Search Results

Multi-sensor information fusion framework is the eyes for unmanned driving and Advanced Driver Assistance System (ADAS) to perceive the surrounding environment. In addition to the perception of the surrounding environment, real-time vehicle localization is also the key and difficult point of unmanned driving technology. The disappearance of high-precision GPS signal suddenly and defect of the lane line will bring much more difficult and dangerous for vehicle localization when the vehicle is on unmanned driving. In this paper, a road boundary feature extraction algorithm is proposed based on multi-sensor information fusion of automotive radar and vision to realize the auxiliary localization of vehicles. Firstly, we designed a 79GHz (78-81GHz) Ultra-Wide Band (UWB) millimeter-wave radar, which can obtain the point cloud information of road boundary features such as guardrail or green belt and so on.

Thin-walled structures have been widely used in automobile body design because of its good lightweight and superior mechanical properties. For the energy-absorbing box of the automobile, it is necessary to consider its working conditions under the axial and oblique impact. In this paper, a novel hierarchical honeycomb is proposed and used as filler for thin-walled structures. Meanwhile, the crashworthiness performances of the conventional honeycomb-filled and the hierarchical honeycomb-filled thin-walled structures under different impact conditions are systematically studied. The results indicate the energy absorption of the hierarchical honeycomb-filled thin-walled structure is higher than that of the conventional honeycomb-filled thin-walled structure, and the impact angle has significant effects on the energy absorption performance of the hierarchical honeycomb-filled structure.

The radar-based advanced driver assistance systems (ADAS) like autonomous emergency braking (AEB) and forward collision warning (FCW) can reduce accidents, so as to make vehicles, drivers and pedestrians safer. For active safety, automotive millimeter-wave radar is an indispensable role in the automotive environmental sensing system since it can work effectively regardless of the bad weather while the camera fails. One crucial task of the automotive radar is to detect and distinguish some objects close to each other precisely with the increasingly complex of the road condition. Nowadays almost all the automotive radar products work in bidimensional area where just the range and azimuth can be measured. However, sometimes in their field of view it is not easy for them to differentiate some objects, like the car, the manhole covers and the guide board, when they align with each other in vertical direction.

Electric Power Steering (EPS) is the actuator of several lateral-dynamic-related Advanced Driver Assistance Systems (ADAS). A driving simulator with EPS will be much helpful for the ADAS development. However, if a real EPS is used in the driving simulator, it is quite difficult to realize the road reaction force accurately and responsively. To overcome this weakness, a virtual EPS platform is established. The virtual EPS platform contains two parts: one is the vehicle and EPS model, the other is the force feedback actuator (FFA) of the Steer-by-Wire (SBW) system. The FFA is an interface between the driver and the EPS/vehicle model. The reactive torque of the FFA is obtained based on the models. Meanwhile, the input of the EPS model is the steering angle of the FFA. Comparing to a real EPS, the virtual EPS platform has a problem of instability because of the actuator lag of the FFA. Therefore, a damping control method is applied to make the system stable.

In the recent years, the interaction between human driver and Advanced Driver Assistance System (ADAS) has gradually aroused people’s concern. As a result, the concept of personalized ADAS is being put forward. As an important system of ADAS, Lane Keeping Assistance System (LKAS) also attracts great attention. To achieve personalized LKAS, driver lane keeping characteristic (DLKC) indices which could distinguish different driver lane keeping behavior should be researched. However, there are few researches on DLKC indices for personalized LKAS. Although there are many researches on modeling driver steering behavior, these researches are not sufficient to obtain DLKC indices. One reason is that most of researches are for double lane change behavior which is different from driver lane keeping behavior. The other reason is that the researches on driver lane keeping behavior only provide model structure and rarely discuss identification procedure such as how to select suitable data.

In China there are many mixed driving roads which cause a lot of safety problems between vehicles and pedalcyclists. Research on driver behavior under risk scenarios with pedalcyclist is relatively few. In this paper driver brake parameters under naturalistic driving are studied and pedalcyclists include bicyclist, tricyclist, electric bicyclist and motorcyclist. Brake reaction time and maximum brake jerk are used to evaluate driver brake reaction speed. Average deceleration is used to evaluate the effect of driver brake operation. Maximum deceleration is used to evaluate driver braking ability. Driver behaviors collected in China are classified and risk scenarios with pedalcyclist are obtained. Driver brake parameters are extracted and statistical characteristics of driver brake parameters are obtained. Influence factors are analyzed with univariate ANOVA and regression analysis.

