Search Results (3,756)

The rapid evolution of intelligent vehicle technology has significantly advanced autonomous decision-making and driving safety. However, the challenge of predicting long-term trajectories in complex urban traffic persists, as traditional methodologies usually handle spatiotemporal attention mechanisms in isolation and are typically limited to short-term trajectory predictions. This paper proposes a Habit-based Temporal–Spatial Attention Long Short-Term Memory (HTSA-LSTM) network, a novel framework that integrates a dual spatiotemporal attention mechanism to capture dynamic dependencies across time and space, coupled with a driving style analysis module. The driving style analysis module employs Sparse Inverse Covariance Clustering and Spectral Clustering (SICC-SC) to extract driving primitives and cluster trajectory data, thereby revealing diverse driving behavior patterns without relying on predefined labels. By segmenting real-world driving data into fundamental behavioral units that reflect individual driving preferences, this approach enhances the model’s adaptability. These behavioral units, in conjunction with the spatiotemporal attention outputs, serve as inputs to the model, ultimately improving prediction accuracy and robustness in multi-vehicle scenarios. The model was evaluated by using the NGSIM dataset and real driving data from Wuhan, China. In comparison to benchmark models, HTSA-LSTM achieved a 20.72% reduction in the root mean square error (RMSE) and a 24.98% reduction in the negative log likelihood (NLL) for 5 s predictions of long-term trajectories. Furthermore, HTSA-LSTM achieved R² values exceeding 97.9% for 5 s predictions on highways and expressways and over 92.7% for 3 s predictions on urban roads, highlighting its excellent performance in long-term trajectory prediction and adaptability across diverse driving conditions. Full article

Lasers with average powers of several kilowatts have become an important tool for industrial applications. Temporal and spatial beam shaping was demonstrated to improve existing and enable novel applications. A very promising technology for both highly dynamic beam shaping and power scaling is the coherent combining of the beams of an array of high-power fundamental mode fibers. However, the limited number of fibers allows only limited spatial resolution of the common phase front. It is therefore favorable to work with plane or spherical common phase fronts, which generate a “point”, i.e., a diffraction pattern with a strong main lobe in the focal plane. By applying a tilt to the common phase front, points can be positioned in the focal plane with high spatial resolution. The Civan DBL 6–14 kW investigated in this work allows switching between positions of the points with 80 MHz. Sequences of points can be used to create arbitrary shapes. The time constants of points and shapes are very critical for this type of shape generation. The current paper analyzes the relevant time constants for setting points and creating shapes and relates them to time constants in laser processes. This is mandatory to deterministically influence laser processes. Full article

In recent years, Out-of-Stock (OOS) occurrences have posed a persistent challenge for both retailers and manufacturers. In the context of grocery retail, an OOS event represents a situation where customers are unable to locate a specific product when attempting to make a purchase. This study analyzes the issue from the manufacturer’s perspective. The proposed system, named the “Multi-modal yield eSTimation System of in-prOmotion Commercial Key-ProductS” (MySTOCKS) platform, is a sophisticated multi-modal yield estimation system designed to optimize inventory forecasting for the agrifood and large-scale retail sectors, particularly during promotional periods. MySTOCKS addresses the complexities of inventory management in settings where Out-of-Stock (OOS) and Surplus-of-Stock (SOS) situations frequently arise, offering predictive insights into final stock levels across defined forecasting intervals to support sustainable resource management. Unlike traditional approaches, MySTOCKS leverages an advanced deep learning framework that incorporates transformer models with self-attention mechanisms and domain adaptation capabilities, enabling accurate temporal and spatial modeling tailored to the dynamic requirements of the agrifood supply chain. The system includes two distinct forecasting modules: TR1, designed for standard stock-level estimation, and TR2, which focuses on elevated demand periods during promotions. Additionally, MySTOCKS integrates Elastic Weight Consolidation (EWC) to mitigate the effects of catastrophic forgetting, thus enhancing predictive accuracy amidst changing data patterns. Preliminary results indicate high system performance, with test accuracy, sensitivity, and specificity rates approximating 93.8%. This paper provides an in-depth examination of the MySTOCKS platform’s modular structure, data-processing workflow, and its broader implications for sustainable and economically efficient inventory management within agrifood and large-scale retail environments. Full article

