A Novel Parallel Control Method for Optimal Consensus of Nonlinear Multiagent Systems.

This work concentrates on the initial introduction of parallel control to investigate an optimal consensus control strategy for continuous-time nonlinear multiagent systems (MASs) via adaptive dynamic programming (ADP). First, the control input is integrated into the feedback system for parallel control, facilitating an augmented system's optimal consensus control with an appropriate augmented performance index function to be established, which is identical to the original system's suboptimal control with a conventional performance index. Second, the feasibility of the proposed control scheme is evaluated based on the policy iteration algorithm, and the convergence of the algorithm is demonstrated. Then, an online learning algorithm becomes available to implement the ADP-based optimal parallel consensus control protocol without prior knowledge of the system. The Lyapunov approach is employed to indicate that the signals are convergent. Ultimately, the experimental data support the theoretical results.

Parallel Driving with Big Models and Foundation Intelligence in Cyber-Physical-Social Spaces.

Recent years have witnessed numerous technical breakthroughs in connected and autonomous vehicles (CAVs). On the one hand, these breakthroughs have significantly advanced the development of intelligent transportation systems (ITSs); on the other hand, these new traffic participants introduce more complex and uncertain elements to ITSs from the social space. Digital twins (DTs) provide real-time, data-driven, precise modeling for constructing the digital mapping of physical-world ITSs. Meanwhile, the metaverse integrates emerging technologies such as virtual reality/mixed reality, artificial intelligence, and DTs to model and explore how to realize improved sustainability, increased efficiency, and enhanced safety. More recently, as a leading effort toward general artificial intelligence, the concept of foundation model was proposed and has achieved significant success, showing great potential to lay the cornerstone for diverse artificial intelligence applications across different domains. In this article, we explore the big models embodied foundation intelligence for parallel driving in cyber-physical-social spaces, which integrate metaverse and DTs to construct a parallel training space for CAVs, and present a comprehensive elucidation of the crucial characteristics and operational mechanisms. Beyond providing the infrastructure and foundation intelligence of big models for parallel driving, this article also discusses future trends and potential research directions, and the "6S" goals of parallel driving.

Exceptionally low mercury concentrations and fluxes from the 2021 and 2022 eruptions of Fagradalsfjall volcano, Iceland.

Mercury (Hg) is naturally released by volcanoes and geothermal systems, but the global flux from these natural sources is highly uncertain due to a lack of direct measurements and uncertainties with upscaling Hg/SO2 mass ratios to estimate Hg fluxes. The 2021 and 2022 eruptions of Fagradalsfjall volcano, southwest Iceland, provided an opportunity to measure Hg concentrations and fluxes from a hotspot/rift system using modern analytical techniques. We measured gaseous Hg and SO2 concentrations in the volcanic plume by near-source drone-based sampling and simultaneous downwind ground-based sampling. Mean Hg/SO2 was an order of magnitude higher at the downwind locations relative to near-source data. This was attributed to the elevated local background Hg at ground level (4.0 ng m-3) likely due to emissions from outgassing lava fields. The background-corrected plume Hg/SO2 mass ratio (5.6 × 10-8) therefore appeared conservative from the near-source to several hundred meters distant, which has important implications for the upscaling of volcanic Hg fluxes based on SO2 measurements. Using this ratio and the total SO2 flux from both eruptions, we estimate the total mass of gaseous Hg released from the 2021 and 2022 Fagradalsfjall eruptions was 46 ± 33 kg, equivalent to a flux of 0.23 ± 0.17 kg d-1. This is the lowest Hg flux estimate in the literature for active open-conduit volcanoes, which range from 0.6 to 12 kg d-1 for other hotspot/rift volcanoes, and 0.5-110 kg d-1 for arc volcanoes. Our results suggest that Icelandic volcanic systems are fed from an especially Hg-poor mantle. Furthermore, we demonstrate that the aerial near-source plume Hg measurement is feasible with a drone-based active sampling configuration that captures all gaseous and particulate Hg species, and recommend this as the preferred method for quantifying volcanic Hg emissions going forward.

Nano-mineral assemblages in mercury- and silver-contaminated soils: records of sequestration, transformation, and release of mercury- and silver-bearing nanoparticles.

