Personnel evacuation modeling method and system for ship poison gas leakage scene
By building personnel intelligent model and coupled toxic gas diffusion model, and optimizing evacuation paths with NSGA-II algorithm, the problem of unsafe evacuation paths in ship toxic gas leakage scenarios is solved, real-time and accurate evacuation guidance and safety management are achieved.
Patent Information
- Application Number
- CN202510822615.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-19
- Publication Date
- 2025-08-26
AI Technical Summary
The existing technology fails to effectively combine personnel evacuation and toxic gas diffusion in the case of ship toxic gas leakage, ignoring personnel characteristics and environmental factors, resulting in unsafe evacuation paths and making it difficult to guide personnel evacuation accurately in real time.
Build a personnel intelligent model, set basic physical properties, consider gas correction and crowding correction, introduce the personnel disturbance quantization module to couple with the Gaussian smoke cluster model, use the NSGA-II algorithm to plan the evacuation path, and optimize the evacuation strategy with real-time data feedback.
It realizes dynamic coupling and closed-loop feedback between personnel evacuation and toxic gas diffusion, improves the accuracy and safety of evacuation paths, and improves emergency response capabilities and safety management levels.
Smart Images
Figure CN120542271A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of ship engineering and safety technology, and in particular to a personnel evacuation modeling method and system for ship toxic gas leakage scenarios. Background Art
[0002] As an important means of transportation, ships may face the risk of toxic gas leaks when transporting various types of cargo. Existing toxic gas diffusion models, such as the Gaussian puff model, can describe toxic gas diffusion trends but fail to consider the impact of evacuation on gas diffusion. Furthermore, existing evacuation models are not effectively coupled with toxic gas diffusion models and fail to reflect the interaction between the two. Furthermore, traditional path planning often relies on the principle of shortest distance, ignoring complex factors such as toxic gas concentration and occupant density, resulting in inadequate evacuation route safety. In emergency evacuation scenarios, the varying mobility capabilities of different populations are also not fully considered, making existing models unable to meet diverse needs. Due to the complex internal structures, dense populations, and narrow passageways of ships, existing technologies struggle to accurately and effectively guide evacuation in real time in the event of a toxic gas leak. These issues demonstrate significant shortcomings in existing evacuation modeling for ship-based toxic gas leaks. There is an urgent need to develop an efficient evacuation modeling method that comprehensively considers toxic gas diffusion, occupant characteristics, and environmental factors to enhance ship emergency response capabilities and safety management in the event of a toxic gas leak. Summary of the Invention
[0003] In view of the deficiencies in the prior art, the present invention provides a personnel evacuation modeling method and system for ship toxic gas leakage scenarios, which solves the problems raised in the above background technology.
[0004] To achieve the above objectives, the present invention is implemented through the following technical solutions: A personnel evacuation modeling method for ship toxic gas leakage scenario, comprising the following steps: S1. Build a ship crew agent model, treating each crew member as an independent agent and setting their basic physical attributes, including age and mobility. Differentiate the mobility of crew members of different age groups based on ergonomic data. S2. Analyze the impact of toxic gas diffusion on personnel movement speed and determine the toxic gas correction factor. That is, when the toxic gas concentration reaches the critical value, the personnel movement speed decreases by 10% for each increase in concentration. S3. Considering the impact of congestion on movement speed during evacuation, a congestion correction coefficient is set. When the number of agents in a unit grid reaches 3 or more, the speed is halved. When the number reaches 5 or more, the speed is further reduced. S4. Based on the Gaussian puff model, a personnel disturbance quantification module is introduced. By analyzing the relationship between personnel movement speed and local flow field, the local wind speed is corrected using the turbulence intensity of personnel movement. The diffusion coefficient is optimized according to the cabin structure and personnel distribution, realizing the dynamic coupling and closed-loop feedback of personnel evacuation and toxic gas diffusion. S5. The NSGA-II algorithm is introduced to solve the objective function that comprehensively considers multiple factors and perform personnel path planning. The planning factors include toxic gas concentration, personnel density, and path distance. According to the toxic gas leakage situation at different stages, the shortest path algorithm, low concentration priority strategy, and strategy of following the crew or ordinary personnel path are used to make evacuation path decisions.
[0005] Preferably, in the step S1, the reaction time of the human agent is randomly distributed with an average of 3 seconds. The human agent remains still during the reaction time and enters the evacuation state after the reaction period ends.
[0006] Preferably, in the step S4, when the local wind speed disturbance is superimposed, inside the closed and windless ship, the background wind speed is zero. Assuming that the movement of personnel generates an additional wind speed field, the direction is consistent with the direction of movement. Assuming that the movement of personnel generates an additional wind speed field, the direction is consistent with the direction of movement, and the moving speed of the i-th person is , the disturbance range radius is , then the disturbance wind speed caused by it is shown in the following formula:
[0007] in, is the disturbance coefficient, is the position coordinate of the mobile person, and the total wind speed field is the vector sum of the disturbance wind speed, as shown in the following formula: .
