Search Results

Driving style and skill classification are of great significance in human-oriented advanced driver-assistance system (ADAS) development. In this paper, we propose Fully-Connected Deep Neural Networks (FC-DNN) to classify drivers’ styles and skills with naturalistic driving data. Followed by the data collection and pre-processing, FC-DNN with a series of deep learning optimization algorithms are applied. In the experimental part, the proposed model is validated and compared with other commonly used supervised learning methods including the k-nearest neighbors (KNN), support vector machine (SVM), decision tree (DT), random forest (RF), and multilayer perceptron (MLP). The results show that the proposed model has a higher Macro F1 score than other methods. In addition, we discussed the effect of different time window sizes on experimental results. The results show that the driving information of 1s can improve the final evaluation score of the model.

The accuracy of tire forces directly affects the vehicle dynamics model precision and determines the ability of the model to develop the simulation platform or design the control strategy. In the high slip angle, due to the complex interactions at tire-road interfaces, the forces generated by the tires are high nonlinearity and uncertainty, which pose issues in calculating tire force accurately. This paper presents a hybrid physical and data-driven tire force calculation framework, which can satisfy the high nonlinearity and uncertainty condition, improve the model accuracy and effectively leverage prior knowledge of physical laws. The parameter identification for the physical tire model and the data-based compensation for the unknown errors between the physical tire model and actual tire force data are contained in this framework. First, the parameters in the selected combined-slip Burckhardt tire model are identified by the nonlinear least square method with tire test data.

An autonomous vehicle is a comprehensive intelligent system that includes environment sensing, vehicle localization, path planning and decision-making control, of which environment sensing technology is a prerequisite for realizing autonomous driving. In the early days, vehicles sensed the surrounding environment through sensors such as cameras, radar, and lidar. With the development of 5G technology and the Vehicle-to-everything (V2X), other information from the roadside can also be received by vehicles. Such as traffic jam ahead, construction road occupation, school area, current traffic density, crowd density, etc. Such information can help the autonomous driving system understand the current driving environment more clearly. Vehicles are no longer limited to areas that can be sensed by sensors. Vehicles with different autonomous driving levels have different adaptability to the environment.

The RON(Research Octane Number) is the most important indicator of motor petrol, and the petrol refining process is one of the important links in petrol production. However, RON is often lost during petrol refining and RON Loss means the value of RON lost during petrol refining. The prediction of the RON loss of petrol during the refining process is helpful to the improvement of petrol refining process and the processing of petrol. The traditional RON prediction method relied on physical and chemical properties, and did not fully consider the high nonlinearity and strong coupling relationship of the petrol refining process. There is a lack of data-driven RON loss models. This paper studies the construction of the RON loss model in the petrol refining process.

Velocity prediction on hilly road can be applied to the energy-saving predictive control of intelligent vehicles. However, the existing methods do not deeply analyze the difference and diversity of road slope driving characteristics, which affects prediction performance of some prediction method. To further improve the prediction performance on road slope, and different road slope driving features are fully exploited and integrated with the common prediction method. A rolling prediction-based multi-scale fusion prediction considering road slope transition driving characteristics is proposed in this study. Amounts of driving data in hilly sections were collected by the advanced technology and equipment. The Markov chain model was used to construct the velocity and acceleration joint state transition characteristics under each road slope transition pair, which expresses the obvious driving difference characteristics when the road slope changes.

Formula One (F1) racing cars are often running at high-speed with the pitching angle changing frequently due to road conditions. These pitching angle changes result in changes to the car’s aerodynamic characteristics that will directly affect handling stability and other performance factors including safety. This paper takes a F1 racing car as the model; the influence of the change of pitching angle on aerodynamic drag force and lift force are investigated. CFD code-PowerFLOW based LBM is used to simulate the aerodynamic characteristics with different pitching angles. The distribution of aerodynamic coefficients, velocity and pressure in the flow field are obtained; and the differences between different pitching angles were analyzed. The results show that as the pitching angle increases, the drag force increases and the lift force decreases. The down-force of the car is mainly supplied by the front wing and the rear wing.

Numerical simulations on the fluid-structure interaction were conducted using commercial software STAR-CCM+ and ABAQUS. The aeroelastic responses of a deflector under several different working conditions were simulated utilizing finite volume and finite element methods to investigate the aeroelastic problem of automotive deflectors. Results showed that the structural response of a top deflector is minimal under the influence of aerodynamics given its large structural stiffness. The size of the top deflector was optimised by using thickness as a variable. The volume and quality of the top deflector were significantly reduced, and its lightweight performance was improved to satisfy the stiffness performance requirement. The vibration of a side deflector structure was mainly induced by the turbulence on the structure surface. The amplitude of vibration was small and the vibration gradually converged in a few seconds without obvious regularity.

