Search Results

Roundabouts are one of the most complex traffic environments in urban roads, and a key challenge for intelligent driving decision-making. Deep reinforcement learning, as an emerging solution for intelligent driving decisions, has the advantage of avoiding complex algorithm design and sustainable iteration. For the decision difficulty in roundabout scenarios, this paper proposes an artificial potential field based Soft Actor-Critic (APF-SAC) algorithm. Firstly, based on the Carla simulator and Gym framework, a reinforcement learning simulation system for roundabout driving is built. Secondly, to reduce reinforcement learning exploration difficulty, global path planning and path smoothing algorithms are designed to generate and optimize the path to guide the agent.

Autonomous driving technology is more and more important nowadays, it has been changing the living style of our society. As for autonomous driving planning and control, vehicle dynamics has strong nonlinearity and uncertainty, so vehicle dynamics and control is one of the most challenging parts. At present, many kinds of specific vehicle dynamics models have been proposed, this review attempts to give an overview of the state of the art of vehicle dynamics models for autonomous driving. Firstly, this review starts from the simple geometric model, vehicle kinematics model, dynamic bicycle model, double-track vehicle model and multi degree of freedom (DOF) dynamics model, and discusses the specific use of these classical models for autonomous driving state estimation, trajectory prediction, motion planning, motion control and so on.

To address the issue of PID control for automotive vibration, this paper supplements and develops the evaluation of automotive vibration characteristics, and proposes a vibration response quantity for evaluating the energy dissipation characteristics of automotive vibration. A two-degree-of-freedom single wheel model for automotive vibration control is established, and the conventional vibration response variables for ride comfort evaluation and the energy consumption vibration response variables for energy dissipation characteristics evaluation are determined. This paper uses the Adaptive Differential Evolution (ADE) algorithm to tune the PID control parameters and introduces an adaptive mutation factor to improve the algorithm's adaptability. Several commonly used adaptive mutation factors are summarized in this paper, and their effects on algorithm improvement are compared.

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.

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.

Active control of low-frequency engine order noise helps to improve the passenger’s sense of hearing, so it has become one of the hot topics in the automotive field. Depth improvement of active noise control (ANC) performance from the perspective of novel algorithms has attracted the attention of researchers. The conventional notch dual-channel filtered-x least mean square (NDFxLMS) algorithm shows acceptable noise reduction for the elimination of engine order noise. To further enhance the steady-state ANC effect, this paper proposed two new notch algorithms: the notch dual-channel filtered-x recursive least square (NDFxRLS) algorithm and the notch dual-channel affine projection (NDAP) algorithm. Vehicle simulation tests show that both the proposed algorithms, especially the NDFxRLS algorithm, have a satisfying performance for the cancellation of interior noise from the engine.

To meet the new requirements of braking system for modern electrified and intelligent vehicles, various novel electro-mechanical brake boosters (Eboosters) are emerging. This paper is aimed at a new type of the Ebooster, which is mainly consisted of a permanent magnet synchronous motor (PMSM), a two-stage reduction transmission and a servo mechanism. Among them, the PMSM is a vital actuator to realize the functions of the Ebooster. To get fast response of the Ebooster system, a novel control strategy employing a maximum torque per ampere (MTPA) control with current compensation decoupling and current-adjusting adaptive flux-weakening control is proposed, which requires the PMSM can operate in a large speed range and maintain a certain anti-load interference capability. Firstly, the wide speed control strategy for the Ebooster’s PMSM is designed in MATLAB/Simulink.

The driving behaviors of young drivers under the influence of anger are analyzed by driving simulator in this paper. A total of 12 subjects are enrolled during the experiment. Standardized videos are utilized to induce the driver's anger emotion. And the driver's electrocardiogram (ECG) signal is collected synchronously and compared before and after emotional trigger, which prove the validity of emotional trigger. Based on the result, the driver's driving performance under the straight road and the curve under normal state and angry state are compared and analyzed. The results of independent sample t-test show that there are significant differences in the running time of straight sections and the standard deviation of steering wheel angle in curves between normal and angry states. In conclusion, the longitudinal and lateral operation of drivers is unstable in angry state and the driver will be more destructive to the regular driving behavior.

