aiming at the problem that the current deep learning-based trajectory prediction methods are not accurate and effective for extracting the interaction between vehicles and the road environment information, we design a target vehicle intention-aware dual attention network (IDAN), which establishes a multi-task learning framework combining intention network and trajectory prediction network, imposing dual constraints. The intention network generates an intention encoding representing the driver’s intention information. It inputs it into the attention module of the trajectory prediction network to assist the trajectory prediction network to achieve better prediction accuracy.

The battery liquid cooling system can ensure that the battery works within a suitable temperature range, improve the safety performance of the battery system, and ensure the cruising range. This paper introduces a design scheme of a stamped double-parallel liquid cooling plate. Based on the STAR-CCM+ simulation software, a thermal simulation model of the battery management system is established to analyze the thermal behavior of the battery system and to study the effect of the inlet mass flow rate on the temperature of the top surface of the batteries. At the same time, with the analysis of the proportion of pressure drop of each component in the liquid cooling plate, an optimization of inserted part in the liquid cooling plate is proposed. The numerical analysis results are compared with the experimental results of the pressure drop to improve the effectiveness of the optimization scheme.

In order to solve the environmental pollution and energy crisis, Electric Vehicles (EVs) have been developed rapidly. Lithium-ion (Li-ion) battery is the key power supply equipment for EVs, and the scientific and accurate prediction of its Remaining Useful Life (RUL) has become a hot topic in the field of new energy research. The internal resistance and capacity are often used to characterize the Li-ion battery State of Health (SOH) from which RUL is obtained. However, in practical applications, it is difficult to obtain internal resistance and capacity information by using the non-intrusive measurement method. Therefore, it is necessary to extract the measurable parameters to characterize the degradation of Li-ion battery. At present, the methods of extracting health indicators based on measurable parameters have gained preliminary results, but most of them are derived from the Li-ion battery discharging data.

Deep Deterministic Policy Gradient (DDPG) is one of the Deep Reinforcement Learning algorithms. Because of the well perform in continuous motion control, DDPG algorithm is applied in the field of self-driving. Regarding the problems of the instability of DDPG algorithm during training and low training efficiency and slow convergence rate. An improved DDPG algorithm based on segmented experience replay is presented. On the basis of the DDPG algorithm, the segmented experience replay select the training experience by the importance according to the training progress to improve the training efficiency and stability of the training model. The algorithm was tested in an open source 3D car racing simulator called TORCS. The simulation results demonstrate the training stability is significantly improved compared with the DDPG algorithm and the DQN algorithm, and the average return is about 46% higher than the DDPG algorithm and about 55% higher than the DQN algorithm.

For intelligent vehicles, a robust perception system relies on training datasets with a large variety of scenes. The architecture of federated learning allows for efficient collaborative model iteration while ensuring privacy and security by leveraging data from multiple parties. However, the local data from different participants is often not independent and identically distributed, significantly affecting the training effectiveness of autonomous driving perception models in the context of federated learning. Unlike the well-studied issues of label distribution discrepancies in previous work, we focus on the challenges posed by scene heterogeneity in the context of federated learning for intelligent vehicles and the inadequacy of a single scene for training multi-task perception models. In this paper, we propose a federated learning-based perception model training system.

In the process of designing the aerodynamic kit for Formula SAE racing cars, there is a lot of repetitive work and low efficiency in optimizing parameters such as wing angle of attack and chord length. Moreover, the optimization of these parameters in past designs heavily relied on design experience and it's difficult to achieve the optimal solution through theoretical calculations. By establishing a parametric model in CAD software and integrating it with CFD software, we can automatically modify model parameters, run a large number of simulations, and analyze the simulation results using statistical methods. After multiple iterations, we achieve fully automatic parameter optimization and obtain higher negative lift. At the same time, the simulation process is optimized, and simulations are run based on GPUs, resulting in a significant increase in simulation speed compared to the original.