Vehicle lane keeping system is becoming a new research focus of drive assistant system except adaptive cruise control system. As we all known, vehicle lateral dynamics show strong nonlinear and time-varying with the variety of longitudinal velocity, especially tire’s mechanics characteristic will change from linear characteristic under low speed to strong nonlinear under high speed. For this reason, the traditional PID controller and even self-tuning PID controller, which need to know a precise vehicle lateral dynamics model to adjust the control parameter, are too difficult to get enough accuracy and the ideal control quality. Based on neural network’s ability of self-learning, adaptive and approximate to any nonlinear function, an adaptive PID control algorithm with BP neural network self-tuning online was proposed for vehicle lane keeping.

Intelligent driving, aimed for collision avoidance and self-navigation, is mainly based on environmental sensing via radar, lidar and/or camera. While each of the sensors has its own unique pros and cons, camera is especially good at object detection, recognition and tracking. However, unpredictable environmental illumination can potentially cause misdetection or false detection. To investigate the influence of illumination conditions on detection algorithms, we reproduced various illumination intensities in a photo-realistic virtual world, which leverages recent progress in computer graphics, and verified vehicle detection effect there. In the virtual world, the environmental illumination is controlled precisely from low to high to simulate different illumination conditions in the driving scenarios (with relative luminous intensity from 0.01 to 400). Sedan cars with different colors are modelled in the virtual world and used for detection task.

Driving assistance system is regarded as an effective method to improve driving safety and comfort and is widely used in automobiles. However, due to the different driving styles of different drivers, their acceptance and comfort of driving assistance systems are also different, which greatly affects the driving experience. The key to solving the problem is to let the system understand the driving style and achieve humanization or personalization. This paper focuses on clustering and identification of different driving styles. In this paper, based on the driver's real vehicle experiment, a driving data acquisition platform was built, meanwhile driving conditions were set and drivers were recruited to collect driving information. In order to facilitate the identification of driving style, the correlation analysis of driving features is conducted and the principal component analysis method is used to reduce the dimension of driving features.

Tire-road friction condition is crucial to the safety of vehicle driving. The emergence of autonomous driving makes it more important to estimate the friction limit accurately and at the lowest possible excitation. In this paper, an early detection method of tire-road friction coefficient based on pneumatic trail under cornering conditions is proposed using an intelligent tire system. The previously developed intelligent tire system is based on a triaxial accelerometer mounted on the inner liner of the tire tread. The friction estimation scheme utilizes the highly sensitive nature of the pneumatic trail to the friction coefficient even in the linear region and its approximately linear relationship with the excitation level. An indicator referred as slip degree indicating the utilization of the road friction is proposed using the information of pneumatic trail, and it is used to decide whether the excitation is sufficient to adopt the friction coefficient estimate.

Driver's driving style has a great impact on lane changing behavior, especially in scenarios such as freeway on-ramps that contain a strong willingness to change lanes, both in terms of inter-vehicle interactions during lane changing and in terms of the driving styles of the two vehicles. This paper proposes a study on game-theoretic decision-making for lane-changing on highway on-ramps considering driving styles, aiming to facilitate safer and more efficient merging while adequately accounting for driving styles. Firstly, the six features proposed by the EXID dataset of lane-changing vehicles were subjected to Principal Component Analysis (PCA) and the three principal components after dimensionality reduction were extracted, and then clustered according to the principal components by the K-means algorithm. The parameters of lane-changing game payoffs are computed based on the clustering centers under several styles.

Unmanned Ground Vehicle (UGV) has a wide range of applications in the military, agriculture, firefighting and other fields. Path planning, as a key aspect of autonomous driving technology, plays an essential role for UGV to accomplish the established driving tasks. At present, there are many global path planning algorithms in grid maps on unstructured roads, while general grid maps do not consider the specific elevation or ground type difference of each grid, and unstructured roads are generally considered as flat and open roads. On the contrary, the unmanned off-road is always a bumpy road with undulating terrain, and meanwhile, the landform is complex and the types of features are diverse. In order to ensure the safety and improve the efficiency of autonomous driving of UGV in off-road environment, this paper proposes a global off-road path planning method for UGV based on the raw image of remote sensing map. Firstly, the raw image is gridded.

