Department of Electrical and Electronics Engineering Korea Maritime and Ocean University Graduate School. An emergency evacuation system is a system that helps people in a room to evacuate safely and quickly from emergencies in the event of danger. For example, managing an evacuation route by deploying the right human resources, or an evacuation route such as an emergency exit point direction that points in one direction only at one location.
Relying on human subjective judgment in a dangerous situation can be very dangerous, and emergency lights and evacuation routes that always point in the same direction cannot flexibly deal with risk factors and can expose the public to danger. In addition, due to the nature of the ship's construction, the initial response is important because it is the rescue time that is delayed rather than the accident on land. Therefore, emergency evacuation systems should be more intelligent in increasingly complex and larger structures and should be able to quickly recognize information about the surrounding situation and suggest an optimal evacuation route.
In particular, the possibility of dangerous elements spreading or becoming dangerous areas along the route where evacuees go cannot be ruled out. Therefore, there is a need for a system that predicts and responds to the near future with sufficient modeling of risk factors. Among the various risk factors, risk factors such as fire, smoke and insulation can be sufficiently collected using sensors or image processing devices.
However, in the case of bottlenecks, it is essential to model the population density at the current node, the direction in which people at that location will evacuate, and whether the path of the chosen route will accommodate the incoming population.
Introduction
- Research background
- Research Trends
- Research Necessity
- Research Summary
- Searching Algorithm
- State Space and Search
- Blind Search
- Heuristic Search
- Searching System
- Feeling Factor
- Risk Predicted Value
- Evacuation System for accident situation
Therefore, in the field of disaster prevention and disaster prevention, such as safety engineering and architectural engineering for unpredictable risk factors, the impact and characteristics of risk factors under disaster conditions have been analyzed and modeled. These studies have analyzed the impacts and characteristics of different risk factors in accidents, which can lead to system requirements for responding to risk factors in a ship accident environment. Therefore, it is necessary to verify the effectiveness of the proposed method based on the performance evaluation based on the actual disaster.
Therefore, a suitable route search algorithm based on technical stability is needed for the specificity of the ship disaster situation. The state here refers to the state of the problem world at a given moment in time, and the world collectively refers to the objects involved in the problem and their situation. However, backward depth-first search and breadth-first search tree structure design features are very useful in designing the proposed state space.
When node 21 is the initial node and is. node 25 is the target node, the principle of operation of the algorithm is as follows. Algorithm operation in step 4. However, the reason why node 5 is chosen instead of node 12 in node 13 is that the search order is determined by the priority of the evaluation function of the algorithm. Depending on the factors that can interfere with the journey in such an environment, the distance that a person actually experiences will be different, even if the distance traveled will be different based on the actual physical distance and risk factors.
This is expressed as the product of the actual distance value () and the feel factor (α) as follows. The weight () is set according to the level of risk or the level of difficulty of the risk, and each risk factor is marked with an index , so that we can define fire as , submergence as and slope as. In the case of a path with risk factors, there is a time difference between the optimal path decision point and the opponent's movement state when the path is represented only by the feeling factor, and in the time interval.
Therefore, the evacuation route decision using only the situation at the time of evacuation cannot be considered the optimal route choice without considering the location, speed and potential risk factors of the evacuee. A device is provided that can reflect the risk level that can be changed dynamically separately from the static weight ( ) of the risk factors. Considering all the risk factors including the risk prediction coefficient (), i.e. the expected risk factors, the estimated weight distance () for the r-th path is expressed as follows.
The optimal route guidance system considering the disaster situation is to detect the presence of risk factors () and the risk level () through the sensor installed in the target structure and calculate the optimal distance of the escape route search. The search procedure for the optimal path described above can be summarized as shown in fig.
Proposed Scheme
Graph Search for inside of ship
Modeling of bottleneck
Using the accommodation rate in the interval, it is possible to obtain the variation of the weight distance through the bottleneck by dividing the number of persons distributed in the corresponding section in each section from the section to the section accommodating rate in the next section. It is assumed that the people distributed in each section when the accident occurs move towards the exit according to the average movement speed. The graph shows the value of for an arbitrary interval according to the proposed method.
As a result, it can be confirmed that blocking occurs when the value of is greater than that of blocking and there is a delay of about 1.7 seconds depending on the input rate compared to the initial number of people until the bottleneck is resolved. tight . Experimental data may be different depending on the number of human access sections and the rate of access to each access section.
Proposed Route Optimization Algorithm
In the event of a disaster, the sensor node of each master node retrieves the sensor data value for the risk element, calculates and stores the risk factors and the risk level.
Simulation and Analysis
Bottleneck occurrence probability
- Experiment environment and result
13 is the graph when only the shortest distance is searched based on the optimal road system, and Conv_2 is the graph when the existing system is applied. It is confirmed that the probability of occurrence of bottlenecks is reduced when the proposed algorithm is applied to the existing emergency routing algorithm. However, it can be seen that the overall stability is improved by reducing the additional delay of a particular node due to the bottleneck.
Weighted distance according to proposed scheme
However, the cost increases by taking the bypass route to avoid congestion, which indicates an increase in the total evacuation time. 14, we have simulated the weight distance increase due to the obstacle at the current node. The parameters applied in the simulation are the same as those used in the simulation.
15 is a graph showing the sensed distance value by the proposed method at the current node. In the case of Conv_2, it can be confirmed that the weight distance value is reduced by applying path optimization. However, in Prop's graph, it can be seen that by applying load balancing to the existing route optimization algorithm, the sense distance is reduced by about 20%.
Evacuation time according to proposed scheme
Conclusiond
Jeon, "A Study on the Optimal Route Planning Algorithm for Rescue of Multiple Victims in Disaster Area." Chang, “Modeling and Optimization of Building Emergency Evacuation Considering Blockage Effects on Crowd Movement,” IEEE Transactions on Automation Science and Engineering, Vol. Hwang, “A Study on Prediction and Improvement Method of Fire Risk for a Newly Built College Dormitory,” Journal of the Korean Society of Marine Engineering, Vol.
