Search Results

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.

Flow channels in proton exchange membrane fuel cells (PEMFC) play an irreplaceable role, and flow channel design in bipolar plates is one of the most active research areas at present. The flow channel on the cathode side needs to discharge liquid water out of the fuel cell in time and allow oxygen to flow to the cathode catalytic layer as much as possible to avoid the phenomenon of cathode water flooding and mass transfer loss. In order to improve the performance of PEMFC, a method of setting both side blockages in the cathode flow channel is proposed. In this paper, lateral blockage models with three shapes are proposed to study the influence of blockage on mass transfer and performance. First, a 3D PEMFC model with a middle channel was built to calculate the fuel cell power at different discharge rates.

Compared with urban areas, the road surface in mountainous areas generally has a larger slope, larger curvature and narrower width, and the vehicle may roll over and other dangers on such a road. In the case of limited driver information, if the two cars on the mountain road approach fast, it is very likely to occur road blockage or even collision. Multi-vehicle cooperative control technology can integrate the driving data of nearby vehicles, expand the perception range of vehicles, assist driving through multi-objective optimization algorithm, and improve the driving safety and traffic system reliability. Most existing studies on cooperative control of multiple vehicles is mainly focused on urban areas with stable environment, while ignoring complex conditions in mountainous areas and the influence of driver status. In this study, a digital twin based multi-vehicle cooperative warning system was proposed to improve the safety of multiple vehicles on mountain roads.

Autonomous driving technology represents a significant direction for future transportation, encompassing four key aspects: perception, planning, decision-making, and control. Among these aspects, vehicle trajectory planning and control are crucial for achieving safe and efficient autonomous driving. This paper introduces a Combined Model Predictive Control algorithm aimed at ensuring collision-free and comfortable driving while adhering to appropriate lane trajectories. Due to the algorithm is divided into two layers, it is also called the Bi-Level Model Predictive Control algorithm (BLMPC). The BLMPC algorithm comprises two layers. The upper-level trajectory planner, to reduce planning time, employs a point mass model that neglects the vehicle's physical dimensions as the planning model. Additionally, obstacle avoidance cost functions are integrated into the planning process.

The active sound generation systems (ASGS) for electric vehicles (EVs) play an important role in improving sound perception and transmission in the car, and can meet the needs of different user groups for driving and riding experiences. The active sound synthesis algorithm is the core part of ASGS. This paper uses an efficient variable-range fast linear interpolation method to design a frequency-shifted and pitch-modified sound synthesis algorithm. By obtaining the operating parameters of EVs, such as vehicle speed, motor speed, pedal opening, etc., the original sound signal is interpolated to varying degrees to change the frequency of the sound signal, and then the amplitude of the sound signal is determined according to different driving states. This simulates an effect similar to the sound of a traditional car engine. Then, a dynamic superposition strategy is proposed based on the Hann window function.

Autonomous vehicles require the collaborative operation of multiple modules during their journey, and enhancing tracking performance is a key focus in the field of planning and control. To address this challenge, we propose a cooperative control strategy, which is designed based on the integration of model predictive control (MPC) and a dual proportional–integral–derivative approach, referred to as collaborative control of MPC and double PID (CMDP for short in this article).The CMDP controller accomplishes the execution of actions based on information from perception and planning modules. For lateral control, the MPC algorithm is employed, transforming the MPC’s optimal problem into a standard quadratic programming problem. Simultaneously, a fuzzy control is designed to achieve adaptive changes in the constraint values for steering angles.

The main objective of platoon control is coordinated motion of autonomous vehicle platooning with small intervehicle spacing while maintaining the same speed and acceleration as the leading vehicle, which can save energy consumption and improve traffic throughput. The conventional platoon control methods are confronted with the problem of manual parameter tuning. In order to addres this isue, a novel bifold platoon control approach leveraging a deep reinforcement learning-based model is proposed, which enables the platoon adapt to the complex traffic environment, and guarantees the safety of platoon. The upper layer controller based on the TD3 tuned PID algorithm outputs the desired acceleration. This integration mitigates the inconvenience of frequent manual parameter tuning asociated with the conventional PID algorithm. The lower layer controller tracks the desired acceleration based on the inverse vehicle dynamics model and feedback control.

