To tackle the challenges of weak sensing capacity for multi-scale objects, high missed detection rates for occluded targets, and difficulties for model deployment in detection tasks of intelligent roadside perception systems, the PDT-YOLO algorithm based on YOLOv7-tiny is proposed. Firstly, we introduce the intra-scale feature interaction module (AIFI) and reconstruct the feature pyramid structure to enhance the detection accuracy of multi-scale targets. Secondly, a lightweight convolution module (GSConv) is introduced to construct a multi-scale efficient layer aggregation network module (ETG), enhancing the network feature extraction ability while maintaining weight. Thirdly, multi-attention mechanisms are integrated to optimize the feature expression ability of occluded targets in complex scenarios, Finally, Wise-IoU with a dynamic non-monotonic focusing mechanism improves the accuracy and generalization ability of model sensing. Compared with YOLOv7-tiny, PDT-YOLO on the DAIR-V2X-C dataset improves mAP50 and mAP50:95 by 4.6% and 12.8%, with a parameter count of 6.1 million; on the IVODC dataset by 15.7% and 11.1%. We deployed the PDT-YOLO in an actual traffic environment based on a robot operating system (ROS), with a detection frame rate of 90 FPS, which can meet the needs of roadside object detection and edge deployment in complex traffic scenes.

The roadside LiDAR(light detection and ranging) sensors can obtain relevant information about all roadway users by collecting real-time 3D point cloud data of surrounding objects. In order to analyze the point cloud data of roadway users, background filtering and object detection are essential steps. This paper proposes a background filtering method based on point correlation by KDTree neighborhood searching and an adaptive threshold Euclidean clustering method. In order to improve the detection speed, background filtering will be performed on the original point cloud data before Euclidean clustering, so as to delete a large number of useless background points, and the clustering results will be distinguished by cuboid frame markers. The experimental results show that the single-frame point cloud clustering speed has increased by 31.3%, the clustering speed has been significantly improved.

Lithium-ion battery plays a key role in electric vehicles, which is critical to the system availability. One of the most important aspects in battery managements systems(BMS) in electric vehicles is the stage of health(SOH) estimation. The state of health (SOH) estimation is very critical to battery management system to ensure the safety and reliability of EV battery operation. The classical approach of current integration(coulomb counting) can't get the accurate values because of accumulative error. In order to provide timely maintenance and replacements of electric vehicles, several estimation approaches have been proposed to develop a reliable and accurate battery state of health estimation. A common drawback of previous algorithm is that the computation quantity is huge and not quite accurate, that is updated partially in this study. In this paper, a training strategy is diagrammed and models based on autoregressiveparticle filter (AR-PF) are developed with novel parameters to provide better prediction of electric vehicle state of health. The estimations are compared to measurements using autoregressive algorithm and particle filter respectively. The proposed training strategy achieves a better prediction precision than other strategies used in this paper. Battery experimental data and AR-PF predicted value are presented in X-Y axis figure as well as other two algorithms in this paper that we can observe them directly. Experiments based on NASA battery data set show that the proposed method yields a good performance in SOH estimation of Lithium-ion battery. The results showed that the health condition parameters accurately reflected the real time SOH of the battery and ensured that an accurate capacity estimation can be calculated.