Faced with intricate traffic conditions, the single sensor has been unable to meet the safety requirements of Advanced Driver Assistance Systems (ADAS) and autonomous driving. In the field of multi-target tracking, the number of targets detected by vision sensor is sometimes less than the current tracks while the number of targets detected by millimeter wave radar is more than the current tracks. Hence, a multi-sensor information fusion algorithm is presented by utilizing advantage of both vision sensor and millimeter wave radar. The multi-sensor fusion algorithm is based on centralized fusion strategy that the fusion center takes a unified track management. At First, vision sensor and radar are used to detect the target and to measure the range and the azimuth angle of the target. Then, the detections data from vision sensor and radar is transferred to fusion center to join the multi-target tracking with the prediction of current tracks.

With the development of autonomous driving, the penetration rate of autonomous vehicles on the road will continue to grow. As a result, the social cooperation ability of autonomous vehicles will have a great effect on the social acceptance of autonomous driving, which can be described as harmony with traffic. In order to research the evaluation method of the harmony with traffic, this paper proposes a subjective and objective mapping evaluation method based on the Mean Impact Value and Backpropagation (MIV-BP) Neural Network, with the merging vehicle on the expressway ramp as the research object. Firstly, by taking 16 original objective indexes obtained by theoretical analysis and the subjective evaluation results as input and output, respectively, the BP Neural Network model is constructed as a baseline model. Secondly, nine selected objective indexes are selected by the MIV method based on the baseline model.

Most of the existing energy management strategies for Extended-Range Electric Vehicles (E-REVs) are heuristic, which restricts coordination between the battery and the Range Extender. This paper presents an instantaneous optimization energy management strategy based on the Minimum Loss Power Algorithm (MLPA) for a fuel cell E-REV. An instantaneous loss power function of power train system is constructed by considering the charge and discharge efficiency of the battery, together with the working efficiency of the fuel cell Range Extender. The battery working mode and operating points of the fuel cell Range Extender are decided by an instantaneous optimization module (an artificial neural network) that aims to minimize the loss power function at each time step.

The prediction of passenger flow is of great significance to facilitate the decision-making processes for local authorities and transport operators to provide an effective bus scheduling. In this work, a backpropagation neural network (BPNN) was adopted to predict the bus passenger flow. To reduce the prediction error and improve the prediction accuracy, a combined model CEEMDAN-BP, which combines CEEMDAN (Complete Ensemble Empirical Mode Decomposition with Adaptive Noise) method and BPNN, has been proposed. CEEMDAN is an improved method based on EEMD, which has been widely applied to signal smoothing and de-noising. Experimental results show that this combined model can exactly achieve an excellent prediction effect and improve the prediction accuracy of the network greatly.

Cut-in scenario is common in traffic and has potential collision risk. Human driver can detect other vehicle’s cut-in intention and take appropriate maneuvers to reduce collision risk. However, autonomous driving systems don’t have as good performance as human driver. Hence a deeper understanding on driving behavior is necessary. How to make decisions like human driver is an important problem for automated vehicles. In this paper, a method is proposed to classify the dangerous cut-in situations and normal ones. Dangerous cases were extracted automatically from naturalistic driving database using specific detection criteria. Among those cases, 70 valid dangerous cut-in cases were selected manually. The largest deceleration of subject vehicle is over 4 m/s2. Besides, 249 normal cut-in cases were extracted by going through video data of 2000km traveled distance. In normal driving cases, subject vehicle may brake or keep accelerating and the largest deceleration was less than 3 m/s2.

Durability performance prediction is a critical issue in fuel cell research. During the demonstration operation of fuel cell commercial vehicles in China, this issue has attracted more attention. In this article, the long short-term memory neural network (LSTMNN), which is an improved recurrent neural network (RNN), and the demonstration operation data are used to establish the prediction model to predict the durability performance of the fuel cell stack. Then, a model based on a back-propagation neural network (BPNN) is established to be a control group. The demonstration operation data is divided into training group and validation group. The former is used to train the prediction model, and the latter is used to verify the validity and accuracy of the prediction model. The outputs of the prediction model, as the durability performance evaluation indexes of the fuel cell, are the polarization curve (current-voltage curve) and the voltage decay curve (time-voltage curve).

As turbulence modeling has become an indispensable approach to perform flow simulation in a wide range of industrial applications, how to enhance the prediction accuracy has gained increasing attention during the past years. Of all the turbulence models, RANS is the most common choice for many OEMs due to its short turn-around time and strong robustness. However, the default setting of RANS is usually benchmarked through classical and well-studied engineering examples, not always suitable for resolving complex flows in specific circumstances. Many previous researches have suggested a small tuning in turbulence model coefficients could achieve higher accuracy on a variety of flow scenarios. Instead of adjusting parameters by trial and error from experience, this paper introduced a new data-driven method of turbulence model recalibration using adjoint solver, based on Generalized k-ω (GEKO) model, one variant of RANS.

