Spatio-temporal Dynamic Model Research Articles

TG-PGAT: An AIS Data-Driven Dynamic Spatiotemporal Prediction Model for Ship Traffic Flow in the Port

Accurate prediction of ship traffic flow is essential for developing intelligent maritime transportation systems. To address the complexity of ship traffic flow data in the port and the challenges of capturing its dynamic spatiotemporal dependencies, a dynamic spatiotemporal model called Temporal convolutional network-bidirectional Gated recurrent unit-Pearson correlation coefficient-Graph Attention Network (TG-PGAT) is proposed for predicting traffic flow in port waters. This model extracts spatial features of traffic flow by combining the adjacency matrix and spatial dynamic coefficient correlation matrix within the Graph Attention Network (GAT) and captures temporal features through the concatenation of the Temporal Convolutional Network (TCN) and Bidirectional Gated Recurrent Unit (BiGRU). The proposed TG-PGAT model demonstrates higher prediction accuracy and stability than other classic traffic flow prediction methods. The experimental results from multiple angles, such as ablation experiments and robustness tests, further validate the critical role and strong noise resistance of different modules in the TG-PGAT model. The experimental results of visualization demonstrate that this model not only exhibits significant predictive advantages in densely trafficked areas of the port but also outperforms other models in surrounding areas with sparse traffic flow data.

Prediction of spatiotemporal dynamics using deep learning: Coupled neural networks of long short-terms memory, auto-encoder and physics-informed neural networks

Several classic reaction-diffusion models using partial differential equations (PDEs) have been established to elucidate the formation mechanism of vegetation patterns. However, predictive modeling of complex spatiotemporal dynamics using traditional numerical methods can be significantly challenging in many practical scenarios. Physics-Informed Neural Networks (PINNs) provide a new approach to predict the solution of PDEs. However, the generalization of PINNs is not satisfactory when pretrained PINNs is directly used in non-trained space (defined as explorations). This may be attributed to the lack of training in the time dimension. Therefore, a framework (LA-PINNs) is proposed to predict the evolutionary solution of the non-dimensional vegetation-sand model. The framework couples neural networks of Long-Short Terms Memory, Auto-Encoder and Physics-Informed Neural Networks. The predictions of LA-PINNs are much better than those of PINNs. Then we studied the effects of hyperparameters on the accuracy of predictions. Based on training in time dimension by LSTM module and pretraining for quick-training strategy, LA-PINNs can improve the accuracy of explorations.

Dynamic Spatiotemporal Scheduling for Construction Building Projects

For building projects, the manager is responsible for coordinating the work of subcontractors at the construction site. This includes operations, material flows, and storage. In summary, one of their main roles is to ensure smooth team rotation, maintain fluid circulation, and avoid congestion or relaxation on the site. However, traditional tools lack the ability to consider the planning and management of worksite spaces when calculating the execution schedule and critical path. Consequently, three-week planning is usually carried out separately on independent plans, often using spreadsheets. In addition, a construction site is highly dynamic and mobile in nature, and the positioning of resources and workers can change daily. This makes the management of available space even more complex, and effective space management becomes an imperative. To address this challenge, this paper develops visual dynamic artifacts that present different operation types. The methodology and the conceptual framework facilitate the calculation of the Occupancy Rate (OR) that enables construction project managers to create simple yet dynamic spatiotemporal models of the construction schedule. By incorporating factors such as crew turnover and occupancy evolution, managers can simplify the calculation process and effectively optimize construction work by utilizing site occupancy rates. In summary, this paper presents the Dynamic Model of the Occupancy Rate Schedule (DMORS), a methodology developed through design science. This model utilizes created artifacts representing various operation types to ensure accurate calculations of dynamic occupancy by floor and sector in a site. Consequently, it enables the construction of a more realistic schedule based on critical space ideologies. The DMORS enables managers to use the OR for different floors and sectors of a site, allowing for better space management. A proof of concept demonstrates that this tool can enhance the efficiency and productivity of construction projects by optimizing crew schedules and resource allocation based on site OR.

