Automotive Innovation最新文献

As intelligent vehicles usually have complex overtaking process, a safe and efficient automated overtaking system (AOS) is vital to avoid accidents caused by wrong operation of drivers. Existing AOSs rarely consider longitudinal reactions of the overtaken vehicle (OV) during overtaking. This paper proposed a novel AOS based on hierarchical reinforcement learning, where the longitudinal reaction is given by a data-driven social preference estimation. This AOS incorporates two modules that can function in different overtaking phases. The first module based on semi-Markov decision process and motion primitives is built for motion planning and control. The second module based on Markov decision process is designed to enable vehicles to make proper decisions according to the social preference of OV. Based on realistic overtaking data, the proposed AOS and its modules are verified experimentally. The results of the tests show that the proposed AOS can realize safe and effective overtaking in scenes built by realistic data, and has the ability to flexibly adjust lateral driving behavior and lane changing position when the OVs have different social preferences.

In the long-term prediction of battery degradation, the data-driven method has great potential with historical data recorded by the battery management system. This paper proposes an enhanced data-driven model for Lithium-ion (Li-ion) battery state of health (SOH) estimation with a superior modeling procedure and optimized features. The Gaussian process regression (GPR) method is adopted to establish the data-driven estimator, which enables Li-ion battery SOH estimation with the uncertainty level. A novel kernel function, with the prior knowledge of Li-ion battery degradation, is then introduced to improve the modeling capability of the GPR. As for the features, a two-stage processing structure is proposed to find a suitable partial charging voltage profile with high efficiency. In the first stage, an optimal partial charging voltage is selected by the grid search; while in the second stage, the principal component analysis is conducted to increase both estimation accuracy and computing efficiency. Advantages of the proposed method are validated on two datasets from different Li-ion batteries: Compared with other methods, the proposed method can achieve the same accuracy level in the Oxford dataset; while in Maryland dataset, the mean absolute error, the root-mean-squared error, and the maximum error are at least improved by 16.36%, 32.43%, and 45.46%, respectively.

Software-defined vehicles have been attracting increasing attentions owing to their impacts on the ecosystem of the automotive industry in terms of technologies, products, services and enterprise coopetition. Starting from the technology improvements of software-defined vehicles, this study systematically combs the impact of software-defined vehicles on the value ecology of automotive products and the automotive industrial pattern. Then, based on the current situation and demand of industrial development, the main challenges hindering the realization of software-defined vehicles are identified, including that traditional research and development models cannot adapt to the iterative demand of new automotive products; the transformation of enterprise capability faces multiple challenges; and many contradictions exist in the industrial division of labor. Finally, suggestions are put forward to address these challenges and provide decision-making recommendations for enterprises on strategy management.

In-wheel motor-drive electric vehicles have the advantage of independently controllable wheel torque and the disadvantages of unsprung mass rise and power restriction. To address the disadvantages, a centralized layout electric torque vectoring drive-axle system (E-TVDS) with dual motors is proposed, which can realize arbitrary distribution of driving torque between the left and right wheels. First, the speed and torque distribution principle of E-TVDS based on velocity diagram are analyzed, and a virtual prototype of the whole vehicle with basic gear ratio relation model of the E-TVDS is built for simulation to verify the theoretical results and the basic effect of E-TVDS on the steering performance of the vehicle. Second, the characteristics of 36 types of the novel E-TVDS topology structure are compared and analyzed, and the optimal structure scheme is selected. Third, the accurate multiple degrees of freedom dynamic model for the optimal structure is established by using the bond graph method, and its dynamic response characteristics are analyzed. The results show that the vehicle equipped with the proposed E-TVDS can distribute the driving torque with the almost identical amount but opposite sign between the left and right wheels in any direction, and varying amount according to different chassis dynamics control requirements, and the torque response performance is great with little delay and overshoot. The function and dynamic response of the proposed E-TVDS show that it has potential application value for various performance improvements of electric vehicles.

Safety assurance of automated driving systems must consider uncertain environment perception. This paper reviews literature addressing how perception testing is realized as part of safety assurance. The paper focuses on testing for verification and validation purposes at the interface between perception and planning, and structures the analysis along the three axes (1) test criteria and metrics, (2) test scenarios, and (3) reference data. Furthermore, the analyzed literature includes related safety standards, safety-independent perception algorithm benchmarking, and sensor modeling. It is found that the realization of safety-oriented perception testing remains an open issue since challenges concerning the three testing axes and their interdependencies currently do not appear to be sufficiently solved.