[0008] Preferably, in step S4, the diffusion coefficient correction is based on the traditional Gaussian diffusion coefficient and the turbulence enhancement effect caused by the superposition of the disturbance, as shown in the following formula:
[0009] in is the direction-dependent turbulence enhancement parameter, is the perturbation coefficient, the axial diffusion perturbation coefficient The maximum value is 0.1 to 0.2, and the radial diffusion disturbance coefficient is the smaller value 0.05 to 0.1.
[0010] Preferably, in step S5, when the evacuation path is planned using the NSGA-II algorithm, the initial population is generated randomly, and genetic operations such as non-dominated sorting and crossover mutation are continuously iterated until the maximum number of iterations is reached and the Pareto optimal solution set is output, thereby providing decision makers with a multi-objective balanced evacuation path plan.
[0011] A personnel evacuation modeling system for ship toxic gas leakage scenarios, characterized by comprising: The human agent construction module is used to model each person on the ship as an independent intelligent agent, set their basic physical properties, including age, mobility, etc., and differentiate the mobility according to the characteristics of people of different age groups; The speed correction module includes a toxic gas correction submodule and a congestion correction submodule. The former is used to determine the toxic gas correction coefficient based on the toxic gas concentration to adjust the movement speed of personnel, while the latter sets the congestion correction coefficient based on the number of agents in the unit grid, so that the speed changes with the degree of congestion. The gas diffusion correction module, based on the Gaussian puff model and coupled with the personnel disturbance quantification submodule, analyzes the relationship between personnel movement speed and local flow field, uses the turbulence intensity of personnel movement to correct the local wind speed, and optimizes the diffusion coefficient based on the cabin structure and personnel distribution, achieving dynamic coupling and closed-loop feedback between personnel evacuation and gas diffusion; The path planning module uses the NSGA-II algorithm to solve the multi-factor comprehensive objective function, and plans the evacuation path considering factors such as toxic gas concentration, personnel density, and path distance. According to different stages of toxic gas leakage, the module adopts the corresponding shortest path algorithm, low concentration priority strategy, and personnel path following strategy to make evacuation decisions.
[0012] Preferably, the personnel disturbance quantification submodule in the poison gas diffusion correction module adopts a dynamic weight superposition rule when processing disturbances caused by multiple personnel, as shown in the following formula:
[0013] Weight It is related to the distance as shown in the following formula:
[0014] Substituting the above correction parameters into the basic model, the dynamic concentration field is obtained, and the formula is as follows:
[0015] Among them, the time integral term reflects the dynamic impact of personnel movement on the migration path of smoke puffs.
[0016] Preferably, the path planning module adopts the shortest path algorithm for planning in the early stage of the toxic gas leakage, and the objective function is shown in the following formula:
[0017] in, The physical path from the personnel location to the safe exit; In the middle stage of toxic gas leakage, a low concentration priority strategy is adopted, and the path objective function is shown in the following formula:
[0018] in, is a node In time The concentration of toxic gas, is the node distance; In the later stage of the gas leak, if a crew member's path is detected, there is an 80% probability of following it. If an ordinary personnel path is detected, there is a 50% probability of following it. If no personnel path is followed, a path is selected according to the decision function. The formula is as follows:
[0019] in, is the physical length of the path, is a node In time The concentration of toxic gas, is a node In time The population density, is the node distance, , weight 、 、 It will be adjusted dynamically according to the diffusion of poison gas. When the diffusion speed of poison gas is fast, it will increase The value of , in order to pay more attention to the exposure to toxic gases on the path, when the crowding situation is more serious, it will increase to avoid selecting overly crowded paths.
[0020] Preferably, it also includes a personnel evacuation priority determination unit and a data collection and update module. The personnel evacuation priority determination unit is used to arrange candidate paths in ascending order according to the path cost function value when selecting a path, and give priority to the path with the lowest cost. When the costs of multiple paths are the same, ordinary personnel give priority to the path with low personnel density, and the elderly, children and other people with limited mobility give priority to the shortest path. The data collection and update module is used to collect information such as the concentration of toxic gas, the location and status of personnel in the ship in real time, and update these data to relevant modules in a timely manner to ensure the accuracy and real-time performance of the modeling process and provide a reliable basis for personnel evacuation.
[0021] Preferably, when the path planning module plans the path using the NSGA-II algorithm, the path distance matrix D, the evacuation path toxic gas concentration matrix C at different times, and the personnel distribution matrix N at different times are set. After the initial population is generated, the population is non-dominated sorted, and after selection, crossover, and mutation operations, the offspring and parent populations are compared. If they are different, the subsequent operations are continued. If they are the same, the individuals are recoded and the personnel evacuation paths are reallocated according to the real-time changes in the toxic gas concentration in the ship and the dynamic movement of the personnel.
