International Journal of Automotive Technology最新文献

This study investigates the use of deep reinforcement learning for active noise control (DRL-ANC) to cancel narrowband noise. The filtered-x least mean square algorithm for ANC, which includes the secondary path model in itself, has been widely used in various applications. If the path model is inaccurate due to the variations of the actual path, control performance and stability of the algorithm can be restricted. To eliminate the effect by the model inaccuracy, it is considered to remove the path model in the novel DRL-ANC strategy. A DRL approach using the deep deterministic policy gradient without any path model is adopted to learn the behavior of a physical environment including the effect of the actual secondary path in real time. However, a temporal credit assignment problem arises due to the time-delayed reward inherent in the secondary path, which means that the current action could not be evaluated by its true. To address this problem, this study proposes a novel definitions of the state and action of the RL agent, specialized in narrowband noise suppression. Additionally, a novel exploration noise is also suggested to enhance effectiveness and practicality of the learning process. Computer simulations and real-time control experiments were conducted, and the results demonstrated that the proposed DRL-ANC algorithm can robustly cope with changes in the secondary path.

In this study, various 3D printed metamaterials are investigated for application in energy absorbing structures in motorsports. Impact attenuating structures are used to decelerate vehicles and protect drivers in the event of a crash. Additive manufacturing enables complex plastic structures which can facilitate improved angular resistance and reduced weight and cost compared with traditional approaches. Metamaterials were 3D printed from PLA using commercially available equipment and include gyroid structures, a novel reinforced gyroid design and a lattice designed using finite-element analysis-based topology optimization. Compression testing was used to measure stress–strain response, compressive modulus, and energy absorption. This demonstrated gyroids and reinforced gyroids have ideal compressive behavior for high energy absorption under impact. The topology optimized metamaterial was found unsuitable for this application due to its high stiffness, revealing a weakness in traditional topology optimization approaches which are not catered to maximize energy absorption. The reinforced gyroid demonstrated the highest specific energy absorption and was used to manufacture an impact attenuator which demonstrated the potential to safely stop a hypothetical 300 kg vehicle crash. This work supports that gyroid-based structures can reduce weight, volume and cost over current materials in all motorsport categories, with improved safety from oblique crashes.

This paper presents a remote smart parking assist (RSPA) control algorithm for hybrid electric vehicles based on a disturbance observer (DOB) and feedback control. The purpose of the proposed control algorithm is to improve the departure/stop agility and speed control performance of the RSPA. The DOB, activated when the hydraulic brake is released and vehicle departure initiates, estimates disturbances such as slippery road, road slope and bump road, and quickly compensates for these disturbances. The feedback controller corrects the motor torque based on the target speed and current vehicle speed. The gain of the feedback controller is adjusted according to the road gradient, which is estimated by a gradient observer. To validate the proposed control algorithm, actual vehicle test is performed using the Kia Niro Plug-in hybrid electric vehicle (PHEV). The results of the vehicle test show that the proposed control algorithm improves the robustness of the RSPA function.

In this research work, a method for integrating a local reinforcement structure in a medium-pressure plate (MPP) for fuel cell electric vehicle (FCEV) applications was developed using steel–aluminium hybrid-casting technology. Using this technology, it is possible to create a bonded enclosure of a steel reinforcement patch with the cast aluminium pressure plate to increase its stiffness and achieve 15% package space savings. A load-compliant, manufacturable patch was chosen and optimised for maximum stiffness gains using non-linear static finite-element (FE) calculations. Special form and process requirements due to hybrid-casting technology were examined and secured with casting simulations. The reinforcement patch was manufactured and coated with a unique aluminium–silicon coating enabling a ductile material connection between the steel and aluminium, and casting trials were conducted to create prototypes. Additionally, the insulating plastic layer on top of the metallic pressure plate carrier was substituted from costly short-fibre-reinforced high-performance plastic to cheaper and stiffer glass-mat reinforced thermoplastic material. Finally, the new hybrid MPP was tested mechanically, and the FE-Model was verified. In summary, through the package gain, 2.1 kW more power output and 11% less weight could be achieved while remaining stiffness neutral.

This article proposes a new formulation for a robust trajectory tracking control law of an autonomous vehicle. Autonomous vehicle navigation highly relies on reliable, robust, and dependable steering mechanism, even under challenging conditions and circumstances. The controller design is based on the higher order quasi-sliding mode control (QSMC) algorithm that provides smooth motion control subjected to steering saturation and curvature constraints. In addition, an adaptive single input fuzzy logic control based on Lyapunov stability theorem is incorporated, which relies on the online estimation of perturbations rather than relying on the requirement of a priori knowledge of the upper bounds of the perturbation. Furthermore, the proposed control scheme exhibits a strong robustness toward the effect of uncertainties like parametric, tire cornering stiffness, surface bonding coefficient, and exogenous noises and disturbances. In addition to that, fuzzy control term offers a fast path-tracking error convergence toward equilibrium condition and reduced steady-state error. The overall control scheme through Lyapunov theory ensures the global asymptotic stability of the autonomous vehicle. Finally, the effectiveness and robustness of the proposed control scheme is demonstrated through numerical simulations MATLAB/SIMULINK platform for linear and nonlinear scenarios. Later, experimental validation is conducted over dSPACE SCALEXIO hardware-in-loop (HIL) platform for trajectory tracking along with the input constraints subjected to parametric uncertainties and disturbances.

