In order to further improve the basic role of emergency shelters in comprehensive disaster prevention and mitigation，a multi-dimensional evaluation index system for emergency shelters was constructed based on the resilience perspective in five aspects: safety of the place，standardization of management，accessibility to transportation，guarantee of service and reconstruction and recovery. A comprehensive evaluation model was established by combining the Analytic Hierarchy Process (AHP) and Geographic Information System (GIS) spatial analysis. Taking the Ouhai District of Wenzhou City as an example，based on the first comprehensive natural disaster risk survey data and open source data，the disaster prevention and mitigation ability of emergency shelters in the district was comprehensively evaluated，and the scientific and effectiveness of the model was verified. The study shows that the constructed index system and comprehensive evaluation model can be used to analyze and evaluate the disaster prevention and mitigation ability of emergency shelters，the weak indicators of emergency shelters can be found，and the direction of optimization and improvement can be provided for managers. In the case analysis，the low resilience emergency shelters in Ouhai District are mainly concentrated in the western mountainous areas，and the resilience of village-level emergency shelters is relatively low，which is consistent with the actual situation，proving.

In order to clarify the relationship between the exterior insulation facade structure and fire spread in high-rise buildings，Pyrosim fire simulation software was used to study the impact of different exterior insulation facade structures on fire spread in high-rise buildings. The results show that during the process of fire spread on the exterior facade of high-rise buildings，the insulation systems of different exterior facade structures reduce the air entrainment capacity and heat release rate as the degree of structural space limitation increases. However，their chimney effect is significantly enhanced，and the smoke flow rate is faster，leading to an accelerated vertical fire spread speed. As the thickness of the air layer increases，the temperature and smoke flow rate of the aluminum curtain wall structure insulation system first increase and then show significant fluctuations. The aging of the performance of the external wall insulation system will increase the risk of fire on the exterior facade of high-rise buildings.

In order to improve the situation of network security education and the quality of network security education in colleges and universities，a cognitive-attitude-behavior theoretical model of college students was constructed. Through the distribution and screening of questionnaires，659 valid questionnaires were obtained. SEM was used to test the influence of network security education on network security cognition，network security attitude and network security behavior in the collected data. The results show that network security education can have a significant influence on network security cognition，network security attitude and network security behavior. During the influenced process，network security cognition and network security attitude play an independent and chain intermediary role. Legal and moral education，knowledge education，practical activities，and leading by example can directly affect network security behavior. Legal and moral education has the strongest influence，followed by knowledge education. Practical activities and leading by example are the weakest. In terms of specific paths of process，legal and moral education mainly affects network security behavior through direct effect，while knowledge education，practical activities and leading by example mainly affect network security behavior indirectly through security cognition.

In order to quickly and effectively extinguish the lithium-ion battery fire，solve the problems concerned in the fire protection field，such as long fire extinguishing time and high water consumption，and explore the inhibition effect of hydrogel extinguishing agent on the thermal runaway of large capacity lithium-ion batteries. Firstly，the microstructure of hydrogel at high temperatures was analyzed by an environmental scanning electron microscope. Then，by building a lithium battery combustion test platform，the hydrogel fire extinguishing test was carried out. Taking the 135 Ah square aluminium case ternary lithium-ion battery pack for vehicles as the test object，the electric heating method was used to induce its thermal runaway and explore the cooling inhibition effect of hydrogel on lithium batteries. The results show that the pore structure of the hydrogel is destroyed after being heated，which is conducive to its adhesion to the surface of the object and continuous cooling. When using hydrogel for fire extinguishing and cooling，the maximum cooling rate of the battery surface is twice that of water. After the hydrogel is sprayed，the temperature of the lithium battery rises slowly，and the heating rate is only half that of water. Compared with water，the hydrogel can delay the thermal runaway of adjacent batteries for a longer time，which can bring longer safety time for rescue and escape.

In order to improve the safety management effect of subway construction and reduce the occurrence of subway collapse accidents，the risk factors of subway construction accidents were analyzed. Firstly，the multi-factor coupling mode of 'human' and 'object' in accident causation were analyzed systematically. On this basis，with the help of accident causation 24Model and GSM，the management model of safety system was constructed. Finally，the multi-factor coupling failure mode of human-object system was applied to 16 existing subway construction collapse accidents. The multi-factor coupling mode and action path of 14 "human" reasons and 6 "object" reasons were given respectively，and their risk levels were calculated. The results show that there are 6 modes and 14 action paths for the multi-factors coupling effect of "human" and "object" in subway construction. The management model is divided into the target layer and the matching layer between the cause of the accident and the safety defense line. The matching layer is constructed from three levels: micro matching，medium matching and macro matching. Based on the established safety system management model of subway construction，the "three defense lines" of subway collapse accident safety construction management were given.

In order to effectively reduce the accident rate of mountain highways，the traffic accident data of mountain highways in Yunnan province from 2016 to 2021 was taken as the research object，based on the DEMATEL-AISM. This paper analyzes the causality of risk factors and draws the UP and DOWN directed topological hierarchical diagrams，and finally determines 19 risk factors，constructs an N-K-coupling degree model to quantify the risk factors，couples the risk factors of mountain highway traffic accidents in all dimensions，explores the relationship between risk factors，and proposes a full-dimensional coupling model of traffic accidents in mountainous areas. The results show that in the single dimension，the coupling value of human factors being too close to the vehicle and fatigue driving is 0.741，and the coupling value of road factors is 0.816，which are the two effects that have a greater impact on the system in the single dimension，and the coupling values of human-vehicle and human-road are 0.157 and 0.124 in the two-dimensional. The maximum effect of human factors is human-road-ring in multi-dimensional，with a coupling value of 0.891，in which the driver's bad driving behavior，the sharp bend of the road and the long downhill，and the rain，fog，and ice and snow days of the environment are easy to be coupled with other factors more than 70%，which constitutes a strong coupling relationship and the probability of traffic accidents is large.