[0022] The present invention provides a personnel evacuation modeling method and system for ship gas leakage scenarios. It has the following beneficial effects: 1. The present invention introduces a personnel disturbance quantification module based on the Gaussian puff model, combines it with the ship cabin environment, constructs a corrected relationship between the turbulence intensity of personnel movement and the local wind speed, and optimizes the diffusion coefficient according to the cabin structure and personnel distribution, thereby realizing the dynamic coupling and closed-loop feedback of personnel evacuation and toxic gas diffusion. This innovative mechanism can accurately predict the diffusion trend of toxic gas inside the ship in real time, and dynamically adjust the toxic gas diffusion model according to the movement and evacuation behavior of personnel. Compared with traditional solutions, this method significantly improves the accuracy of toxic gas diffusion prediction, and provides a more scientific and reliable basis for emergency decision-making and the formulation of personnel evacuation plans in toxic gas leakage accidents, thereby effectively improving the safety management level and emergency response capabilities of ships when facing toxic gas leakage scenarios.
[0023] 2. The present invention innovatively combines the personnel evacuation priority with the NSGA-II algorithm, comprehensively considering multiple factors such as toxic gas concentration, personnel density, and path distance, and adopts the shortest path algorithm, low concentration priority strategy, and strategy of following the crew or ordinary personnel paths to make evacuation path decisions according to the toxic gas leakage situation at different stages. This intelligent evacuation path planning strategy can dynamically adjust the evacuation path according to actual conditions, give priority to the safety of specific groups of people, and take into account the evacuation efficiency of ordinary personnel. In the event of a toxic gas leakage accident, this strategy can significantly reduce the risk of personnel being exposed to a toxic gas environment, reduce evacuation delays caused by problems such as congestion and improper path selection, thereby effectively improving the evacuation success rate and survival probability of personnel, and providing a more powerful guarantee for the life safety of ship personnel.
[0024] 3. The personnel evacuation modeling method and system proposed in the present invention have powerful data-driven optimization and adaptive capabilities. By collecting information such as the concentration of toxic gas, the location and status of personnel in the ship in real time, and updating these data to the relevant modules in a timely manner, the model can dynamically adjust the personnel evacuation path planning and toxic gas diffusion prediction in real time. This adaptive mechanism based on data feedback enables the model to quickly respond to various changes in the ship's toxic gas leakage scenario, such as the sudden acceleration of the toxic gas diffusion rate, the local sharp increase in the density of personnel, etc., thereby ensuring the real-time and effectiveness of the evacuation plan. Compared with traditional static models, the model of the present invention can better adapt to complex and changeable ship environments and emergency situations, and provide more accurate and reliable guidance for the evacuation of personnel in ship toxic gas leakage accidents. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 It is the overall flow chart of the present invention. DETAILED DESCRIPTION
[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the present specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0027] Please see the attached Figure 1 The embodiment of the present invention provides a personnel evacuation modeling method for a ship gas leakage scenario, comprising the following steps: S1. Construct a ship personnel intelligent agent model, treating each person as an independent intelligent agent, setting their basic physical properties, including age, mobility, etc., and differentiate the mobility of personnel of different age groups based on ergonomic data; the reaction time of the personnel intelligent agent is randomly distributed with an average of 3 seconds. During the reaction time, the personnel remain stationary and enter the evacuation state after the reaction period ends.
[0028] Specifically, when building a crew evacuation model for a ship gas leak scenario, the first step is to construct a crew agent model, treating each crew member as an independent agent. This is because each crew member makes their own decisions during an emergency evacuation and possesses a certain degree of independence. Next, the basic physical properties of these agents need to be determined, with age and mobility being key factors. This is because people of different ages vary significantly in their physical fitness and agility, which directly impacts the speed and behavior of evacuees in a gas leak scenario. For example, children and the elderly are generally slower, while young people are more mobile. By referencing relevant ergonomic data, we can set reasonable differentiation in mobility for people of different age groups. For example, setting the mobility of children and the elderly to approximately 33% lower than that of young people can make the model more realistic. Furthermore, the reaction time of the crew agent is a crucial factor. In real-world scenarios, people do not react immediately upon receiving a gas leak alarm. Instead, they take a while to react. This reaction time is randomly distributed, but has a mean of approximately 3 seconds. During this time, people typically remain motionless due to panic or trying to assess the situation. Once the reaction time is over, people officially enter the evacuation state, beginning to follow the pre-determined evacuation routes and strategies. This setup allows for a more accurate simulation of people's reactions during the initial stages of a gas leak, improving the reliability and realism of the entire evacuation model.