Multidimensional effects of history, neighborhood, and proximity on urban land growth: A dynamic spatiotemporal rolling prediction model (STRM)

Multidimensional effects of history, neighborhood, and proximity on urban land growth: A dynamic spatiotemporal rolling prediction model (STRM)

A deep reinforcement learning with dynamic spatio-temporal graph model for solving urban logistics delivery planning problems

ABSTRACT The urban logistics delivery planning problems are a crucial component of urban spatial decision analysis. Most studies typically focus on traditional urban logistics delivery planning problems and ignore real-time traffic information. With the advancement of urbanization, real-time traffic networks play a critical role. However, previous studies have utilized heuristic methods to solve urban logistics delivery planning with real-time traffic information problems, and few studies have applied deep reinforcement learning methods to tackle this problem. Deep reinforcement learning methods solving traditional logistics delivery planning problems overlook the impact of dynamic spatio-temporal features on route planning. In this study, we propose a new deep reinforcement method called DRLDSTG. The method introduces the dynamic spatio-temporal graph model into a deep reinforcement learning method to capture these dynamic features from urban logistics delivery planning tasks. The actor-critic with maximum entropy method is employed to train the model and determine the optimal policy function. The experimental results indicated that the proposed method can achieve a superior solution with faster computational efficiency compared to commercial software and heuristic methods. Compared to other deep reinforcement learning methods, our method can more effectively learn dynamic spatio-temporal features from environments, demonstrating promising applications in cities.

Modeling the effects of PM pollutant on people using a spatio-temporal dynamic model and GIS

Modeling the effects of PM pollutant on people using a spatio-temporal dynamic model and GIS

Modeling of spatiotemporal dynamics of ligand-coated particle flow in targeted drug delivery processes

Nanoparticles tethered with vasculature-binding epitopes have been used to deliver the drug into injured or diseased tissues via the bloodstream. However, the extent that blood flow dynamics affects nanoparticle retention at the target site after adhesion needs to be better understood. This knowledge gap potentially underlies significantly different therapeutic efficacies between animal models and humans. An experimentally validated mathematical model that accurately simulates the effects of blood flow on nanoparticle adhesion and retention, thus circumventing the limitations of conventional trial-and-error-based drug design in animal models, is lacking. This paper addresses this technical bottleneck and presents an integrated mathematical method that derives heavily from a unique combination of a mechanics-based dispersion model for nanoparticle transport and diffusion in the boundary layers, an asperity model to account for surface roughness of endothelium, and an experimentally calibrated stochastic nanoparticle-cell adhesion model to describe nanoparticle adhesion and subsequent retention at the target site under external flow. PLGA-b-HA nanoparticles tethered with VHSPNKK peptides that specifically bind to vascular cell adhesion molecules on the inflamed vascular wall were investigated. The computational model revealed that larger particles perform better in adhesion and retention at the endothelium for the particle sizes suitable for drug delivery applications and within physiologically relevant shear rates. The computational model corresponded closely to the in vitro experiments which demonstrates the impact that model-based simulations can have on optimizing nanocarriers in vascular microenvironments, thereby substantially reducing in vivo experimentation as well as the development costs.

A charge-discharge optimization strategy considering the spatiotemporal distribution of electric vehicles and the operational safety of the power distribution network in the power-transportation coupling network

An electric vehicle (EV) charge-discharge optimization (CDO) strategy that accommodates both grid-side and user-side demands is conducive to mitigating the adverse effects of disordered EV charging on the power distribution network (PDN). To tackle the issue of inaccurate estimation of the schedulable capacity of EVs in existing research, a high temporal resolution dynamic spatiotemporal distribution simulation model for EVs is developed. Furthermore, leveraging the characteristics of the power-transportation coupling network (PTCN), a sub-districted dynamic electricity pricing (DEP) is proposed to assist in improving the PDN node voltage distribution. Subsequently, an incentive coefficient is introduced to incentivize the discharge of EVs during potential peak periods. The proposed CDO strategy, considering the non-cooperative behavior of EV owners (EVOs), can provide personalized charge-discharge plans for them. The temporal sequence experiments reveal a significant decrease in the peak-to-valley disparity of the load, achieving an optimization rate of load mean square error that exceeds 80%. The sub-districted DEP exhibits a significant advantage in maintaining PDN voltage stability compared to traditional time-of-use electricity pricing (TOUEP) and DEP, while also achieving reductions in active power losses of more than 15% and 10% for PDN, respectively. Moreover, in comparison to the conventional scenario, the CDO strategy leads to a maximum reduction of up to 942.5% and 22.3% in the total economic cost for electric private cars and electric taxis respectively. Lastly, the impact of seasonal factors is discussed. Numerical results indicate effective alleviation of load peak-valley differences, load fluctuations, and voltage drop phenomena in both winter and summer. Additionally, the maximum reduction in active power losses reaches 20.1% and 21.9%, respectively, while the total economic cost for EVOs participating in the CDO strategy is reduced by up to 97.7% and 114.7%.