To promote the construction of emergency technology and management majors and strengthen the university talent cultivation of emergency rescue, the majors' characteristics were determined based on the national needs in emergency. Then, a talent development mode for emergency technology and management majors was proposed from aspects of professional ethics, theoretical knowledge systems, and practical technical ability. Firstly, the majors' professional ethics were clarified based on the professional ethics requirements from Teaching Standards for Emergency Rescue Technology and National Vocational Skill Standards for Emergency Rescuers. Secondly, a core curriculum system covering the entire emergency lifecycle was proposed using the PPRR(Prevention, Preparation, Response, Recovery) model including prevention, preparation, response, and recovery. Next, a combination of learning and training, integration of government-industries-research organizations was proposed to advance the cultivation of emergency skills and the enhancement of practical innovation capabilities. Finally, a talent cultivation system was developed from the perspectives of textbooks, faculties, teaching process management, a talent assessment system, and a continuous improvement mechanism. The results indicated that the talent development system integrated with public service and social responsibility attributes, full-cycle work content, disasters, and multi-subjects practical applications. Moreover, the system developed a close connection between talent cultivation and practical work needs and guided for cultivating of high-level applied talents in the emergency field who had moral and professional qualifications and theory and practice capabilities.

To reduce the influences of background interference factors in natural environments such as clouds，mist，dust，lights，sunrise，and sunset on the smoke and flame target detection accuracy，a smoke and fire detection algorithm based on an improved YOLO-V5 algorithm was proposed. Smoke，flame target images，and interference image data sets were obtained from the on-site collection and web crawling approaches to solve sample imbalance and improve model generalization ability. A bidirectional feature pyramid network (BiFPN) was used to replace the original feature pyramid network (FPN) + path aggregation network (PAN) structure，and then multi-scale feature fusion on the target was performed to strengthen the model feature fusion ability. At the same time，distance intersection-over-union(DIoU) non-maximum suppression(NMS) is used to replace the original non-maximum suppression (NMS) to speed up the convergence of the detection box loss function and enhance the model reasoning ability. The results showed that the improved algorithm's accuracy，recall rate，mean average precision(mAP) and FPR were 79.2%，68.6%，74.2%，and 12.8%，respectively. Compared with the original YOLO-V5 algorithm，the proposed algorithm improved accuracy rate，recall rate，and mAP by 1.9%，0.9%，and 2.7%，respectively. Furthermore，the FPR was decreased by 3.7%.

In order to improve traffic safety in the coupling area of the township fair and through-village highway，a secondary road through villages and towns in Yunnan Province was taken as the research object. The core area and the radiation area of the fair were defined according to the affordable travel distance of the township residents and the maximum trading radius of the fair. On this basis，29 risk factors in 6 dimensions of people，vehicles，roads，environment，management and accident form were selected. The risk factor identification model based on association rules was used to construct the risk factor network of highway accidents through villages and towns. By analyzing the characteristics of network topology，accident risk factors were identified. The results show that the traffic conflict in the core area of the fair is more serious，and the probability of injury accidents is 57.99 %. The core area of the township fair is the primary risk factor for accidents and plays a leading role in the spread of accident risks. The types of direct risk factors for property loss accidents and injury accidents are heterogeneous. Overall，the average clustering coefficient of driver factors is the largest，which makes it easy to produce a chain reaction. In the core area of the fair，the probability of injury accidents increases by 23.71% and 21.07%，respectively，when the driver's meeting vehicles are improper，and the driver does not give way according to the regulations.

To solve the problem of severe erosion caused by sand particles passing through separators during shale gas extraction，a well in Changning Shale Gas Field was taken as an example. Based on FLUENT software，numerical simulation was used to simulate the erosion characteristics of sand particles on the drain valve under different speeds，sand content，particle size，shape coefficient and valve opening. The main factors affecting the erosion characteristics of the drain valve were evaluated to clarify the erosion situation of the drain valve. The results indicate that the erosion damage of the valve disc and throttle hole is the main reason for the failure of the sewage valve. As the valve opening decreases，the pressure difference inside the valve increases exponentially，and the flow velocity is the largest at the valve disc and throttle hole. With the decrease of shape coefficient of sand particles，the erosion of the valve is more severe. Based on sensitivity analysis，the degree of influence of various factors on valve erosion characteristics is: ξ_l (Speed)=0.73，ξ_m (Sand particle size)=0.71，ξ_n (Sand content)=0.70 and ξ_q (Shape coefficient)=0.67. Therefore，it is recommended to control the fluid flow rate to within 7 m/s by increasing the internal flow area of the valve，improving the sand removal ability of the desander，and preventing larger sand particles (sand particle size>60 μm) from entering the sewage valve. The sand particle size entering the sewage valve should be controlled within 60 μm，which provides an effective means for the evaluation and optimization of on-site equipment.