A5043520991- Evacuation and Crowd Dynamics
- Simulation and Modeling Applications
- Robotic Path Planning Algorithms
- Metaheuristic Optimization Algorithms Research
- Mobile Agent-Based Network Management
- Opportunistic and Delay-Tolerant Networks
- Vehicular Ad Hoc Networks (VANETs)
- Fire Detection and Safety Systems
- Artificial Immune Systems Applications
- Transportation Planning and Optimization
- Image Retrieval and Classification Techniques
- Smart Parking Systems Research
- Scheduling and Optimization Algorithms
- Underground infrastructure and sustainability
- Advanced Multi-Objective Optimization Algorithms
- Software-Defined Networks and 5G
- Caching and Content Delivery
- Plant Parasitism and Resistance
- Innovative Educational Techniques
- Topic Modeling
- Energy Efficiency in Computing
- Advanced Text Analysis Techniques
- Facility Location and Emergency Management
- Dental Radiography and Imaging
- Peer-to-Peer Network Technologies
Shanghai University of Engineering Science
2019-2024
Tongji University
2016-2019
The Phasmatodea population evolution algorithm (PPE) is a novel metaheuristic proposed in recent years, which simulates the evolutionary trend of stick insect population. In this article, multigroup-based with mutistrategy (MPPE) to further improve overall performance PPE. During initialization period, divided into multiple groups, and step factor flower pollen introduced growth model no more than half groups. This makes diversified prevents from falling local optimal solution certain...
Multicast is an efficient tool to transfer the same data from a publisher multiple subscribers, and its performance relies critically on resiliency. Compared with traditional networks, centralized control flexible programmability of Software Defined Network (SDN) more beneficial for solving failure recovery problem. This paper presents design OpenFlow controller which can recover reconstruct multicast tree rapidly, while reducing Ternary Content Addressable Memory (TCAM) consumption...
Mobile-WebVR based real-time fire evacuation suffers from the limitation of cache and computing Web-browser. This paper proposes a lightweight intelligent solution for popular evacuation. Firstly, lightweighting scenario objects; secondly, collision detection avatars; thirdly, optimal path planning At last, system is created with flattering (Lightweighting Fire Evacuation system, LFE system). The uses real data (the static metro station BIM data, dynamic smoke data). object are reduced to...
The textual similarity task, which measures the between two text pieces, has recently received much attention in natural language processing (NLP) domain. However, due to vagueness and diversity of expression, only considering semantic or syntactic features, respectively, may cause loss critical knowledge. This paper proposes a new type structure tree for sentence representation, exploits both (structural) information known as weight vector dependency (WVD-tree). WVD-tree comprises trees...
Virtual training of metro fire evacuation definitely becomes considerably more efficient and convenient immigration trainees from the PC desktop to mobile web browsers.However, browsers face extreme difficulty in sustaining online visualization large-scale building information modeling (BIM) data on metro, FDS smoking data, crowding passengers, planning path simultaneously due weak rendering capability limited networking bandwidth.Herein, a novel lightweight Web3D technology roadmap is...
Mountain scenario data is large and complex for there are amount of semantic information in it. That the reason why it difficult computer to calculate an accurate path real-time multi-agent marching mountain scenario. The paper proposed m2ACO (multi-agent environment using ACO) algorithm planning Web3D technology. And pgACO (planar grid A* have been implemented this compare with algorithm. experimental results show that more than other kinds algorithms.
The path planning in Web3D for large-scale scenes is a fundamental problem. Especially, there are many agents searching their optimal paths complex and mountain scene. This paper addresses kind of method using an improved ant colony algorithm to find multi-agent environment HTML5. An leader-follower used organize the Blue army soldiers attach Red soldiers. Firstly, proposes BIM information together with ACO form method. Secondly, reality experiment by Egret Engine when marching or fighting...
In mountain battlefield, the path planning of army needs many detail scene information which based on semantic (SI) terrain. This paper basing SI addresses a kind mACO(mountain Ant Colony Optimal) method to compute an optimal in terrain for Web3D. A series solutions proposed inhere: abstracting including ray stone, round wood, entrenchment, bamboo nails trap and fire power information, then mapping every weapon's destruction value into use pheromones, will be reused process planning. And at...
The phasmatodea population evolution algorithm (PPE) is a recently proposed meta-heuristic based on the evolutionary characteristics of stick insect population. simulates features convergent evolution, competition, and growth in process nature realizes above through competition model. Since has slow convergence speed falls easily into local optimality, this paper, it mixed with equilibrium optimization to make easier avoid optimum. Based hybrid algorithm, grouped processed parallel...
目的 随着虚拟现实技术的发展，在虚拟场景中，基于多智能体的逃生路径规划已成为关键技术之一。与传统的火灾演习相比，采用基于虚拟现实的方法完成火灾逃生演练具有诸多优势，如成本低、代价小、可靠性高等，但仍有一定的局限性，为此，提出一种改进的双层深度Q网络（deep Q network，DQN）架构的路径规划算法。方法 基于两个结构相同的双Q网络，优化了经验池的生成方法和探索策略，并在奖励中增加火灾这样的环境因素对智能体的影响。同时，为了提高疏散的安全性和效率，提出了一种基于改进的K-medoids算法的多智能体分组策略方法。结果 相关实验表明提出的改进的双层深度Q网络架构收敛速度更快，学习更加稳定，模型性能得到有效提升。综合考虑火灾场景下智能体的疏散效率和疏散安全性，使用指标平均健康疏散值（average health evacuationvalue，AHEP）评估疏散效果，相较于传统的路径规划方法A-STAR （a star search algorithm）和DIJKSTRA （Dijkstra’...