In intralogistics, autonomous mobile robots (AMRs) operate without predefined paths, leading to complex traffic patterns and potential conflicts that impact system efficiency. This paper proposes a bio-inspired optimization method for autonomously generating spatial movement constraints for autonomous mobile robots (AMRs). Unlike traditional multi-agent pathfinding (MAPF) approaches, which focus on temporal coordination, our approach proactively reduces conflicts by adapting a weighted directed grid graph to improve traffic flow. This is achieved through four mechanisms inspired by ant colony systems: (1) a movement reward that decreases the weight of traversed edges, similar to pheromone deposition, (2) a delay penalty that increases edge weights along delayed paths, (3) a collision penalty that increases weights at conflict locations, and (4) an evaporation mechanism that prevents premature convergence to suboptimal solutions. Compared to the existing approaches, the proposed approach addresses the entire intralogistic problem, including plant layout, task distribution, release and dispatching algorithms, and fleet size. Its autonomous movement rule generation and low computational complexity make it well suited for dynamic intralogistic environments. Validated through physics-based simulations in Gazebo across three scenarios, a standard MAPF benchmark, and two industrial environments, the movement constraints generated using the proposed method improved the system throughput by up to 10% compared to unconstrained navigation and up to 4% compared to expert-designed solutions while reducing the need for conflict-resolution interventions. Full article

Recently, we introduced random complex and phase screen methods as powerful tools for numerically investigating the evolution of partially coherent pulses (PCPs) in nonlinear dispersive media. However, these methods are restricted to the Schell model type. Non-Schell model light has attracted growing attention in recent years for its distinctive characteristics, such as self-focusing, self-shifting, and non-diffraction properties as well as its critical applications in areas such as particle trapping and information encryption. In this study, we incorporate the Monte Carlo method into the pseudo-mode superposition method to derive the random electric field of any PCPs, including non-Schell model pulses (nSMPs). By solving the nonlinear Schrödinger equations through numerical simulations, we systematically explore the propagation dynamics of nSMPs in nonlinear dispersive media. By leveraging the nonlinearity and optical coherence, this approach allows for effective control over the focal length, peak power, and full width at half the maximum of the pulses. We believe this method offers valuable insights into the behavior of coherence-related phenomena in nonlinear dispersive media, applicable to both temporal and spatial domains. Full article

In response to the increasing safety concerns posed by low, slow, and small unmanned aerial vehicles (UAVs), the use of flexible nets for interception emerges as a promising solution due to its high tolerance, minimal requirements, and cost-effectiveness. To enhance the effectiveness of the flexible net capture system for these types of UAVs, an optimization of the system’s parameters is conducted. A dynamic model of the flexible net capture system is developed, and its deployment process is simulated and analyzed through a combination of ABAQUS 2022/Explicit and MATLAB R2020b software. The coverage rate and hang time are proposed as the key performance indicators for quantitatively assessing the interception capabilities of the rope net. A mathematical model is formulated to optimize the capture system parameters, considering both spatial and temporal tolerances. The Multi-objective Wolf Pack Algorithm, which incorporates an Elite Leadership Strategy and a crowding distance-based population update mechanism, is utilized to optimize the design variables. This approach leads to the derivation of the optimized design parameters for the flexible net. Ultimately, the optimal parameter configuration for the flexible net capture system is achieved through the application of the Multi-objective Wolf Pack Algorithm to the design variables. This optimization ensures the system’s peak performance in intercepting low, slow, and small UAVs. Full article

Long-term runoff projection and uncertainty estimates can provide both the changing trends and confidence intervals of water resources, provide basic information for decision makers, and reduce risks for water resource management. In this paper, a grid-based runoff projection and uncertainty framework was proposed through input selection and long short-term memory (LSTM) modelling coupled with uncertainty analysis. We simultaneously considered dynamic variables and static variables in the candidate input combinations. Different input combinations were compared. We employed LSTM to develop a relationship between monthly runoff and the selected variables and demonstrated the improvement in forecast accuracy through comparison with the MLR, RBFNN, and RNN models. The LSTM model achieved the highest mean Kling–Gupta Efficiency (KGE) score of 0.80, representing respective improvements of 45.45%, 33.33%, and 2.56% over the other three models. The uncertainty sources originating from the parameters of the LSTM models were considered, and the Monte Carlo approach was used to provide uncertainty estimates. The framework was applied to the Yellow River Source Area (YRSR) at the 0.25° grid scale to better show the temporal and spatial features. The results showed that extra information about static variables can improve the accuracy of runoff projections. Annual runoff tended to increase, with projection ranges of 148.44–296.16 mm under the 95% confidence level, under various climate scenarios. Full article