Web-based spatial decision support system for precision agriculture: A tool for delineating dynamic management unit zones (MUZs)

Precision agriculture (PA) is a farming management concept that aims to provide agronomic, economic, and environmental benefits. One of the fields of research in PA is the delineation of Management Unit Zones (MUZs). MUZs are the sub-division of fields featuring an inter-zonal variation delineated by agronomists for on-field PA operations. To develop MUZs, three factors typically need to be considered: input multi-dimensional data, procedures to process the information, and the optimal number of zones a field should be divided into. PA uses digital technologies to collect and analyze a large amount of data, outline MUZs, monitor crops, and carry out site-specific crop management. Web-based spatial decision support systems (WB-SDSS) can provide users with tools that ease the complex procedures for PA. The objective of this study is twofold: on the one hand, we developed a free and open source (FOSS) WB-SDSS to facilitate the implementation and use of such tools for delineating MUZs and monitoring crops; on the other, a MUZs outline procedure was developed based on Sentinel-2 and Planetscope time series data, and spatio-temporal dynamic clustering model using fuzzy c-means. Our study highlighted that the WB-SDSS might be a helpful solution for harmonizing data collected from different sources, easing the implementation and use of complex geospatial procedures for PA, and delineating MUZs. We tested the system on a particularly representative farm in the Emilia Romagna region (Northern Italy), with 512 hectares of durum wheat crops. Using the WB-SDSS, we quickly delineated homogeneous zones for 27 fields in the study area during the phenological cycle of durum wheat (November 2018-June 2019).

Modelling foraging strategies of honey bees as agents in a dynamic landscape representation

Introduction: Both intrinsic colony mechanisms and external environmental variables affect the honey bee colony development rates and response and a key aspect of this is the use of resources within the landscape by honey bees. Although several models have been developed to explore the foraging behaviour of bees, none fully considered the spatial and temporal dynamics of landscape resources and the role of the colony in the process. Methodology: Here, we introduce a new honey bee foraging model being developed as a part of the ApisRAM honey bee colony model. Based on agent-based modelling, we used a dynamic ALMaSS landscape model enhanced with floral resource modelling to assess the impacts of weather conditions and resource availability on the possible foraging behaviour of honey bees. Several possible mechanisms (defined, based on honey bee traits) for scouting and foraging were investigated, separately for nectar and pollen collection, including prioritising foraging polygons for nectar foraging according to their distance to the colony, the quality or the energetic efficiency and, for pollen foraging, according to their distance to the colony and pollen quantity. Results: If model foraging bees prioritised the polygons, based on their distance from the colony, the number of unsuccessful flights increased compared to other tested strategies and the total amount of sugar collected showed a high variability. Contrary to expectations, the energetic efficiency strategy did not provide the colony with the highest amount of sugar. Overall, the tested strategies provide different outcomes on the collection of resources, the number of performed flights and their success rate, evidencing that the model's outcome at the colony level arises from the individual types of behaviour. Conclusions and Relevance: Variability in the mass of collected nectar and pollen was found mostly when scout bees applied the distance strategy. This higher variability in sugar collection means that model bees were not able to find the most profitable foraging sites at the landscape level, but only at the local level. Other strategies showed less dependence on the surrounding landscape (i.e. quality or random), but it comes at a cost (i.e. lower production for both nectar and pollen collection). These outputs help us evaluate which strategies could be used for future model development and confirm the models' ability to create dynamic responses. These responses at the colony level were only possible thanks to the implementation of a dynamic landscape model and dynamic spatiotemporal resource model, as well as implementing a social communication mechanism for bees to share information about the resources. Plant nectar production and quality information is essential to predict honey bee foraging distribution. In future model development, the implementation of pollen quality should also be explored to evaluate if it influences the overall pollen collection.