Optimization design of hard point parameters for hinge mechanism has been paid more attention in recent years, attributable to their significant improvement in dynamic performance. In this paper, the experimental analysis and dynamic optimization design of hinge mechanism is performed. The acceleration measurement experiments are carried out at different arrangement points and under different working conditions. Furthermore, the accuracy of established multi-body dynamics model is verified by three-axis accelerometer measurement experiment. In addition, sensitivity analysis for electric strut and gas strut coordinates is performed and shows that the Y coordinate of the lower end point of the electric strut is the design variable that has the greatest impact on the responses.

To improve the durability of Proton-exchange membrane fuel cell (PEMFC) in actual transportation application scenario, the research on fault diagnosis of PEMFC is receiving extensive attention. With the development of artificial intelligence, performing fault diagnosis with the massive sampling data of the fuel cell system has become a popular research topic. But few people have successfully verified the diagnosis performance of these artificial intelligence algorithms on a real high power on-board PEMFC system. Therefore, we intend to make a step forward with these data-driven artificial intelligence algorithms. We applied four data-driven artificial intelligence algorithms to diagnose three common faults of PEMFC (each fault type has two severity levels, slight and severe). AVL CRUISE M was firstly applied for generation of simulation fault dataset to speed up the algorithm screening process. Based on the dataset, these algorithms are trained and optimized.

In recent years, fuel cell vehicles have attracted more attention since the advantages of no environmental pollution and high energy density, however, the cost and durability of fuel cells have been important factors limiting the rapid development of fuel cell vehicles. How to quickly predict the life of fuel cells has always been the emphasis and focus of the industry. Therefore, this paper mainly focuses on two sets of proton exchange membrane hydrogen fuel cell durability test data. In this paper, we establish a fuel cell life prediction model to carry out product prediction research, using Gated Recurrent Unit Neural Network (GRU-NN)—a variant of “Recurrent Neural Networks” (RNN). This article first divides the two sets of fuel cell durability test data into a training group and a verification group and trains the established neural network model with the test data of the training group.

Intrusion Detection Systems (IDS), technically speaking, is to monitor the network, system, and operation status according to certain security policies, and try to find various attack attempts, attacks or attack results to ensure the confidentiality, integrity and availability of network system resources. Automotive intrusion detection systems can identify and alert by analyzing in-vehicle traffic and log when software applications or devices with malicious activity exist, or the in-vehicle network is tampered and injected. But unfortunately, automotive cybersecurity researchers hardly produce a comprehensive detection method due to the confidential nature of Controller Area Network (CAN) DBC format files, which is a standard long maintained by car manufacturers. In this paper, an enhanced intrusion detection method is proposed based on the double-decision-tree to classify different attack models for in-vehicle CAN network without the need to obtain complete DBC files.

High-speed vehicles in low illumination environments severely blur the images used in object detectors, which poses a potential threat to object detector-based advanced driver assistance systems (ADAS) and autonomous driving systems. Augmenting the training images for object detectors is an efficient way to mitigate the threat from motion blur. However, little attention has been paid to the motion of the vehicle and the position of objects in the traffic scene, which limits the consistence between the resulting augmented images and traffic scenes. In this paper, we present a vehicle kinematics-based image augmentation algorithm by modeling and analyzing the traffic scenes to generate more realistic augmented images and achieve higher robustness improvement on object detectors against motion blur. Firstly, we propose a traffic scene model considering vehicle motion and the relationship between the vehicle and the object in the traffic scene.

Image corruptions due to noise, blur, contrast change, etc., could lead to a significant performance decline of Deep Neural Networks (DNN), which poses a potential threat to DNN-based autonomous vehicles. Previous works attempted to explain corruption from a Fourier perspective. By comparing the absolute Fourier spectrum difference between corrupted images and clean images in the RGB color space, they regard the noise from some corruptions (Gaussian noise, defocus blur, etc.) as concentrating on the high-frequency components while others (contrast, fog, etc.) concentrate on the low-frequency components. In this work, we present a new perspective that unifies corruptions as noise from high frequency and thus propose an image augmentation algorithm to achieve a more robust performance against common corruptions. First, we notice the 1/fα statistical rule of the natural image's spectrum and the channels-wise differential sensitivity on the YCbCr color space of the Human Visual System.

Active engine mount (AEM) is an effective approach which can optimize the noise, vibration and harshness (NVH) performance of vehicles. The filtered-x-least-mean-squares (FxLMS) algorithm is widely applicated for vibration attenuation in AEMs. However, the performance of FxLMS algorithm can be deteriorated without an accurate secondary path estimation. First, this paper models the secondary path using finite impulse response (FIR) model, infinite impulse response (IIR) model and back propagation (BP) neural network model and the model errors of which are compared to determine the most accurate and robust modeling method. After that, the influence of operation frequency on accuracy of the secondary path model is analyzed through simulation approach. Then, the impact of reference signal mismatch on the control effect is demonstrated to study the robustness of FxLMS algorithm.