[0029] S2. Analyze the impact of toxic gas diffusion on personnel movement speed and determine the toxic gas correction factor. That is, when the toxic gas concentration reaches the critical value, the personnel movement speed decreases by 10% for each increase in concentration. When a toxic gas leak occurs, the diffusion of toxic gas inside the ship will have a significant impact on the evacuation of personnel. As the concentration of toxic gas increases, the respiratory and nervous systems of personnel will be damaged to varying degrees after inhaling the toxic gas, resulting in symptoms such as difficulty breathing, physical fatigue, and slow movement, which will inevitably affect the speed of personnel movement. Therefore, when constructing a personnel evacuation model, the impact of toxic gas diffusion on the speed of personnel movement must be considered. In order to quantify this impact more accurately, a toxic gas correction coefficient needs to be determined. Specifically, when the toxic gas concentration reaches a certain critical value, the speed of personnel movement begins to be significantly affected. And with each unit increase in toxic gas concentration (kg / m 3 ), the movement speed of personnel will decrease proportionally by 10%. This setting can be derived from a large amount of experimental data and actual case analysis. In this way, the relationship between toxic gas concentration and personnel movement speed can be quantified, thereby more accurately simulating the evacuation of personnel under different toxic gas concentrations in the model, providing a more scientific basis for subsequent evacuation route planning and strategy formulation.
[0030] S3. Considering the impact of congestion on movement speed during evacuation, a congestion correction coefficient is set. When the number of agents in a unit grid reaches 3 or more, the speed is halved. When the number reaches 5 or more, the speed is further reduced. Specifically, in a ship gas leak scenario, evacuation often unfolds rapidly within a confined space (such as a cabin), which can easily lead to crowding. When crowding occurs, factors such as pushing and collisions between people, as well as spatial constraints, significantly impact movement speed. Therefore, when building evacuation models, it is crucial to consider the impact of crowding on movement speed during evacuation. To quantify this effect, a crowding correction factor is established. Specifically, when the number of agents (i.e., individuals) within a unit grid (which can be considered a basic spatial unit within a ship, used for counting and analyzing personnel distribution) reaches three or more, movement space is significantly restricted, and movement speed is halved. When the number of agents within a unit grid increases to five or more, crowding becomes even more severe, making movement nearly impossible, and movement speed further reduced. This setting, based on research on ship interior spatial layout and evacuation behavior, more realistically reflects evacuation behavior in crowded environments. By incorporating the crowding correction coefficient into the model, the actual situation of personnel evacuation in a ship gas leakage accident can be simulated more accurately, thereby improving the accuracy and practicality of the model.
[0031] S4. Based on the Gaussian puff model, a personnel disturbance quantification module is introduced. By analyzing the relationship between personnel movement speed and local flow field, the local wind speed is corrected using the turbulence intensity of personnel movement, and the diffusion coefficient is optimized according to the cabin structure and personnel distribution, so as to realize the dynamic coupling and closed-loop feedback of personnel evacuation and gas diffusion. When the local wind speed disturbance is superimposed, inside the closed and windless ship, the background wind speed is zero. It is assumed that the movement of personnel generates an additional wind speed field, and the direction is consistent with the direction of movement. It is assumed that the movement speed of the i-th person is , the disturbance range radius is , then the disturbance wind speed caused by it is shown in the following formula:
[0032] in, is the disturbance coefficient, is the position coordinate of the mobile person, and the total wind speed field is the vector sum of the disturbance wind speed, as shown in the following formula: .
[0033] The diffusion coefficient correction is based on the traditional Gaussian diffusion coefficient and adds the turbulence enhancement effect caused by the disturbance, as shown in the following formula:
[0034] in is the direction-dependent turbulence enhancement parameter, is the perturbation coefficient, the axial diffusion perturbation coefficient
[0035] The maximum value is 0.1 to 0.2, and the radial diffusion disturbance coefficient is the smaller value 0.05 to 0.1.
[0036] Specifically, while the traditional Gaussian puff model can describe the diffusion of toxic gas to a certain extent, it has significant limitations in the case of a ship-borne gas leak. The main problem is that it ignores the perturbations caused by the evacuation of personnel, resulting in significant deviations between the predicted results and the actual situation. To address this issue, this step improves and expands upon the Gaussian puff model.
[0037] First, a personnel disturbance quantification module was introduced. The core function of this module is to analyze the relationship between personnel movement speed and the local flow field. Because inside a ship, the movement of personnel will drive the surrounding air flow, forming an additional wind speed field, which in turn affects the diffusion path and speed of the toxic gas. Specifically, the speed and direction of personnel movement are consistent with the direction of the additional wind speed field generated. By setting the movement speed and disturbance range radius of the i-th person, the disturbance wind speed caused by them can be calculated. In the case of a closed environment inside a ship with a background wind speed of zero, the total wind speed field is equal to the vector sum of the disturbance wind speeds of all personnel, which can more accurately reflect the actual wind speed field inside the ship.
[0038] Secondly, considering that human disturbance not only changes the wind speed field but also enhances the turbulence effect, thereby affecting the diffusion coefficient of the poison gas, the traditional Gaussian diffusion coefficient is modified by superimposing the turbulence enhancement effect caused by the disturbance. Specifically, a direction-dependent turbulence enhancement parameter and disturbance coefficient are introduced, with the axial diffusion disturbance coefficient taking a larger value (0.1-0.2) and the radial diffusion disturbance coefficient taking a smaller value (0.05-0.1) to more accurately simulate the diffusion of poison gas in different directions.