Mercury-bearing nano-mineral assemblages (Hg-NMAs) are chemically and mineralogically heterogeneous, micrometer-sized aggregates of nanoparticles (NPs) found in contaminated soils and sediments. Although these NMAs control sequestration and release of Hg that is a global contaminant, our understanding is limited with respect to the conditions of different types of Hg-NMAs, the diversity of its minerals, the size distribution of its NPs and whether mineral replacement and alteration reactions in these NMAs result in the release of Hg-bearing NPs. For this purpose, Hg-NMAs in four sediment samples from the Guanajuato Mining District (GMD) in Mexico, a region that was polluted by Hg and silver (Ag) due to historical mining involving Hg amalgamation, are characterized at the micro- and nanoscale. Microscale examinations with SEM show that the majority of Hg-NMAs occurs in mineral surface coatings (MSC) and fillings in fractures within quartz grains and are enriched in Hg and sulfur (S) relative to Ag, and in S and selenium (Se) relative to chloride (Cl). Examinations at the nanoscale show that Hg-NMAs contain (a) residuals of the patio process such as amalgam phases and elemental Ag; (b) associations of Hg- and Ag-sulfide NPs with pyrite and marcasite; (c) associations of Hg- and Ag-sulfide NPs with goethite and clay minerals along the rims of the MSC. The latter minerals replaced the Fe-Si-rich matrix at high-water rock ratios most likely due to an increase in porosity during flooding of the Pastita River. Consequently, the rims are depleted in Hg-Ag-sulfide NPs relative to the unaltered Fe-Si-rich matrices indicating that changes in the physiochemical conditions of soils and sediments in the GMD can result in the release of Hg-Ag-bearing NPs. In this context, this study discusses whether release and dissolution of Hg-Ag-bearing NPs contribute to the recently observed elevated gaseous elemental Hg concentrations in the soil, interstitial air and ambient air, and to the fate and effects of Hg in local aquatic environments.

Large mercury release from the Greenland Ice Sheet invalidated.

The major input of mercury (Hg) to the Arctic is normally ascribed to long-range transport of anthropogenic Hg emissions. Recently, alarming concentrations of Hg in meltwater from the Greenland Ice Sheet (GrIS) were reported with bedrock as the proposed source. Reported Hg concentrations were 100 to 1000 times higher than in known freshwater systems of Greenland, calling for independent validation of the extraordinary concentrations and conclusions. Here, we present measurements of Hg at 21 glacial outlets in West Greenland showing that extreme Hg concentrations cannot be reproduced. In contrast, we find that meltwater from below the GrIS is very low in Hg, has minor implications for the global Hg budget, and pose only a very limited risk for local communities and the natural environment of Greenland.

Knowledge-Embedded Mutual Guidance for Visual Reasoning.

Visual reasoning between visual images and natural language is a long-standing challenge in computer vision. Most of the methods aim to look for answers to questions only on the basis of the analysis of the offered questions and images. Other approaches treat knowledge graphs as flattened tables to search for the answer. However, there are two major problems with these works: 1) the model disregards the fact that the world we surrounding us interlinks our hearing and speaking of natural language and 2) the model largely ignores the structure of the acrlong KG. To overcome these challenging deficiencies, a model should jointly consider two modalities of vision and language, as well as the rich structural and logical information embedded in knowledge graphs. To this end, we propose a general joint representation learning framework for visual reasoning, namely, knowledge-embedded mutual guidance. It realizes mutual guidance not only between visual data and natural language descriptions but also between knowledge graphs and reasoning models. In addition, it exploits the knowledge derived from the reasoning model to boost knowledge graphs when applying the visual relation detection task. The experimental results demonstrate that the proposed approach performs dramatically better than state-of-the-art methods on two benchmarks for visual reasoning.

Deep Rank-Consistent Pyramid Model for Enhanced Crowd Counting.