Trends in the automotive industry are changing rapidly due to environmental factors, including climate change-induced exhaust gas regulations, urban traffic congestion, and the merging of communication and mobility technologies. As a result, new technologies such as electrification, autonomous driving, and mobility as a service (MaaS) are emerging, leading to the development of various types of mobility solutions, including battery electric and hydrogen vehicles, advanced air mobility (AAM), and purpose built vehicle (PBV). In response to these developments, research on a wide range of lightweight materials is being conducted to meet the requirements of these various modes of transportation. Carbon fiber-reinforced plastic (CFRP) is the most effective lightweight material for weight reduction; however, its high cost limits its application. To overcome this limitation, one solution is to produce composite materials using lower cost alternative materials and mass production processes. Another approach is to develop lightweight composite materials that offer additional advantages. Nevertheless, in mobility modes such as AAM, weight reduction is far more critical than cost sensitivity, indicating a potential increase in composite part application in this field.

Since accuracy and adaptation of the closed-loop control in selective catalytic reduction (SCR) and ammonia slip catalytic (ASC) systems quite depend on the detection of ammonia (NH₃), the identification of NH₃ slip is important in the above systems. A model-based NH₃ slip detection method by a brief dosing-off maneuver is proposed, which is utilizing NO_x sensors’ cross-sensitivity between NH₃ and NO_x. When the operating conditions are suitable, NH₃ slip detection function would trigger once the deviation of the DeNO_x efficiency calculated by model and by sensor is over a limit. The NH₃ slip could be detected when the consumption of NH₃ storage exceeded a limit earlier than DeNO_x efficiency below a limit. Validation test results showed this NH₃ slip detection function can effectively detect the NH₃ slip of SCR + ASC system, and immediately send out a NO_x slip result if the tailpipe NO_x emission exceeds the OBD limit.

Because of the volume of software created in the automotive sector, improving software quality has become a need, particularly in areas where safety is crucial, such as autonomous driving. So, continuous software inspection is critical, timely, and essential for ensuring software quality and avoiding worst-case situations in the automotive sector, particularly with autonomous driving. Nonetheless, our idea is inspired by the large number of failures associated with poor software quality. As a result, the primary goal of this research is to provide a solution for continuous software quality inspection in the context of the company Bosch Car Multimedia. A tool named CIAnalyzer Tool was created for this purpose and an architecture was built and deployed in which CIAnalyzer operates as the middleware between SonarQube and Jenkins, delivering an analysis of the software every time someone adds software to the repository. The software is analyzed and, if found to be suitable, is sent to the repository. After implementing this approach, there was a 77.7% reduction in bugs, a 64.6% reduction in code smells, and a 100% reduction in vulnerabilities.

Aiming at the problem of torque fluctuations and intake pressure fluctuations caused by cycle cylinder deactivation, we introduced a refined control strategy encompassing fixed cycle frequency with fixed deactivation frequency, variable cycle frequency with fixed deactivation frequency, and variable cycle frequency with variable deactivation frequency. The cycle cylinder deactivation rate is determined by controlling the cycle frequency and the cylinder deactivation frequency, which provides a new method for exploring the torque fluctuation of the cycle cylinder deactivation. Furthermore, we introduced an intake compensation method based on the cylinder deactivation compensation coordinate system is proposed. The non-uniformity of the intake air volume of each cylinder of the cycle cylinder deactivation is improved by changing the valve lift. After compensation, the non-uniformity of the intake air volume and the mean indicated pressure of each cylinder are within 2%, and the maximum torque fluctuation is reduced by 52%. It provides a useful reference for enhancing the dynamic smoothness of cylinder deactivation in engines.

Simultaneous Localization and Mapping (SLAM) is the foundation for high-precision localization, environmental awareness, and autonomous decision-making of autonomous vehicles. It has developed rapidly, but there are still challenges such as sensor errors, data fusion, and real-time computing. This paper proposes an optimization-based fusion algorithm that integrates IMU data, visual data and LiDAR data to construct a high-frequency visual-inertial odometry. The odometry is employed to obtain the relative pose transformation during the LiDAR data acquisition process, and eliminate the distortion of the point cloud by interpolation. By utilizing the local curvatures, some edge and plane features are extracted by LiDAR after removing the distortion, which are further combined with local map alignment to reconstruct the LiDAR constrains. In addition, the LiDAR odometer can be obtained through the initial values provided by high-frequency visual-inertial odometry. To address the cumulative error in odometers, adjacent keyframe and multi descriptor fusion loop constraints are combined to construct back-end optimization constraints, solving for high-accuracy localization results and constructing a 3D point cloud map of the surroundings. Compared with some classical algorithm, results show that the accuracy of this paper's algorithm is better than the laser SLAM method and the multi-sensor fusion SLAM method. Besides, the laser-assisted multi-feature visual-inertial odometry localization accuracy is also better than that of the single-feature visual-inertial odometry. In summary, the newly proposed SLAM method can largely improve the accuracy of odometry in real traffic scenarios.