Algal bloom is a major ecological and environmental problem caused by abnormal algal reproduction in water, and it poses a serious threat to the aquatic ecosystem, drinking water safety, and public health. Because of the high dynamic and spatiotemporal heterogeneity of bloom outbreaks, the process often presents significant changes in a short time. Therefore, it has important scientific research value and practical application significance to construct an accurate and effective bloom warning model. This study constructs an integrated model combining sequence features, attention mechanisms, and random forest using machine learning algorithms for bloom prediction, based on watercolor geostationary satellite observations and meteorological data from GOCI in South Korea. In the process, high spatial resolution Sentinel-2 satellite data is also utilized for sample extraction. With a 10-m resolution, Sentinel-2 provides more precise spatial information compared to the 500-m resolution of GOCI, which significantly enhances the accuracy of the model, especially in monitoring local water body changes. The experimental results demonstrate that the model exhibits excellent accuracy and stability in the spatiotemporal prediction of water blooms. The average AUC value is 0.88, the F1 score is 0.72, and the accuracy is 0.79 when identifying the dynamic change of water bloom on the hourly scale. At the same time, this study summarized four typical diurnal change modes of effluent bloom, including dispersal mode, persistent outbreak mode, dispersal-regression mode, and subsidence mode, revealing the main characteristics of diurnal dynamic change of bloom. The research results provided strong technical support for water environment monitoring and water quality safety management and showed a good application prospect. Full article

In wireless sensor networks, sensors often collect and transmit a large amount of redundant data, which can lead to excessive battery consumption and subsequent performance degradation. To solve this problem, this paper proposes a Zoom-In Zoom-Out (ZIZO) method based on sparse vectors (SV-ZIZO). It operates in two parts: At the sensor level, given the temporal similarity of the data, a new compression method based on the sparse vector representation of segmented regions is proposed. This method can not only effectively ensure the compression ratio but also improve the accuracy of data restoration. At the cluster-head (CH) level, by utilizing the spatial similarity of the data, the fuzzy clustering theory is introduced to put some sensors into hibernation mode, thereby reducing data transmission. Meanwhile, the sampling frequency of the sensors is dynamically adjusted by calculating the redundancy rate of the collected periodic data. The experimental results show that compared with other existing methods, the algorithm proposed in this paper increases the data compression ratio by 21.8% and can reduce energy consumption by up to 95%. Full article

While existing building energy prediction methods have advanced significantly, they face fundamental challenges in simultaneously modeling complex spatial–temporal relationships between buildings and integrating dynamic weather patterns, particularly in dense urban environments where building interactions significantly impact energy consumption patterns. This study presents an advanced deep learning system combining temporal graph neural networks with weather data parameters to enhance prediction accuracy across diverse building types through innovative spatial–temporal modeling. This approach integrates LSTM layers with graph convolutional networks, trained using energy consumption data from 150 commercial buildings over three years. The system incorporates spatial relationships through a weighted adjacency matrix considering building proximity and operational similarities, while weather parameters are integrated via a specialized neural network component. Performance evaluation examined normal operations, data gaps, and seasonal variations. The results demonstrated a 3.2% mean absolute percentage error (MAPE) for 15 min predictions and a 4.2% MAPE for 24 h forecasts. The system showed robust data recovery, maintaining 95.8% effectiveness even with 30% missing values. Seasonal analysis revealed consistent performance across weather conditions (MAPE: 3.1–3.4%). The approach achieved 33.3% better prediction accuracy compared to conventional methods, with 75% efficiency across four GPUs. These findings demonstrate the effectiveness of combining spatial relationships and weather parameters for building energy prediction, providing valuable insights for energy management systems and urban planning. The system’s performance and scalability make it particularly suitable for practical applications in smart building management and urban sustainability. Full article

Bioconvection can be defined as an aggregation pattern-generating phenomenon characterized by the collective behavior of swimming microorganisms in response to external influences and self-induced fluid flow, and it has attracted significant interest in the study of microalgae. This review aims to provide a comprehensive examination of this phenomenon in a concentrated population of suspended microalgae definable as nonlinear dynamical systems (i.e., generators of spatial and temporal patterns), describing and clarifying the underlying mechanisms, ecological implications, and possible biotechnological applications in different fields for the improvement of bioreactor design for biomass cultivation, wastewater treatment, or bioremediation. The most influential models used to capture the complexity of bioconvection are perused in order to explore the intricate interplay of internal and external influences (biological, physical, and environmental) governing pattern formation dynamics alongside recent advancements in modeling techniques and experimental approaches. Full article