Most conventional crowd counting methods utilize a fully-supervised learning framework to establish a mapping between scene images and crowd density maps. They usually rely on a large quantity of costly and time-intensive pixel-level annotations for training supervision. One way to mitigate the intensive labeling effort and improve counting accuracy is to leverage large amounts of unlabeled images. This is attributed to the inherent self-structural information and rank consistency within a single image, offering additional qualitative relation supervision during training. Contrary to earlier methods that utilized the rank relations at the original image level, we explore such rank-consistency relation within the latent feature spaces. This approach enables the incorporation of numerous pyramid partial orders, strengthening the model representation capability. A notable advantage is that it can also increase the utilization ratio of unlabeled samples. Specifically, we propose a Deep Rank-consistEnt pyrAmid Model (), which makes full use of rank consistency across coarse-to-fine pyramid features in latent spaces for enhanced crowd counting with massive unlabeled images. In addition, we have collected a new unlabeled crowd counting dataset, FUDAN-UCC, comprising 4000 images for training purposes. Extensive experiments on four benchmark datasets, namely UCF-QNRF, ShanghaiTech PartA and PartB, and UCF-CC-50, show the effectiveness of our method compared with previous semi-supervised methods. The codes are available at https://github.com/bridgeqiqi/DREAM.

SASAN: Shape-Adaptive Set Abstraction Network for Point-Voxel 3D Object Detection.

Point-voxel 3D object detectors have achieved impressive performance in complex traffic scenes. However, they utilize the 3D sparse convolution (spconv) layers with fixed receptive fields, such as voxel-based detectors, and inherit the fixed sphere radius from point-based methods for generating the features of keypoints, which make them weak in adaptively modeling various geometrical deformations and sizes of real objects. To tackle this issue, we propose a shape-adaptive set abstraction network (SASAN) for point-voxel 3D object detection. First, the proposal and offset generation module is adopted to learn the coordinates and confidences of 3D proposals and shape-adaptive offsets of the certain number of offset points for each voxel. Meanwhile, an extra offset supervision task is employed to guide the learning of shifting values of offset points, aiming at motivating the predicted offsets to preferably adapt to the various shapes of objects. Then, the shape-adaptive set abstraction module is proposed to extract multiscale keypoints features by grouping the neighboring offset points' features, as well as features learned from adjacent raw points and the 2-D bird-view map. Finally, the region of interest (RoI)-grid proposal refinement module is used to aggregate the keypoints features for further proposal refinement and confidence prediction. Extensive experiments on the competitive KITTI 3D detection benchmark demonstrate that the proposed SASAN gains superior performance as compared with state-of-the-art methods.

Comprehensive safety risk evaluation of fireworks production enterprises using the frequency-based ANP and BPNN.

The fireworks industry has long struggled with the problem of safety. Scientific, reasonable, and operable evaluation models are prerequisites of reducing risk. Based on the data from over 100 fireworks production safety accidents in China from 2010 to 2022, two evaluation models were established from the perspective of safety risk definition. Firstly, a weight calculation derivative method, the frequency-based analytic network process (ANP), was proposed creatively. This method optimized the importance ranking index calculation process in the ANP by considering the causal frequency of risk factors in the historical accident samples, thus determining how much each indicator affects the likelihood of accidents. Secondly, utilizing the historical accident samples as the dataset, a back propagation neural network (BPNN) model was developed to extract the mathematical relationship between each risk factor and the severity of accident consequence. Finally, the frequency-based ANP and BPNN models were combined to determine the safety risk level of the fireworks production enterprises. Meanwhile, the safety evaluation research samples were used as the comparison set for empirical study with historical accident samples, involving 100 fireworks production enterprises in China evaluated from 2017 to 2020. The significance result of zero shows that there is a statistically significant difference between the likelihood evaluation results of the accident and non-accident companies. Additionally, the severity evaluation model exhibits an excellent result, revealing a classification accuracy of 98.21 %, a mean square error of 8.97 × 10-4, a percent bias of 1.24 %, and a correlation coefficient and Nash-Sutcliffe efficiency coefficient both of 0.96. The frequency-based ANP and BPNN models integrate self-learning, self-adaptive, and fuzzy information processing, obtaining more accurate and objective evaluation results. This work provides a new strategy for the promotion and application of artificial intelligence in the field of safety risk evaluation, thus offering real-time safety risk evaluation and decision support of the safety management for the enterprises.

Toward parallel intelligence: An interdisciplinary solution for complex systems.