The study of DDPG based spatiotemporal dynamic deployment optimization of Air-Ground ad hoc network for disaster emergency response

In situations where natural disasters damage public communication networks, self-organized emergency communication networks play a vital role as important resources for disaster monitoring and emergency response. Geographical conditions, communication capacity, power availability, terminals’ position changes during disasters, and data volume, on the other hand, have a direct impact on emergency terminals’ effectiveness in disaster monitoring and data processing. As a result, the ability to detect calamities quickly and give real-time reactions suffers. To ensure effective emergency communication coverage when public networks are disrupted, intelligent, rapid and reliable optimization of the spatiotemporal dynamic deployment of Ad Hoc Network (ANET) is crucial for disaster monitoring and emergency response. The Deep Deterministic Policy Gradient (DDPG) method is used in this study to assist the spatiotemporal dynamic deployment of air-ground ANET (AGANET) nodes, including unmanned aerial vehicle (UAV) ANET nodes and ground-based ANET nodes. Its goal is to develop a reliable model for ensuring the deployment of AGANET nodes for multi-terminal disaster monitoring. The paper describes the spatiotemporal dynamic deployment of AGANET nodes and develops a reinforcement learning model. Subsequently, it describes a DDPG reinforcement-based optimization method for the spatiotemporal dynamic deployment of AGANET nodes. Specifically, this method includes a greedy matching strategy based on real-time environmental information of AGANET nodes and ground-based sensing terminals, a DDPG AGANET node three-dimensional initialization algorithm, and a DDPG dynamic solution algorithm for spatiotemporal dynamic deployment reinforcement learning model of AGANET. AGANET nodes and ground-based sensing terminals can all work together in this way to provide high-quality emergency AGANET services. When compared to traditional methods, it offers better communication dependability, enhanced efficiency, and cheaper costs associated with building emergency networks. It is also crucial for emergency response in instances where the public network fails.

Modelling EEG Dynamics with Brain Sources

An electroencephalogram (EEG), recorded on the surface of the scalp, serves to characterize the distribution of electric potential during brain activity. This method finds extensive application in investigating brain functioning and diagnosing various diseases. Event-related potential (ERP) is employed to delineate visual, motor, and other activities through cross-trial averages. Despite its utility, interpreting the spatiotemporal dynamics in EEG data poses challenges, as they are inherently subject-specific and highly variable, particularly at the level of individual trials. Conventionally associated with oscillating brain sources, these dynamics raise questions regarding how these oscillations give rise to the observed dynamical regimes on the brain surface. In this study, we propose a model for spatiotemporal dynamics in EEG data using the Poisson equation, with the right-hand side corresponding to the oscillating brain sources. Through our analysis, we identify primary dynamical regimes based on factors such as the number of sources, their frequencies, and phases. Our numerical simulations, conducted in both 2D and 3D, revealed the presence of standing waves, rotating patterns, and symmetric regimes, mirroring observations in EEG data recorded during picture naming experiments. Notably, moving waves, indicative of spatial displacement in the potential distribution, manifested in the vicinity of brain sources, as was evident in both the simulations and experimental data. In summary, our findings support the conclusion that the brain source model aptly describes the spatiotemporal dynamics observed in EEG data.

Analysis of sustainable water resource management and driving mechanism in arid region: a case study of Xinjiang, China, from 2005 to 2020.

The long-term dynamic comprehensive evaluation of the water resource carrying capacity (WRCC) and the analysis of its potential driving mechanism in arid areas are contemporary research issues and technical means of mitigating and coordinating the conflict between severe resource shortages and human needs. The purpose of this study was to explore the distribution of the WRCC and the spatiotemporal heterogeneity of drivers in arid areas based on an improved two-dimensional spatiotemporal dynamic evaluation model. The results show that (1) the spatial distribution of the WRCC in Xinjiang, China, is high in the north, low in the south, high in the west, and low in the east. (2) From 2005 to 2020, the centers of gravity of the WRCC in northern and southern Xinjiang moved to the southeast and west, respectively, and the spatial distribution exhibited slight diffusion. (3) The factors influencing the WRCC exhibit more obvious spatial and temporal heterogeneity. The domestic waste disposal rate and ecological water use rate were the main factors influencing the WRCC in the early stage, while the GDP per capita gradually played a dominant role in the later stage. (4) In the next 30 years, the WRCC in Xinjiang will increase. The results provide a theoretical reference for the sustainable development of water resources in arid areas.