Performance testing and evaluation always plays an important role in the developmental process of a vehicle, which also applies to autonomous vehicles. The complex nature of an autonomous vehicle from architecture to functionality demands even more quality-and-quantity controlled testing and evaluation than ever before. Most of the existing testing methodologies are task-or-scenario based and can only support single or partial functional testing. These approaches may be helpful at the initial stage of autonomous vehicle development. However, as the integrated autonomous system gets mature, these approaches fall short of supporting comprehensive performance evaluation. This paper proposes a novel hierarchical and systematic testing and evaluation approach to bridge the above-mentioned gap.

As one of advanced driver assistance systems (ADAS), automatic parking system has great market prospect and application value. In this paper, based on an intelligent vehicle platform, an automatic parking system is designed by using multi-point preview theory. The vehicle kinematics model was established, based on Ackermann steering principle. By analyzing working conditions of parallel parking, complex constraint condition of parking trajectory is established and reference trajectory based on sine wave is proposed. In addition, combined with multi-point preview theory, the design of trajectory following controller for automatic parking is completed. The cost function is designed, which consider the trajectory following effect and the degree of easy handling. The optimization of trajectory following control is completed by using the cost function.

Although advanced driver assistance systems (ADAS) have been widely introduced in automotive industry to enhance driving safety and comfort, and to reduce drivers’ driving burden, they do not in general reflect different drivers’ driving styles or customized with individual personalities. This can be important to comfort and enjoyable driving experience, and to improved market acceptance. However, it is challenging to understand and further identify drivers’ driving styles due to large number and great variations of driving population. Previous research has mainly adopted physical approaches in modeling drivers’ driving behavior, which however are often very much limited, if not impossible, in capturing human drivers’ driving characteristics. This paper proposes a reinforcement learning based approach, in which the driving styles are formulated through drivers’ learning processes from interaction with surrounding environment.

Since steering wheel torque feedback is one of the crucial factors for drivers to gain road feel and ensure driving safety, it is especially important to simulate the steering torque feedback for a driving simulator. At present, steering wheel feedback torque is mainly simulated by an electric motor with gear transmission. The torque response is typically slow, which can result in drivers’ discomfort and poor driving maneuverability. This paper presents a novel torque feedback device with magnetorheological (MR) fluid and coil spring. A phase separation control method is also proposed to control its feedback torque, including spring and damping torques respectively. The spring torque is generated by coil spring, the angle of coil spring can be adjusted by controlling a brushless DC motor. The damping torque is generated by MR fluid, the damping coefficient of MR fluid can be adjusted by controlling the current of excitation coil.

Improvement of shift quality evaluation has become more prevalent over the past few years in the development of automatic transmission electronic control system. For the problems of the subjective shift quality evaluation that subjectivity is too strong, the standard cannot be unified and the definition of the objective evaluation index is not clear at present, this paper studies on the methods of objective evaluation of shift quality based on the multi-hierarchical grey relational analysis. Firstly, objective evaluation index system is constructed based on physical quantities, such as the engine speed, the longitudinal acceleration of the vehicle and so on, which broadens the scope of the traditional objective evaluation index further.

Due to growing interest in automated driving, the need for better understanding of human driving behavior in uncertain environment, such as driving behavior at un-signalized crossroad and roundabout, has further increased. Driving behavior at roundabout is greatly influenced by different dynamic factors such as speed, distance and circulating flow of the potentially conflicting vehicles, and drivers should choose whether to leave or wait at the upcoming exit according to these factors. In this paper, the influential dynamic factors and driving behavior characteristics at the roundabout is analyzed in detail, random forest method is then deployed to predict the driving behavior. For training the driving behavior model, four typical roundabout layouts were created under a real-time driving simulator with PanoSim-RT and dSPACE. Traffic participants with different motion style were also set in the simulation platform to mimic real driving conditions.