The growing complexity of real-world systems necessitates interdisciplinary solutions to confront myriad challenges in modeling, analysis, management, and control. To meet these demands, the parallel systems method rooted in the artificial systems, computational experiments, and parallel execution (ACP) approach has been developed. The method cultivates a cycle termed parallel intelligence, which iteratively creates data, acquires knowledge, and refines the actual system. Over the past two decades, the parallel systems method has continuously woven advanced knowledge and technologies from various disciplines, offering versatile interdisciplinary solutions for complex systems across diverse fields. This review explores the origins and fundamental concepts of the parallel systems method, showcasing its accomplishments as a diverse array of parallel technologies and applications while also prognosticating potential challenges. We posit that this method will considerably augment sustainable development while enhancing interdisciplinary communication and cooperation.

Can language models be used for real-world urban-delivery route optimization?

Language models have contributed to breakthroughs in interdisciplinary research, such as protein design and molecular dynamics understanding. In this study, we reveal that beyond language, representations of other entities, such as human behaviors, that are mappable to learnable sequences can be learned by language models. One compelling example is the real-world delivery route optimization problem. We here propose a novel approach based on the language model to optimize delivery routes on the basis of drivers' historical experiences. Although a broad range of optimization-based approaches have been designed to optimize delivery routes, they do not capture the implicit knowledge of complex delivery operating environments. The model we propose integrates this knowledge in the route optimization process by learning from driving behaviors in experienced drivers. A real-world delivery route that preserves drivers' implicit behavioral patterns is first analogized to a sentence in natural language. Through unsupervised learning, we then learn the vector representations of words and infer the drivers' delivery chains on the basis of the tailored chain-reaction-based algorithm. We also provide insights into the fusion of language models and operations research methods. In our approach, language models are applied to learn drivers' delivery behaviors and infer new deliveries at the delivery zone level, while the classic traveling salesman problem (TSP) model is embedded into the hybrid framework for intra-zone optimization. Numerical experiments performed on real-world data from Amazon's delivery service demonstrate that the proposed approach outperforms pure optimization, supporting the effectiveness, efficiency, and extensibility of our model. As a versatile approach, the proposed framework can easily be extended to various disciplines in which the data follow certain grammar rules. We anticipate that our work will serve as a stepping stone toward the understanding and application of language models in tackling interdisciplinary research problems.

Robust Optimal Parallel Tracking Control Based on Adaptive Dynamic Programming.

This article focuses on a novel robust optimal parallel tracking control method for continuous-time (CT) nonlinear systems subject to uncertainties. First, the designed virtual controller facilitates the transformation of the original nonlinear system into an affine system with an augmented state vector, which promotes the introduction of the optimal parallel tracking control problem. Then, this article generates fresh insight into counteracting the effects of uncertainty by developing a novel parallel control system that invokes the formulated virtual control law and an auxiliary variable obtained from the relationship between the solutions of the optimal control problems for the uncertain system and the nominal one. Next, critic neural networks (NNs) approximate the Hamilton-Jacobi-Bellman (HJB) equations' solution to implement the proposed robust optimal control method via adaptive dynamic programming (ADP). Finally, simulation experiments demonstrate the proposed method's remarkable effectiveness.

3D printing in materials manufacturing industry: A realm of Industry 4.0.

Additive manufacturing (AM), also known as 3D printing, is a new manufacturing trend showing promising progress over time in the era of Industry 4.0. So far, various research has been done for increasing the reliability and productivity of a 3D printing process. In this regard, reviewing the existing concepts and forwarding novel research directions are important. This paper reviews and summarizes the process flow, technologies, configurations, and monitoring of AM. It started with the general AM process flow, followed by the definitions and the working principles of various AM technologies and the possible AM configurations, such as traditional and robot-assisted AM. Then, defect detection, fault diagnosis, and open-loop and closed-loop control systems in AM are discussed. It is noted that introducing robots into the assisting mechanism of AM increases the reliability and productivity of the manufacturing process. Moreover, integrating machine learning and conventional control algorithms ensures a closed-loop control in AM, a promising control strategy. Lastly, the paper addresses the challenges and future trends.

Quality of Stormwater Infrastructure Systems in Vulnerable Communities: Three Case Studies from Texas.