[0039] These improvements achieve dynamic coupling and closed-loop feedback between evacuation and gas diffusion. In other words, in the model, evacuation behavior influences gas diffusion, which in turn influences evacuation decisions (such as movement speed and path selection). This two-way dynamic interaction enables the model to more realistically reflect the complexities of a ship-borne gas leak scenario, improving its accuracy and reliability and providing strong support for developing scientific and rational evacuation strategies.
[0040] S5. The NSGA-II algorithm is introduced to solve an objective function that comprehensively considers multiple factors for personnel path planning. Planning factors include gas concentration, personnel density, and path distance. Evacuation path decisions are made using the shortest path algorithm, a low-concentration priority strategy, and a strategy that follows the paths of crew members or ordinary personnel, depending on the gas leakage situation at different stages. When planning evacuation paths using the NSGA-II algorithm, the initial population is randomly generated. Genetic operations such as non-dominated sorting and crossover mutation are continuously iterated until the maximum number of iterations is reached, and a Pareto optimal solution set is output, providing decision makers with a multi-objective balanced evacuation path solution.
[0041] A personnel evacuation modeling system for ship toxic gas leakage scenarios, characterized by comprising: The human agent construction module is used to model each person on the ship as an independent intelligent agent, set their basic physical properties, including age, mobility, etc., and differentiate the mobility according to the characteristics of people of different age groups; The speed correction module includes a toxic gas correction submodule and a congestion correction submodule. The former is used to determine the toxic gas correction coefficient based on the toxic gas concentration to adjust the movement speed of personnel, while the latter sets the congestion correction coefficient based on the number of agents in the unit grid, so that the speed changes with the degree of congestion. The gas diffusion correction module is based on the Gaussian puff model and is coupled with the personnel disturbance quantification submodule. It analyzes the relationship between personnel movement speed and local flow field, uses the turbulence intensity of personnel movement to correct the local wind speed, and optimizes the diffusion coefficient based on the cabin structure and personnel distribution to achieve dynamic coupling and closed-loop feedback between personnel evacuation and gas diffusion. The personnel disturbance quantification submodule in the gas diffusion correction module adopts a dynamic weight superposition rule when dealing with disturbances from multiple personnel, as shown in the following formula:
[0042] Weight It is related to the distance as shown in the following formula:
[0043] Substituting the above correction parameters into the basic model, the dynamic concentration field is obtained, and the formula is as follows: Among them, the time integral term reflects the dynamic impact of personnel movement on the migration path of smoke puffs.
[0044] The path planning module uses the NSGA-II algorithm to solve the multi-factor comprehensive objective function, and plans the evacuation path considering factors such as toxic gas concentration, personnel density, and path distance. According to different stages of toxic gas leakage, the module adopts the corresponding shortest path algorithm, low concentration priority strategy, and personnel path following strategy to make evacuation decisions.
[0045] The path planning module uses the shortest path algorithm for planning in the early stage of the gas leak. The objective function is shown in the following formula:
[0046] in, The physical path from the personnel location to the safe exit; In the middle stage of toxic gas leakage, a low concentration priority strategy is adopted, and the path objective function is shown in the following formula:
[0047] in, is a node In time The concentration of toxic gas, is the node distance; In the later stage of the gas leak, if a crew member's path is detected, there is an 80% probability of following it. If an ordinary personnel path is detected, there is a 50% probability of following it. If no personnel path is followed, a path is selected according to the decision function. The formula is as follows:
[0048] in, is the physical length of the path, is a node In time The concentration of toxic gas, is a node In time The population density, is the node distance, , weight 、 、 It will be adjusted dynamically according to the diffusion of poison gas. When the diffusion speed of poison gas is fast, it will increase The value of , in order to pay more attention to the exposure to toxic gases on the path, when the crowding situation is more serious, it will increase to avoid selecting overly crowded paths.
[0049] When the path planning module uses the NSGA-II algorithm to plan the path, it sets the path distance matrix D, the gas concentration matrix C of the evacuation path at different times, and the personnel distribution matrix N at different times. After the initial population is generated, the population is non-dominated sorted, and after the selection, crossover, and mutation operations, the offspring and parent populations are compared. If they are different, the subsequent operations are continued. If they are the same, the individuals are recoded and the personnel evacuation paths are reallocated according to the real-time changes in the gas concentration in the ship and the dynamic movement of the personnel.
[0050] It also includes a personnel evacuation priority determination unit and a data collection and update module. The personnel evacuation priority determination unit is used to arrange candidate paths in ascending order according to the path cost function value when selecting a path, and give priority to the path with the lowest cost. When the costs of multiple paths are the same, ordinary personnel give priority to the path with low population density, and the elderly, children and other people with limited mobility give priority to the shortest path. The data collection and update module is used to collect information such as the toxic gas concentration, personnel location and status in the ship in real time, and update this data to the relevant modules in a timely manner to ensure the accuracy and real-time performance of the modeling process, providing a reliable basis for personnel evacuation.
