This paper reviews the main progress in the integration of pedestrian dynamics and epidemiology for the risk management of highly infectious diseases. Pedestrian dynamics models can characterize individual movement, contact relationships, and spatial exposure processes, providing effective tools for assessing disease transmission risks and optimizing crowd management measures. This paper summarizes the research methods of pedestrian microsimulation models for simulating disease transmission processes, and analyzes the characteristics of disease transmission, including transmission pathways, transmission patterns, and risk assessment indicators. On this basis, the paper identifies existing limitations in parameter setting, data validation, and model application, providing a reference for transmission risk assessment and refined management of infectious diseases in crowded places.

This paper focuses on the global LNG trade network from 2013 to 2022, aiming to identify potential LNG trade relationships through an improved link prediction algorithm. It is important to assist LNG-consuming countries in finding potential partners and ensuring energy supply security. External factors, such as geographical distance between nodes, trade volumes and poli-tical stability of countries, are fully considered, and these are coupled with internal network indices through a multi-index approach. A random forest classifier is used to enhance prediction accuracy. The results show that the average prediction accuracy can reach 97.55%. And, 120 potential links had been successfully identified, of which 99 was transformed into real links in 2022, and 94 links were transformed into real links within following three years. These results not only confirm the high precision of the algorithm in forecasting unknown links but also highlight its effective capacity for identifying potential future links.

In the influence maximization problem, existing methods based on greed and the centrality of network structure are difficult to strike a balance between accuracy and time complexity. Aiming at the limitations of existing algorithms, an improved sine-cosine optimization search influence maximization algorithm DSCA is proposed. This method takes into account the two-hop neighbors of nodes, and combines the position update concept of the optimization algorithm with dynamic small-range search to optimize a single node. The sine-cosine search, alignment and mutation operators are used to search nodes based on ordering. Experiments on four networks show that the proposed method has wider propagation range and lower time complexity than the advanced benchmark method, which proves the superiority and accuracy of the proposed algorithm.

The paper addresses an overlooked phenomenon in opinion dynamics research: the clustering of identical opinions on network topology. The paper investigates the distribution and clustering of steady-state opinions on networks through the probability distribution function of the proportion of neighbors holding different opinion around a specific type of opinion. Simulations of the Deffuant-Weisbuch model′s opinion diffusion process reveal two key characteristics of steady-state opinion distributions in ER random graphs converging into two opinion clusters: the probability distribution function follows a binomial distribution, and identical opinion aggregation induces a probability bias, where the product of the bias and the network′s average degree remains constant. Building upon these findings, this study extends the results to small-world networks and examines how clustering coefficient influences the steady-state distribution and clustering of opinions, thereby further validating and broadening the general applicability of the research outcomes.

To enhance the science and effectiveness of dynamic temporal networks analysis, a systematic, data-driven approach to selecting time scales is proposed. Firstly, the multi-scale entropy theory is introduced to describe the complexity and randomness of the passenger flow time series to determine the appropriate time scale selection range. Secondly, the network characteristic indexes are selected from node characteristics, edge characteristics and overall structure, and the time scale is selected by combining principal component analysis and random forest algorithm. Three networks are used as examples for validation, and the results show that the method is applicable to the data requirements of different transportation systems with high rationality and applicability.

In order to more accurately predict the spread trend of infectious diseases and seek the best strategy to control the development of diseases. In this paper, a spatial diffusion SVEIQR epidemic model with a saturated incidence rate and temporary immunity is presented. We define the basic reproduction number R₀ and establish the threshold criteria for the global dynamics. By constructing a Lyapunov function, we ascertain that the disease-free equilibrium point is globally asymptotically stable when R₀<1, and when R₀>1, the endemic equilibrium point exhibits the same property. Finally, the correctness of the theoretical derivation was verified by numerical simulation and it was found that the diffusion effect will not fundamentally change the global threshold stability of the epidemic model. However, a large diffusion coefficient will accelerate the spread of the epidemic and thus accelerate the speed of the epidemic to reach a stable state.

Existing infectious disease models are usually based on compartment models and model optimisation is performed by adjusting the number of compartments and transfer paths. However, the finite states delineated by the compartment model often do not adequately reflect the actual states that individuals live in the real world. In this study, we model infectious disease based on quantum mechanics and use quantum superposition states to represent individual states, achieving a more accurate representation of individual states. Firstly, this paper analyses the individual infection process and evolution of the model, and derives the basic reproduction number and disease-free equilibrium point of the model. Secondly, the model simulation is carried out on a quantum circuit, and the parameter sensitivity of the model is analysed and the reasonableness of the model is verified. The simulation results show that the predictions of the model are consistent with the general law of virus propagation and can be implemented to simulate the structure of the compartment model. Finally, the applicability of the model is further verified by simulating real COVID-19 epidemic data.