A Novel Method to Identify Mild Cognitive Impairment Using Dynamic Spatio-Temporal Graph Neural Network.

Resting-state functional magnetic resonance imaging (rs-fMRI) has been widely used in the identification of mild cognitive impairment (MCI) research, MCI patients are relatively at a higher risk of progression to Alzheimer's disease (AD). However, almost machine learning and deep learning methods are rarely analyzed from the perspective of spatial structure and temporal dimension. In order to make full use of rs-fMRI data, this study constructed a dynamic spatiotemporal graph neural network model, which mainly includes three modules: temporal block, spatial block, and graph pooling block. Our proposed model can extract the BOLD signal of the subject's fMRI data and the spatial structure of functional connections between different brain regions, and improve the decision-making results of the model. In the study of AD, MCI and NC, the classification accuracy reached 83.78% outperforming previously reported, which manifested that our model could effectively learn spatiotemporal, and dynamic spatio-temporal method plays an important role in identifying different groups of subjects. In summary, this paper proposed an end-to-end dynamic spatio-temporal graph neural network model, which uses the information of the temporal dimension and spatial structure in rs-fMRI data, and achieves the improvement of the three classification performance among AD, MCI and NC.

Self-organization of nested patterns in a coupled spatiotemporal phytoplankton-zooplankton system

This research theoretically investigates the emergence of nested structures in plankton patchiness. Considering the vertical migration of plankton populations, we develop a spatiotemporal dynamic model that describes the coupling of two phytoplankton-zooplankton systems. The conditions of pattern self-organization are derived, and importantly, the dispersion relation with two peaks is found to be the key for the occurrence of nested patterns. Based on the analysis of amplitude equations and simulations, four cases of nested patterns that are characterized by the nestedness of different sizes of hexagon and stripe patches are found. As the ratio of the wavenumbers corresponding to the critical Turing modes of patterns increases, the difference of patch sizes in the nested structures rises accordingly, leading to more small patches growing in the large ones. Moreover, the effect of advection on the self-organization of nested patterns is also discussed. The findings in this research may contribute to a new comprehension of the nestedness phenomenon occurring in the spatial distribution of plankton, as observed in field investigations.

MODELING LAND DEFORMATIONS IN MOUNTAINS BY COMBINING TIME-SERIES L-BAND SAR IMAGES AND SPATIOTEMPORAL STATISTICAL MODELS

Abstract. Predicting land deformations using remote-sensing technology is extremely challenging because the simulation models that are involved have many parameters, including geological features, that must be calibrated against areas of interest, and no universal model is available. Instead, this paper proposes a method for modeling land deformations with time-series L-band synthetic aperture radar (SAR) images and spatiotemporal statistical models for prediction. First, time-series SAR analysis generates temporal land deformation. Then, water-related products are obtained from the Today’ Earth system of the Japan Aerospace Exploration Agency. The three-day consecutive maximum values are calculated and then interpolated spatially. Finally, a descriptive or dynamic spatiotemporal statistical model is applied. The low-rank Gaussian process model, a descriptive model, is found to be highly accurate for explaining land deformations, even for actual slid areas, as long as the areas of analysis are limited, e.g., 150 m × 50 m. It is concluded that the proposed method is effective for predicting land deformations spatially and temporally and so can be used to identify possible landslide locations.

A Dynamic Spatiotemporal Stochastic Volatility Model with an Application to Environmental Risks

A dynamic spatiotemporal stochastic volatility (SV) model is introduced, incorporating explicit terms accounting for spatial, temporal, and spatiotemporal spillover effects. Alongside these features, the model encompasses time-invariant site-specific factors, allowing for differentiation in volatility levels across locations. The statistical properties of an outcome variable within this model framework are examined, revealing the induction of spatial dependence in the outcome variable. Additionally, a Bayesian estimation procedure employing the Markov Chain Monte Carlo (MCMC) approach, complemented by a suitable data transformation, is presented. Simulation experiments are conducted to assess the performance of the proposed Bayesian estimator. Subsequently, the model is applied in the domain of environmental risk modeling, addressing the scarcity of empirical studies in this field. The significance of climate variation studies is emphasized, illustrated by an analysis of local air quality in Northern Italy during 2021, which underscores pronounced spatial and temporal clusters and increased uncertainties/risks during the winter season compared to the summer season.