[0051] In an emergency evacuation scenario involving a ship's toxic gas leak, the basic physical characteristics of different individuals vary significantly, significantly impacting their evacuation behavior and capabilities. To more realistically simulate the evacuation process, it's necessary to properly configure the basic physical properties of the agent.
[0052] During the evacuation process, people of different ages have different mobility capabilities. Children and the elderly move relatively slowly, while young people have stronger mobility. Based on this, the speed will be adjusted according to environmental factors, as shown in the following formula: The above formula is based on ergonomic data. The mobility of children and the elderly decreases by about 33%.
[0053] In the scenario of a ship-borne gas leak, the spread of the gas will change the environmental conditions within the ship, and the congestion caused by the evacuation process will also affect the movement of personnel. These factors will significantly affect the movement speed of personnel. Therefore, it is necessary to dynamically correct the basic movement speed of personnel to more accurately simulate the actual evacuation process, as shown in the following formula: in, is the corrected personnel movement speed, is the poison gas correction factor, is the crowding correction factor.
[0054] When the concentration of poisonous gas reaches a certain value, it will affect the respiratory system and nervous system of personnel, causing breathing difficulties, slow movement, etc., thereby reducing the movement speed. As the concentration of poisonous gas increases further, each increase (Unit: kg / m 3 ), this effect will cause the speed to decrease proportionally by 10%. The poison gas correction factor is as follows:
[0055] in, is the critical value that affects personnel movement.
[0056] During the evacuation process, when the number of people in a unit grid increases, that is, when congestion occurs, people will block and collide with each other, resulting in limited movement space and difficulty maintaining normal movement speed. When the number of agents in a unit grid reaches 3 or more, the movement of people will be significantly hindered and the speed will be halved; and when the number reaches 5 or more, the congestion will be further intensified, making it more difficult for people to move and the speed will be further reduced. Congestion correction coefficient As shown in the following formula: After receiving the gas leak alarm, the agent enters the reaction period and remains stationary for the reaction time (randomly distributed, with an average of 3 seconds). After the reaction period, it enters the evacuation state, as shown below:
[0057] in, for The status of personnel at all times, The status is not responding. In evacuation status, For the reaction time.
[0058] The impact of toxic gas on personnel evacuation is determined by the toxic gas exposure dose. The toxic gas exposure dose is continuously accumulated during the reaction time and used for subsequent health status judgment. The exposure dose calculation method is shown in the following formula: in, for The exposure dose at the location at the moment, in kg / m 3 , for The concentration of toxic gas at the location at the moment, in kg / m 3 , is the time step, set to 1 second.
[0059] The health status of personnel will decline with the increase of exposure dose. This article uses numerical values to represent the health status of personnel, as shown in the following formula: in, For personnel The health value at the moment, the initial health value is 100, and the exposure dose increases (kg / m3), the health value decreases by 10 points. When it drops to 0, it enters the poisoned state and loses the ability to move, as shown in the following formula:
[0060] The Gaussian model is generally used for toxic gas diffusion. The Gaussian model can be divided into the plume model and the puff model. The Gaussian plume model is suitable for continuous leakage scenarios, while the puff model is suitable for instantaneous leakage. This paper uses the puff model to quantitatively analyze the transient leakage of toxic gas in ships. The uncorrected Gaussian puff model describes the instantaneous leakage concentration distribution of toxic gas as shown in the following formula: in, For coordinates in The gas concentration at the moment (kg / m 3 ), is the total instantaneous leakage (kg), is the background wind speed (m / s), 、 、 is the diffusion coefficient in each direction, is the height of the leak source.
[0061] The traditional Gaussian puff model describes the fixed distribution of toxic gas concentration and does not consider the impact of the disturbed wind field generated by the movement of people during the evacuation process on the diffusion of toxic gas, which leads to errors. Therefore, this paper regards human activities as the dynamic disturbance source of toxic gas diffusion.
[0062] The scene of poison gas spreading in the ship is as follows: 1. Single leak source with multiple puffs released, or multiple leak sources leaking simultaneously For the Leakage source, ,exist The moment generated A puff of smoke at a point in space The concentration at The improved Gaussian puff model is calculated as follows:
[0063] in It is The leakage source produces A puff of smoke at a moment At a point in space The concentration at It is The leakage source produces The initial mass of the puff; is the initial position of the puff Comprehensive consideration Leakage source, at point in space Total concentration of toxic gas at The sum of the smoke concentrations generated by all leakage sources:
[0064] in, It is The leakage source at time The number of puffs generated and still expanding 2. Improved obstacle reflection distribution Obstacle reflection expression: When the smoke encounters an obstacle, the reflected part of the poisonous gas will form a new distribution near the surface of the obstacle. The leakage source produces A puff, when it diffuses with the The collision of an obstacle causes reflection, and the concentration contribution of the reflected part is as follows:
[0065] in, is the concentration of the reflected part, are the coordinates of the collision point, is the rebound coefficient, which is related to the obstacle type and surface characteristics. It is the center of the reflected smoke cloud. is the diffusion coefficient of the reflected puff, which is determined by the angle of incidence and the characteristics of the obstacle.