This study analyzes the carbon footprint and implicit carbon transfer of advanced manufacturing in the western region, revealing the spatial distribution characteristics of carbon emissions and transfer pathways, providing a basis for formulating regional collaborative emission reduction policies. Using data from 11 provinces between 2007 and 2017, the study applies the IO model and MRIO model to measure carbon footprints and analyze the dynamic changes in inter-provincial and inter-industry implicit carbon transfers. It also constructs a provincial implicit carbon transfer network and examines its overall and individual node characteristics. The results show that: The carbon footprint of advanced manufacturing in the western region has increased overall, with significant growth in provinces such as Inner Mongolia, Guangxi, and Chongqing, mainly concentrated in the non-metallic mineral products and equipment manufacturing industries. The implicit carbon transfer into Sichuan, Guizhou, and Yunnan has increased, while Inner Mongolia and Guangxi experienced a significant growth in implicit carbon transfer out in 2017, mainly from the non-metallic mineral products industry. The implicit carbon transfer network exhibits a "core-periphery" structure, with Sichuan and Shaanxi in the core and Xinjiang, Qinghai, and Gansu in the periphery.

This study explores the influence of cross-platform user mobility on the dissemination of corporate negative information, focusing on the dissemination characteristics and patterns of multi-topic public opinion regarding corporate negative events. First, based on infectious disease theory, a cross-platform dissemination model for multi-topic public opinion on corporate negative events is constructed. The basic reproduction number is then derived using dynamic equations. Real-world case data are employed to estimate model parameters through equation fitting, followed by simulation analysis. The simulation results indicate that, the closer the cumulative number of corporate negative information posts approaches stability, the higher the degree of fit between the model curve and the real data; competition effects among topics lead to an inverse relationship between the basic reproduction number and the infection rate of a specific topic; user mobility from Weibo to Bilibili has a more significant impact on the spread of corporate negative topics.

In order to identify and analyze the multi-relational network of cross-industry convergence fields, we used the LDA topic model, and applied the topic intensity, text similarity, co-presence rate, co-presence network analysis and the complex network analysis method to identify the multi-relational network. The findings reveal the evolutionary characteristics of multi-relational network topology, maximum communities, and key nodes. The findings indicate that universities have consistently occupied key node positions within these networks and have maintained a dominant control over network resources. Enterprises have gradually taken on more significant node positions in these networks. Furthermore, the core-periphery structures and the phenomenon of enterprises clustering together have become increasingly evident. This research provides a new approach for uncovering the associations and growth patterns among heterogeneous-relation agents in complex networks. It offers empirical evidence for enterprises seeking to acquire technological resources and gain a competitive edge in the market, while also serving as a valuable reference for government departments aiming to optimize policies for industrial convergence and development.

This paper investigates the herdability of multi-agent systems with general linear and signed matrix-weighted networks under signed directed graphs. Firstly, we present the sufficient conditions for herdability of general linear multi-agent systems by utilizing controllability structure decomposition. Moreover, leveraging the concepts of controllable subspaces and graph theory, we propose a set of graph-theoretic conditions for the herdability of multi-agent systems under different matrix weight scenarios. Finally, we introduce the definition of signed matrix-weighted distance-equivalent partitioning and provide the herdability criteria for general matrix-weighted networks under this partitioning. It is found that the system is herdable when the non-zero elements of the weight matrix are either all positive or all negative.

To address the issue of limited flexibility in homogeneous agents within persistent monitoring systems and to investigate the impact of sensor models on system performance, this study proposes an optimal persistent monitoring scheme based on heterogeneous agents. First, an optimal control framework is established, and the necessary conditions for the optimal control strategy are derived using the minimum principle, transforming the original problem into a parameter optimization problem. Subsequently, the infinitesimal perturbation analysis method is applied to calculate the gradient of the objective function with respect to the parameters, and the optimal control parameters are obtained through the gradient descent method. The results demonstrate that heterogeneous agents exhibit superior fault tolerance and flexibility compared to homogeneous agents. Although different sensor models influence the objective function, they do not alter the fundamental characteristics of the optimal trajectory.

