Research On The Establishment And Situational Awareness Of Safety Zone In The Complex Flow Field Of Emergency Events

Safety of the indoor environment is an important task related to personnel health and normal activities.In recent years,arson and chemical attack incidents occurred inside the building cause concern about terrorism,which makes building safety has caused widespread attention.As the result of the space limitation and personnel concentration,once the poisonous area cannot be cleaned in time,irreparable casualties will happen.At present,researches on indoor safety issues are limited.The primary technical challenge currently is indoor flow field,affected by the ventilation airflow,presents a chaotic turbulent motion state.Harmful substances are affected by turbulence and rapidly diffuse in indoor space,and it is difficult for us to predict its development because of the high coupling and dimensionality of physical quantity parameters.On the other hand,we need to consider the construction of the safety zone/ path that facilitates evacuation and escape in the turbulent environment.Therefore,how to quickly build a safety zone in indoor space under biochemical attack and predict the evolution of the harmful substance concentration field online have become important issues in indoor safety field.At present,domestic and foreign countries have not yet solved the problem of online evolution of safe area under emergency scenarios.The research on rapid construction and situation awareness of indoor safety zone is still important.In view of the above two main problems,the research work as follows:(1)We propose the concept of the safety zone and through flow-field control couple with the decontamination agent technology to minimize the harm and construct the safety zone based on computational fluid dynamics(CFD)software – STAR-CCM +.Aiming at three typical environmental modes(natural diffusion,outdoor ventilation and air conditioning ventilation),CFD simulations provide great accuracy and more details about distribution of concentration,as well as velocity in spatiotemporal scales.Corresponding experiments are provided for validations.Simulation results show that the method can effectively control the transportation of toxic substances and establish the safe area.The results also reveal the presence of coherent correlation between the airflow state and the evolution of safety zone,which is of great value for the safe design of ventilation system.(2)Based on deep learning technology,a Multi-Step Spatial-Temporal Situational Awareness Network(MSSTP-SA net)is proposed.The model consists of a spatial feature extraction network and a temporal feature extraction natwork.It can perform online identification and prediction of the safety zone evolution in indoor space when provide real-time data of monitoring points arranged in indoor inactive areas.The dataset covers a great number of biochemical attack scenarios through orthogonal multi-boundary conditions.The data are of high-precision and reliable characters obtained through CFD numerical simulation.The test experiments suggest that the evolution of concentration field distribution could be predicted faithfully in millisecond computation time costs,which ensure the real-time prediction could be provided in emergency scenarios.Therefore,MSSTP-SA net has application potential in indoor emergency chemical attack events.In summary,this article proposes the technology to construct the security zone and a situation awareness model.We also discuss the algorithm deployment for application practice.We conduct a relatively systematic study which is of certain theoretical innovation and application value.