[0066] Determination of reflection direction and diffusion coefficient: The determination of the reflection direction follows the analogy principle of the law of reflection of light in the fluid diffusion scenario. In geometric optics, when light encounters a reflective surface, the angle of incidence is equal to the angle of reflection. For the case where smoke diffusion encounters an obstacle, we can compare the direction of movement of the smoke to the propagation direction of light, and the surface of the obstacle to the reflective surface. The collision process will change the motion state of the smoke, generate new turbulence, etc., and then cause the degree of diffusion of the smoke in all directions to change. Reflection direction It can be expressed as follows:
[0067] in, is the puff incident velocity vector, is the normal vector of the obstacle surface. is a vector dot product operation, the result of which is a scalar, namely:
[0068] The above result is multiplied by 2 and Multiplication is based on the principle that the incident angle and the reflection angle are equal, and accurately calculates the change in the reflection direction relative to the incident direction. Subtracting this change, we get the reflection direction vector .
[0069] When the smoke puff has an incident speed When hitting the obstacle surface, its velocity can be decomposed into components parallel to the obstacle surface and the component perpendicular to the obstacle surface , as shown in the following formula:
[0070] in, is the incident velocity vector and the obstacle normal vector The increment of the diffusion coefficient in each direction is linearly related to the component of the incident velocity in the direction parallel to and perpendicular to the obstacle. Therefore, the increment of the diffusion coefficient in each direction can be obtained as follows:
[0071] in, It is a coefficient related to the roughness of the obstacle surface. Combining the original diffusion and the additional diffusion caused by the obstacle, we can get the diffusion coefficient of the reflected smoke and the incident angle. The relationship is as follows:
[0072] In summary, the improved total concentration field is as follows:
[0073] in, is the total number of collisions of the current puff, It is Contribution of reflected puff concentration produced by secondary collision.
[0074] These improvements enable the model to more accurately simulate the diffusion of toxic gases in the complex environment inside a ship, including the impact of obstacles, thereby providing more reliable data support for personnel evacuation and emergency response.
[0075] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.
Claims
1. A personnel evacuation modeling method for ship gas leakage scenario, characterized in that: The following steps are involved: S1. Build a ship crew agent model, treating each crew member as an independent agent and setting their basic physical attributes, including age and mobility. Differentiate the mobility of crew members of different age groups based on ergonomic data. S2. Analyze the impact of toxic gas diffusion on personnel movement speed and determine the toxic gas correction factor. That is, when the toxic gas concentration reaches the critical value, the personnel movement speed decreases by 10% for each increase in concentration. S3. Considering the impact of congestion on movement speed during evacuation, a congestion correction coefficient is set. When the number of agents in a unit grid reaches 3 or more, the speed is halved. When the number reaches 5 or more, the speed is further reduced. S4. Based on the Gaussian puff model, a personnel disturbance quantification module is introduced. By analyzing the relationship between personnel movement speed and local flow field, the local wind speed is corrected using the turbulence intensity of personnel movement. The diffusion coefficient is optimized according to the cabin structure and personnel distribution, realizing the dynamic coupling and closed-loop feedback of personnel evacuation and toxic gas diffusion. S5. The NSGA-II algorithm is introduced to solve the objective function that comprehensively considers multiple factors and perform personnel path planning. The planning factors include toxic gas concentration, personnel density, and path distance. According to the toxic gas leakage situation at different stages, the shortest path algorithm, low concentration priority strategy, and strategy of following the crew or ordinary personnel path are used to make evacuation path decisions.
2. A personnel evacuation modeling method for ship gas leakage scenario according to claim 1, characterized in that: In the step S1, the reaction time of the human agent is randomly distributed with an average of 3 seconds. During the reaction time, the human agent remains still and enters the evacuation state after the reaction period ends.
3. The personnel evacuation modeling method for ship gas leakage scenario according to claim 1 is characterized in that: In the step S4, when the local wind speed disturbance is superimposed, inside the closed and windless ship, the background wind speed is zero. Assuming that the movement of personnel generates an additional wind speed field, the direction is consistent with the direction of movement. Assuming that the movement of personnel generates an additional wind speed field, the direction is consistent with the direction of movement, and the moving speed of the i-th person is , the disturbance range radius is , then the disturbance wind speed caused by it is shown in the following formula: ； in, is the disturbance coefficient, is the position coordinate of the mobile person, and the total wind speed field is the vector sum of the disturbance wind speed, as shown in the following formula: 。 4. The personnel evacuation modeling method for ship gas leakage scenario according to claim 1 is characterized in that: In the S4 step, the diffusion coefficient correction is based on the traditional Gaussian diffusion coefficient and the turbulence enhancement effect caused by the superposition of the disturbance, as shown in the following formula: ； in is the direction-dependent turbulence enhancement parameter, is the perturbation coefficient, axial diffusion perturbation coefficient The maximum value is 0.1 to 0.2, and the radial diffusion disturbance coefficient is the smaller value 0.05 to 0.