This paper addresses the issues of long-term signal loss in localization and cumulative drift in SLAM-based online mapping and localization in autonomous valet parking scenarios. A GPS, INS, and SLAM fusion localization framework is proposed, enabling centimeter-level localization with wide scene adaptability at multiple scales. The framework leverages the coupling of LiDAR and Inertial Measurement Unit (IMU) to create a point cloud map within the parking environment. The IMU pre-integration information is used to provide rough pose estimation for point cloud frames, and distortion correction, line and plane feature extraction are performed for pose estimation. The map is optimized and aligned with a global coordinate system during the mapping process, while a visual Bag-of-Words model is built to remove dynamic features.

Lane detection and tracking play a key role in autonomous driving, not only in the LKA System but help estimate the pose of the vehicle. While there has been significant development in recent years, traditional outdoor SLAM algorithms still struggle to provide reliable information in challenging dynamic environments such as lack of roadside landscape or surrounding vehicles at almost the same speed or on the road in the woods. On the structured road, lane markings as static semantic features may provide a stable landmark assist in robust localization. As most of the current lane detection work mainly on separated images ignoring the relationship between adjacent frames, we propose a pixel-level lane tracking method for autonomous vehicles. In this paper, we introduce a deep network to detect and track lane features. The network has two parallel branches. One branch detects the lane position, while the other extracts the point description on a pixel level.

Coordinated control of mode transition is an important part of the multi-mode hybrid vehicles' control strategy, combined with a vehicle torque distribution strategy to realize an optimal working condition of the power sources, as well as achieve smooth mode switching. This paper builds hybrid electric vehicle driveline dynamics model and depth analyzes drive mode transition process, coordinated control methods were provided to solve three types of mode switching, neural network algorithm was provided to estimate the engine torque. The results show that coordinated control can reduce torque fluctuations and decrease jerk during the transition of different modes to improve the vehicle drivability.

This paper studies a multi-objective optimization design of interior noise for an automotive body. An acoustic-structure coupled model with materials and properties was established to predict the interior noise based on a passenger car. Moreover, three kinds of approximation models related damping thickness and the root mean square of the driver’s ear sound pressure level were established through Latin hypercube method and the corresponding experiments. The prediction accuracy was analyzed and compared for the approximate response surface model, Kriging model and Radial Basis Function neural network model. On this basis, multi-objective optimization of the vehicle interior noise was conducted by using NSGA-II. According to the optimization results, the damping composite structure was applied on the car body structure. Then, the comparison of sound pressure level response at driver’s ear location before and after optimization was performed at speed of 60 km/h on a smooth road.

It’s hard to achieve complete self-driving using hand-crafting generalized decision-making rules, while the end-to-end self-driving system is low in complexity, does not require hand-crafting rules, and can deal with complex situations. Modular-based self-driving systems require multi-task fusion and high-precision maps, resulting in high system complexity and increased costs. In end-to-end self-driving, we usually only use camera to obtain scene status information, so image processing is very important. Numerous deep learning applications benefit from multi-task learning, as the multi-task learning can accelerate model training and improve accuracy with combine all tasks into one model, which reduces the amount of calculation and allows these systems to run in real-time. Therefore, the approach of obtaining rich scene state information based on multi-task learning is very attractive. In this paper, we propose an approach to multi-task learning for semantics, geometry and motion.

In this research, the interior noise of a passenger car was measured, and the sound quality metrics including sound pressure level, loudness, sharpness, and roughness were calculated. An artificial neural network was designed to successfully apply on automotive interior noise as well as numerous different fields of technology which aim to overcome difficulties of experimentations and save cost, time and workforce. Sound pressure level, loudness, sharpness, and roughness were estimated by using the artificial neural network designed by using the experiment values. The predicted values and experiment results are compared. The comparison results show that the realized artificial intelligence model is an appropriate model to estimate the sound quality of the automotive interior noise. The reliability value is calculated as 0.9995 by using statistical analysis.

In order to solve the abnormal vibration of a light bus, order tracking analysis of finite element simulation and road test was made to identify the vibration source, finding that the rotation angular frequency of the wheels and the first two natural frequency of the body structure overlaps, resonance occurring which lead to increased vibration. To stagger the first two natural frequency and excitation frequency of the body, thickness of sheet metal and skeleton of the body-in-white were chosen as the design variables, rise of the first two natural frequency of the body-in-white as the optimization objective, optimal design and sensitivity analysis of the body-in-white was carried out with the modal analysis theory. Combining with the modal sensitivity and mass sensitivity of sheet metal and skeleton, the optimum design was achieved and tests analysis was conducted.