In order to analyze the operating rules of the financial system, the dynamic behavior and synchronization control of a financial chaotic system are studied. By analyzing the dissipativity and the variation of state variables with parameters of the system, the existence of attractor has been proven, and when the system parameters change, the state variables traverses the stable, period doubling bifurcations, chaos, and other states. With the further analysis, the existence of the global exponential attraction set of the system is conducted. Based on the theories of the global exponential attraction set and the Dini derivative, a nonlinear controller is set up in the system to achieve global exponential synchronization control of the driving and response systems of the financial chaotic system. The maximum estimated value of system variables and the range of control parameter values are obtained, which can achieve synchronization between two systems. The correctness of the theoretical analysis is verified through numerical simulation analysis.

In order to analyze the effect of speed on shipborne tethered unmanned aerial vehicle (UAV) system, a nonlinear dynamic model of the system is established, and the cable tension and UAV movement in different speed regions are numerically analyzed. In region I, the UAV phase trajectory is closed, showing single-period motion characteristics, and the end tension under the cable produces frequency doubling response; In region II, there are intersection points in the UAV phase trajectory, showing period-doubling motion characteristics; In region III, the UAV phase trajectory is not closed, the motion trajectory is not coincident, showing chaotic motion characteristics, and the end tension spectrum under the cable is continuous spectrum. The research results of this paper can provide theoretical reference for the safe flight of tethered UAVs following ships.

In order to solve the problems of slow planning speed and large memory consumption of Path Planning Algorithm (RRT) for fast search random tree in the process of unmanned vehicle trajectory planning, an improved RRT algorithm combined with Artificial Potential Field Method (APF) is proposed. In this paper, a probability value is introduced into the random tree of the basic RRT algorithm to accelerate the convergence of the random tree to the target node, and then a gravity component is added to guide the random tree to grow to the target point to speed up the search process. At the same time, a repulsion field is established around the obstacle to limit the search area and reduce the randomness of the path. In this paper, we not only improve the RRT algorithm (APF-RRT), but also apply the above improvements to the RRT^* algorithm (APF-RRT^*). After a large number of simulation experiments, it is shown that the improved algorithm in this paper has a great improvement compared with the original algorithm and other latest improved RRT algorithms in terms of path length, optimization time and number of iterations.

In order to alleviate the problems of inefficient landing schedules, elevated operation costs of aircraft, and inadequate attention to passenger satisfaction, considering the expectations of multiple stakeholders, including air traffic control, airports, airlines and passengers, the concept of passenger expected time satisfaction is proposed. A model of arrival aircraft sequencing grounded in multi-stakeholder requirements is established, which aims at minimizing fuel consumption and delay time while maximizing the passenger expectation time satisfaction, and vortex wake separation standards, max position shift and point merge system airspace structure as constraints. The elite-preserved non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) is selected to solve it. Based on the data of actual airspace environment and the historical data of actual aircraft, the airspace structure of point merge system is designed as the simulation environment, the simulation experiment is conducted across three different types of air traffic flow. The results show that compared with the First-Come, First-Served (FCFS) model, under the three types of air traffic flow, the passenger satisfaction metrics are increased by 48.47% on average, the delay time is reduced by 59.37% on average, and the operating costs are reduced by 0.94% on average, verifying the effectiveness of the proposed model.

Vehicle driving behavior is constrained by multi-dimensional elements of complex traffic scenes, and its lane-changing decision is the key in the field of automatic driving. In order to solve the defect of traditional lane-changing model ignoring inter-vehicle interactions, the relationship between inter-vehicle interactions and inter-molecular forces is analyzed, and based on the molecular interaction Buckingham potential function, a vehicle lane-changing decision-making model is established. Relying on SUMO software, the model is compared and analyzed with SL2015 lane changing model in terms of safety, efficiency and comfort. The results show that the interaction potential decision model reduces the number of dangerous lane changes by 45%; the average speed increases by 3.38% and the average number of passes improves by 5.12%. The interaction potential model systematically describes the inter-vehicle interaction relationship, and provides a theoretical basis for internet-connected autonomous vehicles to make safe, efficient and comfortable lane-changing decisions.

An adaptive fixed-time sliding mode control method based on neural network is proposed for uncertain manipulator systems with actuator input dead zone. Based on the adaptive neural network and the minimum parameter learning method, the adaptive neural network sliding mode controller of the closed loop system of the manipulator is designed to effectively compensate the input dead zone of the manipulator, and only one parameter is updated online. Based on the fixed-time stability theory, it is proved that the controller can not only achieve the reachability of the sliding mode surface, but also ensure that the position tracking error signal of the closed-loop system converges to any small neighborhood with zero as the center in a fixed time, and then exponentially converges to zero. Finally, the simulation results verify the effectiveness of the proposed control method, and improve the transient and steady-state performance of the manipulator's tracking control to a certain extent.