Zhifu Wang

Lane detection is a key technology in automatic driving environment perception, and its accuracy directly affects vehicle positioning, path planning, and driving safety. In this study, an enhanced real-time model for lane detection based on an improved DeepLabV3+ architecture is proposed to address the challenges posed by complex dynamic backgrounds and blurred road boundaries in suburban road scenarios. To address the lack of feature correlation in the traditional Atrous Spatial Pyramid Pooling (ASPP) module of the DeepLabV3+ model, we propose an improved LC-DenseASPP module. First, inspired by DenseASPP, the number of dilated convolution layers is reduced from six to three by adopting a dense connection to enhance feature reuse, significantly reducing computational complexity. Second, the convolutional block attention module (CBAM) attention mechanism is embedded after the LC-DenseASPP dilated convolution operation. This effectively improves the model’s ability to focus on key features through the adaptive refinement of channel and spatial attention features. Finally, an image-pooling operation is introduced in the last layer of the LC-DenseASPP to further enhance the ability to capture global context information. DySample is introduced to replace bilinear upsampling in the decoder, ensuring model performance while reducing computational resource consumption. The experimental results show that the model achieves a good balance between segmentation accuracy and computational efficiency, with a mean intersection over union (mIoU) of 95.48% and an inference speed of 128 frames per second (FPS). Additionally, a new lane-detection dataset, SubLane, is constructed to fill the gap in the research field of lane detection in suburban road scenarios.

The rapid devolopment of Internet of Vehicles (IoV) and Autonomous Connected Vehicles (ACVs) has increased the complexity of in-vehicle networks, exposing security vulnerabilities in traditional Controller Area Network (CAN) systems. CAN security faces dual challenges: stringent computational constraints imposed by automotive functional safety requirements and the impracticality of protocol modifications in multi-device networks. To address this, we propose a lightweight intrusion detection algorithm leveraging information entropy to analyze side-channel CAN message ID distributions. Evaluated in terms of detection accuracy, false positive rate, and sensitivity to bus load variations, the algorithm was implemented on an NXP MPC-5748G embedded platform through the AutoSar Framework. Experimental results demonstrate robust performance under low computational resources, achieving high detection accuracy with high recall (>80%) even at 10% bus load fluctuation thresholds. This work provides a resource-efficient security framework compatible with existing CAN infrastructures, effectively balancing attack detection efficacy with the operational constraints of automotive embedded systems.

The electric motor emulator (EME) represents a practical application of power hardware-in-the-loop (PHIL) technology, facilitating the testing of electric motor drive units (MDUs). Its main function is to replicate both the steady-state and dynamic current characteristics of electric motors. Typically, the current loop in the EME functions as the inner loop of that in the MDU, requiring a high control bandwidth to ensure superior dynamic response performance. This article analyzes the limitations of the control strategy for the EME introduced in a recent study and proposes a modified digital complex vector controller (MDCVC) with better dynamic performance and enhanced disturbance rejection capability. First, an internal loop is introduced to replace the complex poles of the original platform with real poles, followed by the design of a current controller to achieve overshoot-free current tracking performance. Furthermore, to improve the disturbance suppression performance, an additional active damping term is incorporated into the control loop. As a result, the dynamic characteristics of the target electric motor are accurately reproduced. Comparative evaluation tests are carried out to validate the effectiveness of the proposed control scheme under various operating conditions.

Discrete space vector modulation (DSVM) and weighting factor elimination have been investigated to improve the system performance of finite-set model predictive control (FS-MPC). However, the existing DSVM-based FS-MPC without weighting factors for three-level inverters suffers from high computational burden, low algorithm flexibility, and substantially narrowed operating range. To address these issues, this article proposes a novel DSVM-based FS-MPC with the virtual medium voltage vectors (MVVs), which greatly reduces the current harmonics and computational burden, realizes a flexibly adjustable number of time intervals in DSVM, and eliminates the weighting factor while ensuring balanced neutral-point voltage (NPV) over the full range of operating frequencies and load conditions. The introduced virtual MVVs allow the candidate set to be always narrowed down to the voltage vector closest to the reference vector. A simple judgment scheme is proposed to determine whether a real or virtual MVV should be used, which enables a flexible constraint on the maximum NPV error. Carrier-based implementation is also achieved. Experimental results validate the proposed algorithm.

This article proposes a novel space vector modulation strategy with hybrid medium voltage vectors (MVVs) for three-level neutral-point-clamped (3L-NPC) inverters, which achieves high algorithm flexibility, low switching losses, and low computational burden while ensuring neutral-point voltage (NPV) balance over the full range of modulation indices and power factors. The proposed strategy introduces six additional virtual MVVs at the locations of the six actual MVVs, which allows the space vector distribution to be maintained without modifying the synthesis of the reference voltage vector. A simple judgment scheme is presented to determine whether a virtual or actual MVV should be used, which can be flexibly tuned to greatly reduce switching losses while ensuring always balanced NPV. Besides, a novel and simple method for determining the optimal switching sequence is proposed. Furthermore, carrier-based modulation with adjustable switching patterns is achieved to further simplify implementation and improve flexibility. Simulation and experimental results validate the proposed strategy.

This article proposes a simplified finite-set model predictive control (FS-MPC) algorithm for full-frequency operation of five-level nested neutral-point-clamped (5L-NNPC) inverters. Compared to the existing FS-MPC of 5L-NNPC inverters, the proposed algorithm substantially reduces the computational burden and simplifies the tuning of the weighting factor while ensuring low current harmonics and full-frequency capacitor voltage balance. The full-frequency operating conditions that the candidate sets should satisfy are first acquired considering the periodic characteristics of the capacitor voltages. On this basis, a computationally efficient varying candidate set selection method is proposed for high-frequency operation, while a simple compensation set is introduced for low-frequency operation. Besides, an online approach is proposed to calculate available switching states for different candidate sets, eliminating tedious offline preprocessing. Furthermore, a staged cost function evaluation method is presented to further simplify the computation. Experimental results validate the proposed algorithm.