A properly functioning local stormwater drainage system is essential for mitigating flood risks. This study evaluates the quality of roadside drainage channels in three underserved communities in Texas: the Sunnyside neighborhood in Houston (Harris County), a neighborhood in the City of Rockport (Aransas County), and the Hoehn colonia (Hidalgo County). These communities have a history of flooding, are highly socially vulnerable, and rely on roadside ditches as their principal stormwater drainage system for runoff control. Mobile lidar (Light Detection and Ranging) measurements were collected for 6.09 miles of roadside channels in these communities. The raw lidar measurements were processed to evaluate drainage conditions based on the channel's geometric properties, hydraulic capacity, and level of service. The assessment results are linked to a Geographic Information System (GIS) tool for enhanced visualization. Finally, the paper provides insights regarding the quality of stormwater infrastructure in the study communities and discusses their practical implications.

Modeling the Mercury Cycle in the Sea Ice Environment: A Buffer between the Polar Atmosphere and Ocean.

Sea ice (including overlying snow) is a dynamic interface between the atmosphere and the ocean, influencing the mercury (Hg) cycling in polar oceans. However, a large-scale and process-based model for the Hg cycle in the sea ice environment is lacking, hampering our understanding of regional Hg budget and critical processes. Here, we develop a comprehensive model for the Hg cycle at the ocean-sea ice-atmosphere interface with constraints from observational polar cryospheric data. We find that seasonal patterns of average total Hg (THg) in snow are governed by snow thermodynamics and deposition, peaking in springtime (Arctic: 5.9 ng/L; Antarctic: 5.3 ng/L) and minimizing during ice formation (Arctic: 1.0 ng/L, Antarctic: 0.5 ng/L). Arctic and Antarctic sea ice exhibited THg concentration peaks in summer (0.25 ng/L) and spring (0.28 ng/L), respectively, governed by different snow Hg transmission pathways. Antarctic snow-ice formation facilitates Hg transfer to sea ice during spring, while in the Arctic, snow Hg is primarily moved through snowmelt. Overall, first-year sea ice acts as a buffer, receiving atmospheric Hg during ice growth and releasing it to the ocean in summer, influencing polar atmospheric and seawater Hg concentrations. Our model can assess climate change effects on polar Hg cycles and evaluate the Minamata Convention's effectiveness for Arctic populations.

Tissue composition and storage duration affect the usefulness of generic wet-to-dry mass conversion factors in toxicology studies.

All ecosystems are exposed to a variety of anthropogenic contaminants. The potential threat posed by these contaminants to organisms has prompted scores of toxicology studies. Contaminant concentrations in wildlife toxicology studies are inconsistently expressed in wet or dry mass units, or even on a lipid-normalized basis, but tissue composition is rarely reported, and the conversion between dry and wet mass units, notably, is often based on assumed empirical moisture contents in tissues. However, diverse factors (e.g., tissue, storage conditions) may affect tissue composition and render comparisons between studies difficult or potentially biased. Here, we used data on the concentration of mercury, a global pollutant, in tissues of red foxes (Vulpes vulpes) to quantify the effects of diverse variables on moisture and lipid contents, and their consequences on contaminant concentration in different tissues, when converting between wet and dry mass units (lipid extracted or not). We found that moisture content differed largely between organs, enough to preclude the use of a single conversion factor, and decreased by 1% per year when stored at -80 °C. Although most fox tissues had low lipid concentrations, lipid content affected water content and their extraction affected the wet to dry mass conversion factor. We thus recommend reporting tissue composition (at least water and lipid contents) systematically in toxicology studies of mercury specifically and of contaminants in general, and using tissue/species specific conversion factors to convert between dry and wet mass concentration.

One Size Fits All: A Unified Traffic Predictor for Capturing the Essential Spatial-Temporal Dependency.