Accurately predicting the future trajectories of surrounding traffic agents is important for ensuring the safety of autonomous vehicles. To address the scenario of frequent interactions among traffic agents in the highway merging area, this paper proposes a trajectory prediction method based on interactive graph attention mechanism. Our approach integrates an interactive graph model to capture the complex interactions among traffic agents as well as the interactions between these agents and the contextual map of the highway merging area. By leveraging this interactive graph model, we establish an agent-agent interactive graph and an agent-map interactive graph. Moreover, we employ Graph Attention Network (GAT) to extract spatial interactions among trajectories, enhancing our predictions. To capture temporal dependencies within trajectories, we employ a Transformer-based multi-head self-attention mechanism.

Decision-making of lane-change for autonomous vehicles faces challenges due to the behavioral differences among human drivers in dynamic traffic environments. To enhance the performances of autonomous vehicles, this paper proposes a game theoretic decision-making method that considers the diverse Social Value Orientations (SVO) of drivers. To begin with, trajectory features are extracted from the NGSIM dataset, followed by the application of Inverse Reinforcement Learning (IRL) to determine the reward preferences exhibited by drivers with distinct Social Value Orientation (SVO) during their decision-making process. Subsequently, a reward function is formulated, considering the factors of safety, efficiency, and comfort. To tackle the challenges associated with interaction, a Stackelberg game model is employed.

The phenomenon of liquid transfer in the liquid tank of the semi-trailer vehicle for transporting dangerous cargo (SVTDC) during braking is analyzed and the relevant mathematical model is established. The braking dynamic model of the SVTDC considering the liquid sloshing in the tank is established, and the model is verified based on the co-simulation method. Based on the typical conditions, the braking deceleration and axle load calculation functions of the model are simulated and analyzed, and the application prospect of the model in the development of driving automation control strategy is discussed.

During the drivability test process, a large amount of noise generated by a series of internal and external factors of the vehicle reduces the accuracy of the drivability evaluation. To solve this problem, this paper introduces the EEMD denoising method and compares the denoising effects of the EMD denoising method and EEMD denoising method on the original signal using the entropy weight evaluation index. In addition, the optimal parameter setting is obtained by comparing the denoising results of different parameter settings in the EEMD denoising method. The results show that when the white noise is integrated 3000 times and the standard deviation of white noise is 0.1, the EEMD noise reduction method is the best, and the comprehensive score of noise reduction is 0.732 points higher than that of EMD.

This paper presents a low-speed assisted steering control approach for distributed drive electric vehicles. When the vehicle is driven at low speed, the braking of the inner-rear wheel is combined with differential drive to reduce the turning radius. A hierarchical control structure has been designed to achieve comprehensive control. The upper-level controller tracks the expected yaw rate and vehicle side-slip angle through a Linear Quadratic Regulator (LQR) algorithm. The desired yaw rate and vehicle side-slip angle are obtained according to the reference vehicle model, which can be regulated by the driver through the accelerator pedal. The lower-level controller uses a quadratic programming algorithm to distribute the yaw moment and driving moment to each wheel, aiming to minimize tire load rate variance.

Because of the long driving range and good power performance, plug-in hybrid electric vehicles (PHEV) have drawn much attention. And the current fuel-saving effect of PHEV still has a lot of room for improvement. The complex powertrain structure of PHEV makes the requirements for control strategy be increasing. Therefore, it is crucial to develop an efficient control strategy to ensure that the PHEV operates at optimal performance with an improved driving range. This paper establishes a mathematical model for fuel economy control of PHEV, by treating the torque distribution problem as a multi-stage decision optimization problem, and establishes a global energy management strategy based on a dynamic programming (DP) algorithm. Based on the actual physical model, this paper creatively solves the correction range of battery SOC value according to the charge and discharge power of the motor, which greatly reduces the calculation time of the DP algorithm.

