Explore the vast range of titles available.

Real-time object detection is significant for industrial and research fields. On edge devices, a giant model is difficult to achieve the real-time detecting requirement, and a lightweight model built from a large number of the depth-wise separable convolutional could not achieve the sufficient accuracy. We introduce a new lightweight convolutional technique, GSConv, to lighten the model but maintain the accuracy. The GSConv accomplishes an excellent trade-off between the accuracy and speed. Furthermore, we provide a design suggestion based on the GSConv, slim-neck (SNs), to achieve a higher computational cost-effectiveness of the real-time detectors. The effectiveness of the SNs was robustly demonstrated in over twenty sets comparative experiments. In particular, the real-time detectors of ameliorated by the SNs obtain the state-of-the-art (70.9% AP 50 for the SODA10M at a speed of ~ 100 FPS on a Tesla T4) compared with the baselines. Code is available at https://github.com/alanli1997/slim-neck-by-gsconv .

Monocular camera and Lidar are the two most commonly used sensors in unmanned vehicles. Combining the advantages of the two is the current research focus of SLAM and semantic analysis. In this paper, we propose an improved SLAM and semantic reconstruction method based on the fusion of Lidar and monocular vision. We fuse the semantic image with the low-resolution 3D Lidar point clouds and generate dense semantic depth maps. Through visual odometry, ORB feature points with depth information are selected to improve positioning accuracy. Our method uses parallel threads to aggregate 3D semantic point clouds while positioning the unmanned vehicle. Experiments are conducted on the public CityScapes and KITTI Visual Odometry datasets, and the results show that compared with the ORB-SLAM2 and DynaSLAM, our positioning error is approximately reduced by 87%; compared with the DEMO and DVL-SLAM, our positioning accuracy improves in most sequences. Our 3D reconstruction quality is better than DynSLAM and contains semantic information. The proposed method has engineering application value in the unmanned vehicles field.

In the realm of autonomous driving tort, a significant disparity exists in the parties’ access to autonomous driving data and essential technical information, resulting in challenges in unilateral proof. The traditional burden of proof framework in driving litigation is inadequate for direct application in the autonomous driving sphere. As we approach the era of widespread autonomous driving operations, there is an urgent need to clarify and redefine the allocation of the burden of proof in specific litigations. Utilizing comparative legal analysis and case studies, this paper delves into the disparities in the legislative provisions concerning the burden of proof for autonomous driving in Germany and China. China can learn from Germany’s legislative precedence in shifting the burden of proof for “product defect” and “fault” onto the manufacturer, thereby requiring the infringed party to merely furnish preliminary evidence indicating a “causal relationship between the defect and the damage”. This approach mitigates the evidentiary burden on the aggrieved party, clarifies the litigation procedures, incentivizes manufacturers to enhance the technology, reinforces risk management, and ultimately facilitates the progression of autonomous driving technology.

PHEVs (plug-in hybrid electric vehicles) equipped with diesel engines have multiple model transitions in the driving cycle for their particular structure. The high frequency of start–stop of a diesel engine will increase fuel consumption and reduce the catalytic efficiency of SCR (Selective Catalyst Reduction) catalysts, which will increase cold start emissions. For comprehensive optimization of fuel consumption and emissions, an optimal control strategy of PHEVs that originated from the PER-TD3 algorithm based on DRL (deep reinforcement learning) is proposed in this paper. The priority of samples is assigned with greater sampling weight for high learning efficiency. Experimental results are compared with those of the DP (dynamic programming)-based strategy in HIL (hardware in loop) equipment. The engine fuel consumption and NOX emissions were 2.477 L/100 km and 0.2008 g/km, nearly 94.1% and 90.1% of those of the DP-based control strategy. By contrast, the fuel consumption and NOx of DDPG (Deep Deterministic Policy Gradient)-based and TD3(Twin Delayed Deep Deterministic Policy Gradient) -based control strategy were 2.557, 0.2078, 2.509, and 0.2023, respectively. By comparative results, we can see that the comprehensive control strategy of PHEVs based on the PER-TD3 algorithm we proposed can achieve better performance with comparison to TD3-based and DDPG-based, which is the state-of-the-art strategy in DRL. The HIL-based experimental results prove the effectiveness and real-time potential of the proposed control strategy.

Great advances had been achieved in the discipline of environmental perception, motion planning and control strategy implementation, however, fully autonomous vehicle is still far from large-scale commercial application. The concept of “human-automation shared control” provides a promising solution to enhance autonomous driving safety, to which great research effort has been contributed in recent years. Nevertheless, more attention should be given to the following aspects. The present shared control strategy either only considers the discontinuous switching control between driver and ADS or investigates the simple effect of driver’s behavior in specific scenarios. The adaptive authority allocation between the driver’s active assistance and ADS hasn’t been investigated yet. In this paper, a shared control experiment with driver’s active assistance is conducted in scheduled traffic scenarios to observe the state of vehicle and arm’ EMG signal. After that, we construct a feature classification algorithm for shared control authority by clustering the experimental data. Then, a SCS with incremental PID controller and 2 DOF vehicle dynamic model is proposed. For validation of the SCS, the comparison of vehicle performance for different control authority illustrates that SCS can allocate appropriate control authority to improve the safety.