1.
5. The personnel evacuation modeling method for ship gas leakage scenario according to claim 1 is characterized in that: In step S5, when the NSGA-II algorithm is used to plan the evacuation path, the initial population is generated randomly, and genetic operations such as non-dominated sorting and crossover mutation are continuously iterated until the maximum number of iterations is reached, and the Pareto optimal solution set is output to provide decision makers with a multi-objective balanced evacuation path plan.
6. A personnel evacuation modeling system for ship gas leakage scenarios, according to a personnel evacuation modeling method for ship gas leakage scenarios according to any one of claims 1 to 5, characterized in that: include: The human agent construction module is used to model each person on the ship as an independent intelligent agent, set their basic physical properties, including age, mobility, etc., and differentiate the mobility according to the characteristics of people of different age groups; The speed correction module includes a toxic gas correction submodule and a congestion correction submodule. The former is used to determine the toxic gas correction coefficient based on the toxic gas concentration to adjust the movement speed of personnel, while the latter sets the congestion correction coefficient based on the number of agents in the unit grid, so that the speed changes with the degree of congestion. The gas diffusion correction module, based on the Gaussian puff model and coupled with the personnel disturbance quantification submodule, analyzes the relationship between personnel movement speed and local flow field, uses the turbulence intensity of personnel movement to correct the local wind speed, and optimizes the diffusion coefficient based on the cabin structure and personnel distribution, achieving dynamic coupling and closed-loop feedback between personnel evacuation and gas diffusion; The path planning module uses the NSGA-II algorithm to solve the multi-factor comprehensive objective function, and plans the evacuation path considering factors such as toxic gas concentration, personnel density, and path distance. According to different stages of toxic gas leakage, the module adopts the corresponding shortest path algorithm, low concentration priority strategy, and personnel path following strategy to make evacuation decisions.
7. The personnel evacuation modeling system for ship gas leakage scenario according to claim 6, characterized in that: The personnel disturbance quantification submodule in the poison gas diffusion correction module adopts a dynamic weight superposition rule when processing disturbances from multiple personnel, as shown in the following formula: ； Weight It is related to the distance as shown in the following formula: ； Substituting the above correction parameters into the basic model, the dynamic concentration field is obtained, and the formula is as follows: ; Among them, the time integral term reflects the dynamic impact of personnel movement on the migration path of smoke puffs.
8. The personnel evacuation modeling system for ship gas leakage scenario according to claim 6, characterized in that: The path planning module uses the shortest path algorithm for planning in the early stage of the toxic gas leakage, and the objective function is shown in the following formula: ； in, The physical path from the personnel's location to the safe exit; In the middle stage of toxic gas leakage, a low concentration priority strategy is adopted, and the path objective function is shown in the following formula: ； in, is a node In time The concentration of toxic gas, is the node distance; In the later stage of the gas leak, if a crew member's path is detected, there is an 80% probability of following it. If an ordinary personnel path is detected, there is a 50% probability of following it. If no personnel path is followed, a path is selected according to the decision function. The formula is as follows: ； in, is the physical length of the path, is a node In time The concentration of toxic gas, is a node In time The population density, is the node distance, , weight 、 、 It will be adjusted dynamically according to the diffusion of poison gas. When the diffusion speed of poison gas is fast, it will increase The value of , in order to pay more attention to the exposure to toxic gases on the path, when the crowding situation is more serious, it will increase to avoid selecting overly crowded paths.
9. The personnel evacuation modeling system for ship gas leakage scenario according to claim 6, characterized in that: It also includes a personnel evacuation priority determination unit and a data collection and update module. The personnel evacuation priority determination unit is used to arrange candidate paths in ascending order according to the path cost function value when selecting a path, and give priority to the path with the lowest cost. When the costs of multiple paths are the same, ordinary personnel give priority to the path with low population density, and the elderly, children and other people with limited mobility give priority to the shortest path. The data collection and update module is used to collect information such as the toxic gas concentration, personnel location and status in the ship in real time, and update this data to the relevant modules in a timely manner to ensure the accuracy and real-time performance of the modeling process, providing a reliable basis for personnel evacuation.
10. The personnel evacuation modeling system for ship gas leakage scenario according to claim 6, characterized in that: When planning a path using the NSGA-II algorithm, the path planning module sets a path distance matrix D, a gas concentration matrix C of the evacuation path at different times, and a personnel distribution matrix N at different times. After the initial population is generated, the population is non-dominated sorted. After selection, crossover, and mutation operations, the offspring and parent populations are compared. If they are different, the subsequent operations are continued. If they are the same, the individuals are recoded and the personnel evacuation paths are reallocated based on the real-time changes in the gas concentration in the ship and the dynamic movement of the personnel.