Simulation has been considered as one of the key enablers on the development and testing for autonomous driving systems as in-vehicle and field testing can be very time-consuming, costly and often impossible due to safety concerns. Accurately modeling traffic, therefore, is critically important for autonomous driving simulation on threat assessment, trajectory planning, etc. Traditionally when modeling traffic, the motion of traffic vehicles is often considered to be deterministic and modeled based on its governing physics. However, the sensed or perceived motion of traffic vehicles can be full of errors or inaccuracy due to the inaccurate and/or incomplete sensing information. In addition, it is naturally true that any future trajectories are unknown. This paper proposes a novel modeling method on traffic considering its motion uncertainties, based on Gaussian process (GP).

In this work, a XD132 Road Roller from XCMG in China was employed as a research basis to study the heat exchange performance of the heat dissipation module under varied working conditions. The module in the XD132 consists of a cooling fan and three radiators. At first, the numerical investigation on the elementary units of radiators was performed to obtain Colburn j factor and Fanning friction f factor, which were used for the ε-NTU method to predict the radiator performance. The fan was numerically tested in a wind test tunnel to acquire the performance curve. The performance data from both investigations were transformed into the boundary conditions of the numerical vehicle model in a virtual tunnel. A field experiment was carried out to validate the simulation accuracy, and an entrance coefficient was proposed to discuss the performance regularity under four working conditions.

In order to predict the fatigue life of thrust rod heavy duty commercial vehicle balanced suspension, based on the continuum mechanics theory, the fatigue life prediction model of rubber with equivalent effect as damage parameter is established. Based on the equivalent stress and fatigue cumulative damage theory, the fatigue damage evolution equation of rubber material expressed by stress is derived by using the strain energy function. The general fatigue life model is established by using the maximum logarithmic principal strain as the damage parameter. The finite element model of the thrust rod is established, and the stress distribution of the spherical hinge rubber layer and the easy damage area are analyzed. Based on the equivalent stress calculation results and the axial tension stress and strain data of the rubber material, the accuracy of the results of the finite element calculation is verified.

Vehicle Longitudinal Control (VLC) algorithm is the basis function of automotive Cruise Control system. The main task of VLC is to achieve a longitudinal acceleration tracking controller, performance requirements of which include fast response and high tracking accuracy. At present, many control methods are used to implement vehicle longitudinal control. However, the existing methods are need to be improved because these methods need a high accurate vehicle dynamic model or a number of experiments to calibrate the parameters of controller, which are time consuming and costly. To overcome the difficulties of controller parameters calibration and accurate vehicle dynamic modeling, a vehicle longitudinal control algorithm based on iterative learning control (ILC) is proposed in this paper. The algorithm works based on the information of input and output of the system, so the method does not require a vehicle dynamics model.

At the collision moment, a driver’s lower extremity will be in different foot position, which leads to the different posture of the lower extremity with various muscle activations. These will affect the driver’s injury during collision, so it is necessary to investigate further. A simulated collision scene was constructed, and 20 participants (10 male and 10 female) were recruited for the test in a driving simulator. The braking posture and muscle activation of eight major muscles of driver’s lower extremity (both legs) were measured. The muscle activations in different postures were then analyzed. At the collision moment, the right leg was possible to be on the brake (male, 40%; female, 45%), in the air (male, 27.5%; female, 37.5%) or even on the accelerator (male, 25%; female, 12.5%). The left leg was on the floor all along.

In this paper, the dynamic stress of the front subframe of a passenger car was obtained using modal stress recovery method to estimate the fatigue life. A finite element model of the subframe was created and its accuracy was checked by modal test in a free hanging state. Furthermore, the whole vehicle rigid-flexible coupling model of the passenger car was built up while taking into account the flexibility of the subframe. Meanwhile, the road test data was used to verify the validity of the dynamic model. On this basis, the modal displacement time histories of the subframe were calculated by a dynamic simulation on virtual proving ground consisting of Belgian blocks, cobblestone road and washboard road. By combining the modal displacement time histories with modal stress tensors getting from normal mode analysis, the dynamic stress time histories of the subframe were obtained through modal stress recovery method.