Ensuring safety in autonomous driving is crucial for effective motion planning and navigation. However, most end-to-end planning methodologies lack sufficient safety measures. This study tackles this issue by formulating the control optimization problem in autonomous driving as Constrained Markov Decision Processes (CMDPs). We introduce an innovative, model-based approach for policy optimization, employing a conditional Value-at-Risk (VaR)-based soft actor-critic (SAC) to handle constraints in complex, high-dimensional state spaces. Our method features a worst-case actor to ensure strict compliance with safety requirements, even in unpredictable scenarios. The policy optimization leverages the augmented Lagrangian method and leverages latent diffusion models to forecast and simulate future trajectories. This dual strategy ensures safe navigation through environments and enhances policy performance by incorporating distribution modeling to address environmental uncertainties. Empirical evaluations conducted in both simulated and real environments demonstrate that our approach surpasses existing methods in terms of safety, efficiency, and decision-making capabilities.
Border patrol object detection is an important basis for obtaining information about the border patrol area and for analyzing and determining the mission situation. Border Patrol Staffing is now equipped with medium to close range UAVs and portable reconnaissance equipment to carry out its tasks. In this paper, we designed a detection algorithm TP-ODA for the border patrol object detection task in order to improve the UAV and portable reconnaissance equipment for the task of border patrol object detection, which is mostly performed in embedded devices with limited computing power and the detection frame imbalance problem is improved; finally, the PDOEM structure is designed in the neck network to optimize the feature fusion module of the algorithm. In order to verify the improvement effect of the algorithm in this paper, the Border Patrol object dataset BDP is constructed. The experiments show that, compared to the baseline model, the TP-ODA algorithm improves mAP by 2.9%, reduces GFLOPs by 65.19%, reduces model volume by 63.83% and improves FPS by 8.47%. The model comparison experiments were then combined with the requirements of the border patrol tasks, and it was concluded that the TP-ODA model is more suitable for UAV and portable reconnaissance equipment to carry and can better fulfill the task of border patrol object detection.
The growing demand for efficient last-mile delivery has made routing optimization a critical challenge for logistics providers. Traditional vehicle routing models typically minimize a single criterion, such as travel distance or time, without considering broader social and environmental impacts. This paper proposes a novel Multi-Criteria Adaptive Hybrid Ant Colony Optimization (MCAH-ACO) algorithm for solving the delivery vehicle routing problem formulated as a Multiple Traveling Salesman Problem (MTSP). The proposed MCAH-ACO introduces three key innovations: a multi-criteria pheromone decomposition strategy that maintains separate pheromone matrices for each optimization objective, an adaptive weight balancing mechanism that dynamically adjusts criterion weights to prevent dominance by any single objective, and a 2-opt local search enhancement integrated with elite archive diversity preservation. A comprehensive cost function is designed to integrate four categories of factors: distance, time, social-environmental impact, and safety. Extensive experiments on real-world data from the Greater Toronto Area demonstrate that MCAH-ACO significantly outperforms existing approaches including Genetic Algorithm (GA), Adaptive GA, and standard Max–Min Ant System (MMAS), achieving 12.3% lower total cost and 18.7% fewer safety-critical events compared with the best baseline while maintaining computational efficiency.
Wildfires often spread rapidly and cause significant casualties and economic losses. Firefighting drones carrying water capsules provide an efficient way for wildfire extinguishing, but their operational capabilities are limited by their payloads. This weakness can be compensated by using ground vehicles to provide mobile water supply. To this end, this paper presents an optimization problem of scheduling multiple drones and water supply trucks for wildfire fighting, which allocates burning subareas to drones, routes drones to perform fire-extinguishing operations in burning subareas and reload water between every two consecutive operations, and routes trucks to provide timely water supply for drones. To solve the problem within the limited emergency response time, we propose a deep reinforcement learning method, which consists of an encoder for embedding the input instance features and a decoder for generating a solution by iteratively predicting the subarea selection decision through attention. Computational results on test instances constructed upon real-world wilderness areas demonstrate the performance advantages of the proposed method over a collection of heuristic and metaheuristic optimization methods.
Modern email spam and phishing attacks have evolved far beyond keyword blacklists or simple heuristics. Adversaries now craft multi-modal campaigns that combine natural-language text with obfuscated URLs, forged headers, and malicious attachments, adapting their strategies within days to bypass filters. Traditional spam detection systems, which rely on static rules or single-modality models, struggle to integrate heterogeneous signals or to continuously adapt, leading to rapid performance degradation. We propose EvoMail, a self-evolving cognitive agent framework for robust detection of spam and phishing. EvoMail first constructs a unified heterogeneous email graph that fuses textual content, metadata (headers, senders, domains), and embedded resources (URLs, attachments). A Cognitive Graph Neural Network enhanced by a Large Language Model (LLM) performs context-aware reasoning across these sources to identify coordinated spam campaigns. Most critically, EvoMail engages in an adversarial self-evolution loop: a ''red-team'' agent generates novel evasion tactics---such as character obfuscation or AI-generated phishing text---while the ''blue-team'' detector learns from failures, compresses experiences into a memory module, and reuses them for future reasoning.Extensive experiments on real-world datasets (Enron-Spam, Ling-Spam, SpamAssassin, and TREC) and synthetic adversarial variants demonstrate that EvoMail consistently outperforms state-of-the-art baselines in detection accuracy, adaptability to evolving spam tactics, and interpretability of reasoning traces. These results highlight EvoMail's potential as a resilient and explainable defense framework against next-generation spam and phishing threats.