A quantitative model for spatio-temporal dynamics of root gravitropism.

Plant organs adapt their morphology according to environmental signals through growth-driven processes called tropisms. While much effort has been directed towards the development of mathematical models describing the tropic dynamics of aerial organs, these cannot provide a good description of roots due to intrinsic physiological differences. Here we present a mathematical model informed by gravitropic experiments on Arabidopsis thaliana roots, assuming a subapical growth profile and apical sensing. The model quantitatively recovers the full spatio-temporal dynamics observed in experiments. An analytical solution of the model enables us to evaluate the gravitropic and proprioceptive sensitivities of roots, while also allowing us to corroborate the requirement for proprioception in describing root dynamics. Lastly, we find that the dynamics are analogous to a damped harmonic oscillator, providing intuition regarding the source of the observed oscillatory behavior and the importance of proprioception for efficient gravitropic control. In all, the model provides not only a quantitative description of root tropic dynamics, but also a mathematical framework for the future investigation of roots in complex media.

Modelling of wind and photovoltaic power output considering dynamic spatio-temporal correlation

The simulation technology of wind and solar power output can provide data support for the planning of new energy stations and the optimization and scheduling of power systems. In order to solve the problem that the existing output models can not accurately describe the dynamic spatio-temporal dependence between wind and solar output, a dynamic spatiotemporal correlation model of wind and solar output based on Markov stochastic process and Copula function theory is proposed. Firstly, wind power and photovoltaic output are regarded as a stochastic process, and the time autocorrelation models of wind power and photovoltaic output are constructed based on a one-dimensional Markov chain and hybrid Copula function. Next, two one-dimensional Markov chains were coupled to form a two-dimensional Markov chain, and a spatial correlation model between wind and solar output was constructed using the dynamic SJC Copula function. Take the measured data of adjacent wind farms and photovoltaic power stations in Hami, Xinjiang as an example for simulation. The simulation results show that the proposed model can effectively reflect the spatio-temporal correlation of the original data and reflect the dynamic changes in the correlation between wind and solar energy.

Model of shrimp pond-mediated spatiotemporal dynamic distribution of antibiotic resistance genes in the mangrove habitat of a subtropical gulf

Aquacultures are the main reason for the environmental selection of antibiotic resistance genes (ARGs), resulting in the enrichment of ARGs. As a filter, a marine mangrove ecosystem can reduce antimicrobial resistance (AMR) or eliminate ARGs; however, its elimination mechanism remains unclear. This study investigated the spatiotemporal dynamic distribution of ARGs in two different types of mangrove habitats (shrimp ponds and virgin forests), within a subtropical gulf located in the Beibu Gulf, China, during dry and wet seasons by using metagenomics and real time quantitative polymerase chain reaction (RT-qPCR) analysis. As the key environmental factors, sulfide, salinity, and mobile genetic elements significantly were found to contribute to ARGs distribution, respectively. Wet and dry seasons influenced the dispersal of ARGs but did not affect the microbial community structure. Three potential biomarkers, TEM-116, smeD, and smeE, played key roles in seasonal differences. The key different genes in the biological relevance of absolute abundance were demonstrated by RT-qPCR. Co-occurrence network analysis indicated that high-abundance ARGs were distributed in a modular manner. For the first time, a risk index weighted by risk rank (RIR) was proposed and used to quantify the human risk of ARGs in the mangrove metagenome. The shrimp ponds during the wet season showed the highest RIR detected. In addition to offering a perspective on reducing AMR in mangrove wetlands, this study constructed the first spatiotemporal dynamic model of ARGs in the Beibu Gulf, China and contributed to revealing the global spread of ARGs. Meanwhile, this study proposes a new pipeline for assessing the risk of ARGs, while also exploring the concept of “One Health.”