High-definition map has become a vital cornerstone in the navigation of autonomous vehicles in complex traffic scenarios. Thus, the construction of high-definition maps has become crucial. Traditional methods relying on expensive mapping vehicles equipped with high-end sensor equipment are not suitable for mass map construction because of the limitation imposed by its high cost. Hence, this paper proposes a new method to create a high-definition road semantics map using multi-vehicle sensor data. The proposed method implements crowdsourced point-based visual SLAM to align and combine the local maps derived by multiple vehicles. This allows users to modify the extraction process by using a more sophisticated neural network, thus achieving a more accurate detection result when compared with traditional binarization method. The resulting map consists of road marking points suitable for autonomous vehicle navigation and path-planning tasks. Finally, the method is evaluated on the real-world KAIST urban dataset and Shougang dataset to demonstrate the level of detail and accuracy of the proposed map with 0.369 m in mapping errors in ideal condition.

Lithium-ion batteries have become one of the most promising technologies for speeding up clean automotive applications, where electrode plays a pivotal role in determining battery performance. Due to the strongly-coupled and highly complex processes to produce battery electrode, it is imperative to develop an effective solution that can predict the properties of battery electrode and perform reliable sensitivity analysis on the key features and parameters during the production process. This paper proposes a novel tree boosting model-based framework to analyze and predict how the battery electrode properties vary with respect to parameters during the early production stage. Three data-based interpretable models including AdaBoost, LPBoost, and TotalBoost are presented and compared. Four key parameters including three slurry feature variables and one coating process parameter are analyzed to quantify their effects on both mass loading and porosity of battery electrode. The results demonstrate that the proposed tree model-based framework is capable of providing efficient quantitative analysis on the importance and correlation of the related parameters and producing satisfying early-stage prediction of battery electrode properties. These can benefit a deep understanding of battery electrodes and facilitate to optimizing battery electrode design for automotive applications.

The lane-change transportation research usually focuses on the efficiency and stability of the macro traffic flow while ignoring the driving comfort of individual vehicles. And many studies of lane-change models are often limited to the performance of a single vehicle, which leads to a lack of macroscopic evaluation. To solve the above limitations, an automatic lane-change generalized dynamic model is adopted. In this model, the lane-change behavior of an individual vehicle is considered as the generalized excitation and the restraining force between vehicles is described with the car-following model. Macro and micro evaluation indexes are also adopted to evaluate the automatic lane-change behavior in traffic flow. Furthermore, this paper proposes a modified intelligent driver model (IDM) to describe the state change process during lane change. The hyperbolic tangent transition function is used to eliminate the vehicle state mutation. The simulation results show that the proposed automatic lane-change generalized dynamic model can reflect the macro and micro parameters of the traffic flow. And compared with the traditional IDM model, the proposed HC-IDM model achieves higher comfort performance and lower fluctuation of traffic flow.

Fleets of autonomous vehicles including shuttle buses, freight trucks, and road sweepers will be deployed in the Olympic Village during Beijing 2022 Winter Olympics. This requires intelligent charging infrastructure based on wireless power transfer technology to be equipped. To increase the misalignment tolerance of a high-power wireless charger, the robustness of the magnetic coupler should be optimized. This paper presents a new type of unipolar coupler, which is composed of three connected coils in series. The dimensional configuration of the coils is analyzed by the finite element method. The characteristic parameters of the coil are identified with their influence on the self-inductance and coupling coefficient. An expert model is built, whose feasibility can be verified in the aimed design domain. Combined with the expert model, an improved simulated annealing algorithm with a backtracking mechanism is proposed. The primary coil can reach the expected characteristics from any starting parameter combination through the proposed optimization algorithm. Under the same conditions in terms of external circuit parameters, ferrite usage, and aluminum shielding, the offset sensitivity of the magnetic coupler can be reduced from 58.79% to 18.89%. A prototype is established, validating the feasibility of the proposed coil structure with the optimized parameter algorithm.