Traffic prediction is a keystone for building smart cities in the new era and has found wide applications in traffic scheduling and management, environment policy making, public safety, and so on. Instead of creating a traffic predictor for each city, this article focuses on designing a unified network model that could be directly applied for traffic prediction in any city, by learning the essential spatial-temporal dependencies, i.e., the mutual relationship between traffic and the corresponding fine-grained road network. To achieve this goal, this article proposes a joint knowledge-and data-driven mechanism that novelly divides dependencies into three kinds of correlations, i.e., road segment, intra-intersection, and inter-intersection correlation, which capture the microcosmic, middle, and macroscopic dependencies between traffic and the road network, respectively. Specifically, we first construct traffic datasets that could cover all road segments from real-world trajectory datasets, which makes it possible to model the whole road network as a graph, with the help of fine-grained road topology. Then, we propose meta road segment learner, connection-aware spatial-temporal graph convolutional network (GCN), and multiscale residual networks for capturing the microcosmic, middle, and macroscopic dependencies, respectively. Our experiments on three real-world datasets demonstrate that our proposed method could: 1) achieve better prediction accuracy compared with several approaches and 2) capture the mutual relationship between traffic and the fine-grained road network since our model trained only using data from the source city achieves good performance when it is directly applied for traffic prediction in the target city, without any fine-tuning. The codes will be made publicly available.

GLaLT: Global-Local Attention-Augmented Light Transformer for Scene Text Recognition.

Recent years have witnessed the growing popularity of connectionist temporal classification (CTC) and attention mechanism in scene text recognition (STR). CTC-based methods consume less time with few computational burdens, while they are not as effective as attention-based methods. To retain computational efficiency and effectiveness, we propose the global-local attention-augmented light Transformer (GLaLT), which adopts a Transformer-based encoder-decoder structure to orchestrate CTC and attention mechanism. The encoder integrates the self-attention module with the convolution module to augment the attention, where the self-attention module pays more attention to capturing long-term global dependencies and the convolution module focuses on local context modeling. The decoder consists of two parallel modules: one is the Transformer-decoder-based attention module and the other is the CTC module. The first one is removed in the testing phase and can guide the second one to extract robust features in the training phase. Extensive experiments on standard benchmarks demonstrate that GLaLT achieves state-of-the-art performance for both regular and irregular STR. In terms of tradeoffs, the proposed GLaLT is at or near the frontiers for maximizing speed, accuracy, and computational efficiency at the same time.

Webly Supervised Knowledge-Embedded Model for Visual Reasoning.

Visual reasoning between visual images and natural language remains a long-standing challenge in computer vision. Conventional deep supervision methods target at finding answers to the questions relying on the datasets containing only a limited amount of images with textual ground-truth descriptions. Facing learning with limited labels, it is natural to expect to constitute a larger scale dataset consisting of several million visual data annotated with texts, but this approach is extremely time-intensive and laborious. Knowledge-based works usually treat knowledge graphs (KGs) as static flattened tables for searching the answer, but fail to take advantage of the dynamic update of KGs. To overcome these deficiencies, we propose a Webly supervised knowledge-embedded model for the task of visual reasoning. On the one hand, vitalized by the overwhelming successful Webly supervised learning, we make much use readily available images from the Web with their weakly annotated texts for an effective representation. On the other hand, we design a knowledge-embedded model, including the dynamically updated interaction mechanism between semantic representation models and KGs. Experimental results on two benchmark datasets demonstrate that our proposed model significantly achieves the most outstanding performance compared with other state-of-the-art approaches for the task of visual reasoning.

Conservative-Progressive Collaborative Learning for Semi-Supervised Semantic Segmentation.

Pseudo supervision is regarded as the core idea in semi-supervised learning for semantic segmentation, and there is always a tradeoff between utilizing only the high-quality pseudo labels and leveraging all the pseudo labels. Addressing that, we propose a novel learning approach, called Conservative-Progressive Collaborative Learning (CPCL), among which two predictive networks are trained in parallel, and the pseudo supervision is implemented based on both the agreement and disagreement of the two predictions. One network seeks common ground via intersection supervision and is supervised by the high-quality labels to ensure a more reliable supervision, while the other network reserves differences via union supervision and is supervised by all the pseudo labels to keep exploring with curiosity. Thus, the collaboration of conservative evolution and progressive exploration can be achieved. To reduce the influences of the suspicious pseudo labels, the loss is dynamic re-weighted according to the prediction confidence. Extensive experiments demonstrate that CPCL achieves state-of-the-art performance for semi-supervised semantic segmentation.