With the development of highway transportation and automobile industry technology, highway truck overload phenomenon occurs frequently, which poses a danger to road safety and personnel life safety. So it is very important to identify the overload phenomenon. Traditionally, static detection is adopted for overload identification, which has low efficiency. Aiming at this phenomenon, a dynamic overload identification method is proposed. Firstly, the coupled road excitation model of vehicle speed and speed bump is established, and then the 4-DOF vehicle model of half car is established. At the same time, considering that the double input vibration of the front and rear wheels will be coupled when vehicle passes through the speed bump, the model is decoupled. Then, the vertical trajectory of the body in the front axle position is obtained by Carsim software simulation.

At present, the development of autonomous driving technology is still immature, and there is still a long way until fully driverless vehicles. Therefore, the state of the driver is still an important factor affecting traffic safety, and it is of great significance to detect the driver’s distracted behavior. In the task of driver distracted behavior detection, some characteristics of driver behavior in the cockpit can be further utilized to improve the detection performance. Compared with general human behaviors, driving behaviors are confined to enclosed space and are far less diverse. With this in mind, we propose a sparse spatiotemporal transformer which extracts local spatiotemporal features by segmenting the video at the low level of the model, and filters out local key spatiotemporal information associated with larger attention values based on the attention map in the middle layer, so as to enhance the high-level global semantic features.

Hydrogen recirculation system (HRS) is one of the main subsystems of proton exchange membrane fuel cell (PEMFC) system, and an HRS with high performance and low cost is essential to improve the fuel cell lifetime and efficiency. The ejector is becoming an effective alternative to the hydrogen circulation pump in the hydrogen recirculation system because of its small size, low cost and no parasitic power. However, the conventional ejector can only operate at the limited output power of the fuel cell. In order to improve this drawback, a hydrogen recirculation system with a separator-ejector integrated device is proposed. The hydrogen recirculation system based on the integrated device can make the space more compact, reduce the condensation effect of the piping, improve the fuel cell performance and reduce the cost. The design and integration method of the ejector and gas-water separator in this hydrogen recirculation system are introduced.

This paper proposed a model predictive control(MPC) based car-following control strategy for electric vehicles considering comfort, in order to improve the comfort of the car-following control system of electric vehicles. The MPC algorithm is improved in the following three aspects to improve the comfort: Firstly, a five-state longitudinal car-following model is adopted, so that the MPC algorithm can optimize the acceleration and acceleration change rate of the ego vehicle. Secondly, for the weight coefficients of the output vector and the input vector of the objective function, the fixed weight coefficients are changed into variable weight coefficients by the way of Nash equilibrium game, so that the control system can improve the weight of the parameters used to control the comfort under suitable driving conditions.

Regenerative braking is an effective way to increase the cruising range of vehicles. In commercial vehicles with large vehicle mass, regenerative braking can be maintained in a high-power working state for a long time theoretically because of the large braking torque and long braking time. But in fact, it is often impossible to run at full power because of battery safety problems. In this paper, a control strategy is designed to maintain the maximum power operation of regenerative braking as much as possible. The maximum charging power of the battery is obtained through the battery model, and it is set as the battery limiting parameter. The regenerative braking torque and power are obtained by using the motor model. The eddy current retarder is used to absorb the excess power that the battery can't bear, and the braking torque of the eddy current retarder is calculated. Finally, mechanical braking is used to make up the insufficient braking torque.

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.

With the improvement of sensor accuracy, sensor data plays an increasingly important role in intelligent vehicle motion control. Good use of sensor data can improve the control of vehicles. However, data-based end-to-end control has the disadvantages of poorly interpreted control models and high time costs; model-based control methods often have difficulties designing high-fidelity vehicle controllers because of model errors and uncertainties in building vehicle dynamics models. In the face of high-speed steering conditions, vehicle control is difficult to ensure stability and safety. Therefore, this paper proposes a hybrid model and data-driven control method. Based on the vehicle state data and road information data provided by vehicle sensors, the method constructs a deep neural network based on LSTM and Attention, which is used as a compensator to solve the performance degradation of the LQR controller due to modeling errors.