Climate change is driving the restructuring of global biological communities. As a species sensitive to climate change, studying the response of small rodents to climate change is helpful to indirectly understand the changes in ecology and biodiversity in a certain region. Here, we use the MaxEnt (maximum entropy) model to predict the distribution patterns, main influencing factors, and range changes of various small rodents in the Ordos desert steppe in China under different climate change scenarios in the future (2050s: average for 2041–2060). The results show that when the parameters are FC = LQHPT, and RM = 4, the MaxEnt model is optimal and AUC = 0.833. We found that NDVI (normalized difference vegetation index), Bio 12 (annual precipitation), and TOC (total organic carbon) are important driving factors affecting the suitability of the small rodent habitat distribution in the region. At the same time, the main influencing factors were also different for different rodent species. We selected 4 dominant species for analysis and found that, under the situation of future climate warming, the high-suitability habitat area of Allactaga sibirica and Phodopus roborovskii will decrease, while that of Meriones meridianus and Meriones unguiculatus will increase. Our research results suggest that local governments should take early preventive measures, strengthen species protection, and respond to ecological challenges brought about by climate change promptly. Full article

As the frequency of disasters increases worldwide, it has become increasingly important to raise awareness of the risks and mitigate their effects through effective disaster management. Anticipating disaster risks and ensuring timely evacuations are crucial. This paper proposes SafeWitness, which dynamically captures the evolving characteristics of disasters by integrating crowdsensing and GIS-based geofencing. It not only enables real-time disaster awareness and evacuation support but also provides spatial context awareness by mapping the disaster area based on GIS road information and temporal context awareness by using crowdsensing to track the progress of the disaster. This approach increases the effectiveness of disaster management by providing explicit, data-driven insights for timely decision making and risk mitigation. The experimental results reveal that the proposed method improved the F1-scores in the hazard and warning zones compared to the domain-based approach. The result increased by 12% in the hazard zone and by 55% in the warning zone compared to the traditional technique. Through user sampling, we enhanced the SafeWitness F1-score in the hazard zone by 6 times and in the warning zone by 2.8 times compared to the method without user sampling. In conclusion, SafeWitness offers a more precise perception of disaster areas than traditional domain-based area definitions, and the experimental results demonstrate the effectiveness of user sampling. Decision-makers and disaster management professionals can use the proposed method in urban disaster scenarios. Full article

There is a close correlation between brain aging and age. However, traditional neural networks cannot fully capture the potential correlation between age and brain aging due to the limited receptive field. Furthermore, they are more concerned with deep spatial semantics, ignoring the fact that effective temporal information can enrich the representation of low-level semantics. To address these limitations, a local attention spatio-temporal graph inference network (LSTGINet) was developed to explore the details of the association between age and brain aging, taking into account both spatio-temporal and temporal perspectives. First, multi-scale temporal and spatial branches are used to increase the receptive field and model the age information simultaneously, achieving the perception of static correlation. Second, these spatio-temporal feature graphs are reconstructed, and large topographies are constructed. The graph inference node aggregation and transfer functions fully capture the hidden dynamic correlation between brain aging and age. A new local attention module is embedded in the graph inference component to enrich the global context semantics, establish dependencies and interactivity between different spatio-temporal features, and balance the differences in the spatio-temporal distribution of different semantics. We use a newly designed weighted loss function to supervise the learning of the entire prediction framework to strengthen the inference process of spatio-temporal correlation. The final experimental results show that the MAE on baseline datasets such as CamCAN and NKI are 6.33 and 6.28, respectively, better than the current state-of-the-art age prediction methods, and provides a basis for assessing the state of brain aging in adults. Full article

To address the limitations of traditional two-stream networks, such as inadequate spatiotemporal information fusion, limited feature diversity, and insufficient accuracy, we propose an improved two-stream network for human action recognition based on multi-scale attention Transformer and 3D convolutional (C3D) fusion. In the temporal stream, the traditional 2D convolutional is replaced with a C3D network to effectively capture temporal dynamics and spatial features. In the spatial stream, a multi-scale convolutional Transformer encoder is introduced to extract features. Leveraging the multi-scale attention mechanism, the model captures and enhances features at various scales, which are then adaptively fused using a weighted strategy to improve feature representation. Furthermore, through extensive experiments on feature fusion methods, the optimal fusion strategy for the two-stream network is identified. Experimental results on benchmark datasets such as UCF101 and HMDB51 demonstrate that the proposed model achieves superior performance in action recognition tasks. Full article