Chicken coccidiosis is an intestinal disease caused by the parasite Eimeria, which severely damages the growth of chickens and causes significant economic losses in the poultry industry. Improvement of the immune protective effect of antigens to develop high efficiency subunit vaccines is one of the hotspots in coccidiosis research. Sporozoite‐specific surface antigen 1 (SAG1) of Eimeria tenella (E. tenella) is a well‐known protective antigen and is one of the main target antigens for the development of subunit, DNA and vector vaccines. However, the production and immunoprotective effects of SAG1 need to be further improved. Here, we report that both SAG1 from E. tenella and its fusion protein with the xylanase XynCDBFV‐SAG1 are recombinant expressed and produced in Pichia pastoris (P. pastoris). The substantial expression quantity of fusion protein XynCDBFV‐SAG1 is achieved through fermentation in a 15‐L bioreactor, reaching up to about 2 g/L. Moreover, chickens immunized with the fusion protein induced higher protective immunity as evidenced by a significant reduction in the shedding of oocysts after E. tenella challenge infection compared with immunized with recombinant SAG1. Our results indicate that the xylanase enhances the immunogenicity of subunit antigens and has the potential for developing novel molecular adjuvants. The high expression level of fusion protein XynCDBFV‐SAG1 in P. pastoris holds promise for the development of effective recombinant anti‐coccidial subunit vaccine. We report that both SAG1 from E. tenella and its fusion protein with the xylanase XynCDBFV‐SAG1 are recombinant expressed and produced in Pichia pastoris (P. pastoris). The substantial expression quantity of fusion protein XynCDBFV‐SAG1 is achieved through fermentation in a 15‐L bioreactor, reaching up to about 2 g/L. Moreover, chickens immunized with the fusion protein induced higher protective immunity as evidenced by a significant reduction in the shedding of oocysts after E. tenella challenge infection compared with immunized with recombinant SAG1. Our results indicate that the xylanase enhances the immunogenicity of subunit antigens and has the potential for developing novel molecular adjuvants.

Ocean carbon sink is an emerging and interdisciplinary research area that plays a vital role in the global carbon cycle. This paper reviews recent scientific advancements in ocean carbon sink research, focusing on the mechanisms for capturing, utilizing, and sequestering atmospheric CO 2 , and highlights its contribution to climate change mitigation and adaptation. Using bibliometric analysis based on CiteSpace and data from the Web of Science and Scopus, we examine research hotspots and topic evolution through country collaboration, journal co-citation, and keyword co-occurrence networks. The findings show that ocean carbon sink research is shaped by complex scientific uncertainties and the integration of multiple disciplines. Current research hotspots include scientific advances, technological innovation, and governance challenges related to sustainable development. In general, recent studies emphasize the role of carbon sink, the value of nature, and the importance of precautionary management. This paper underlines the need for coordination between scientific and social dimensions of carbon sink functions, and it draws attention to the ethical aspects of carbon sink governance. It advocates for multi-stakeholder participation, precautionary governance, and policy-based financial system to support climate resilience and foster the sustainable development of the oceans.

For investigating driver characteristic as well as control authority allocation during the process of human–vehicle shared control (HVSC) for an autonomous vehicle (AV), a HVSC dynamic mode with a driver’s neuromuscular (NMS) state parameters was proposed in this paper. It takes into account the driver’s NMS characteristics such as stretch reflection and reflex stiffness. By designing a model predictive control (MPC) controller, the vehicle’s state feedback and driver’s state are incorporated to construct the HVSC dynamic model. For the validation of the model, a field experiment was conducted. The vehicle state signals are collected by V-BOX, and the driver’s state signals are obtained with the electromyography instrument. Subsequently, the hierarchical least square (HLS) parameter identification algorithm was implemented to identify the parameters of the model based on the experimental results. Moreover, the Unscented Kalman Filter (UKF) was utilized to estimate the important NMS parameters which cannot be measured directly. The experimental results showed that the model we proposed has excellent accuracy in characterizing the vehicle’s dynamic state and estimating the driver’s NMS parameter. This paper will serve as a theoretical basis for the new control strategy allocation between human and vehicle for L3 class AVs.

In this paper, the generation mechanism of the negative airgap eccentricity effect for the in-wheel switched reluctance motor (SRM) driving system is analyzed. An independent current chopping control strategy is proposed to achieve optimum control between the response characteristic of the in-wheel motor driving system and the dynamic performance of electric vehicle (EV). Firstly, the electromagnetic characteristic of the studied SRM under airgap eccentricity is studied based on electromagnetic coupling model and circuit driving equation, and the radial electromagnetic force under different airgap eccentricity is verified by adopting the built experiment device. Then, combined with the excitation characteristics of the radial electromagnetic force, the negative dynamic effect of the in-wheel motor driving system is analyzed in the time–frequency domain. Finally, an independent current chopping control strategy for the in-wheel SRM driving system based on vehicle vibration feedback is proposed. The controller parameters including the turn-off angle and chopping current threshold are optimized by data interpolation. Results show that the proposed control strategy can achieve the optimum control between the response characteristics of the in-wheel motor driving system and the vehicle dynamic performance, especially to suppress the vehicle sprung mass acceleration and tire bounce while starting EV.

Costate value plays a significant role in the application of PMP-based control strategy for PHEV. It is critical for terminal SOC of battery at destination and corresponding equivalent fuel consumption. However, it is not convenient to choose the approximate costate in real driving condition. In the paper, the optimal control problem of PHEV based on PMP has been converted to nonlinear programming problem. By means of KKT condition costate can be approximated as KKT multipliers of NLP divided by the LGL weights. A kind of general costate estimation approach is proposed for predefined driving condition in this way. Dynamic model has been established in Matlab/Simulink in order to prove the effectiveness of the method. Simulation results demonstrate that the method presented in the paper can deduce the closer value of global optimal value than constant initial costate value. This approach can be used for initial costate and jump condition estimation of PMP-based control strategy for PHEV.