Cruise ship personnel optimal evacuation path planning method considering storage position of life jacket

Abstract

The invention relates to the technical field of cruise ship evacuation, in particular to a cruise ship personnel optimal evacuation path planning method considering a life jacket storage position, which comprises the following steps: step 1, constructing a cruise ship three-dimensional model; step 2, construction and simulation of an evacuation scene; step 3, data collection; step 4, classifying personnel evacuation features; and step 5, constructing and solving a personnel evacuation path optimization planning model. According to the invention, the setting of the life jacket storage scene can be optimized, the evacuation efficiency of class personnel is improved, and a more effective safe evacuation strategy is formulated.

Description

Technical Field

[0001] The present invention relates to the technical field of cruise ship evacuation, and in particular to an optimal evacuation path planning method for cruise ship personnel taking into account storage positions of life jackets. Background Art

[0002] A cruise ship is a large ship that sails on the sea, usually accommodating thousands of passengers and crew. Although cruise travel is considered a very safe way to travel, the internal structure of a cruise ship is complex and may cause serious casualties in the event of a fire, collision, flooding or other emergency.

[0003] Compared with the ground space, the passages in the cabins of large cruise ships are usually narrower and more crowded. In emergency scenarios, people are overly panicked, and there are certain obstacles to the organization and evacuation of people in an ideal state. Existing research focuses on the diversity of movement characteristics of evacuated personnel, the influence of ship movement state and the structure of the guest room on the evacuation behavior, speed and time of personnel, and pays less attention to the influence of the design of life-saving equipment inside the cruise ship on the safety of personnel evacuation. Among them, the choice of life jacket storage has a great impact on the evacuation time of personnel. At present, there are two main ways to set the placement of cruise life jackets: one is to store all life jackets near the assembly point or the lifeboat boarding point so that they can be used at any time in an emergency; the other is that some large luxury cruise ships place life jackets in each guest room, and only place a small proportion of life jackets at the assembly point. However, for large cruise ships, whether the setting of the life jacket storage place and the storage ratio can ensure the safe evacuation of all personnel in emergency scenarios needs further study. Summary of the invention

[0004] The present invention provides an optimal evacuation route planning method for cruise ship personnel taking into account the storage location of life jackets, which can optimize the life jacket storage scene setting, improve the evacuation efficiency of different types of personnel and formulate a more effective safety evacuation strategy.

[0005] The optimal evacuation path planning method for cruise ship personnel considering the storage position of life jackets according to the present invention comprises the following steps:

[0006] Step 1: Construction of cruise ship 3D model;

[0007] Step 2: Evacuation scenario construction and simulation;

[0008] Step 3: Data collection;

[0009] Step 4: Classification of personnel evacuation characteristics;

[0010] Step 5: Construct and solve the personnel evacuation path optimization planning model.

[0011] Preferably, in step 1, specifically:

[0012] 1.1) Collect the cruise ship plan, determine the cabin structure layout, emergency exit location, main vertical area, floor height, number of stairs, assembly point location information, and use 3D modeling software to build a 3D model of the cruise ship;

[0013] 1.2) Determine the evacuation routes for personnel in each main vertical zone, and plan the width and capacity of the evacuation passages from the personnel activity area to the safe exit;

[0014] 1.3) According to the actual passenger capacity of the model, the number of people to be evacuated is set, and the evacuation parameters are constructed according to gender, age, movement speed and initial position distribution;

[0015] 1.4) Combined with the number, load and distribution of passengers inside the cruise ship stipulated in the advanced evacuation analysis in the IMO circular, it is assumed that the cruise ship is in daytime operation, 3 / 4 of the passengers are distributed in public places, and 1 / 4 of the passengers are distributed in the guest rooms. The passengers to be evacuated are imported into the cruise ship 3D model in proportion.

[0016] Preferably, in step 2, specifically:

[0017] 2.1) Determine the storage location and proportion of life jackets in the model, taking into account the IMO circular that life jackets should be easily accessible and quickly distributed inside cruise ships, and place life jackets at different proportions at the assembly point and inside the guest rooms;

[0018] 2.2) Use Pathfinder personnel evacuation simulation software to identify the target safe evacuation area and establish personnel evacuation scenarios when different life jackets are stored in different proportions at the assembly point and in the guest rooms;

[0019] 2.3) Observe the evacuation process during the simulation, analyze the changes in the evacuation time of the cruise ship, the length of the evacuation route and the congestion time parameters during the evacuation process, and provide basic data for subsequent analysis.

[0020] Preferably, in step 3, specifically:

[0021] 3.1) Collect data on personnel evacuation under different life jacket storage scenarios, and clarify the use of life jackets at assembly points and in guest rooms;

[0022] 3.2) Collect data on the density of people in the evacuation stairwells, evacuation time and the length of the evacuation route to ensure that the research is based on real data.

[0023] Preferably, in step 4, specifically:

[0024] 4.1) The data of evacuation time, evacuation route length and evacuation congestion time of all personnel under different life jacket storage scenarios are sorted into a data set to ensure that the correspondence between scenarios and personnel is complete and the values ​​are correct;

[0025] 4.2) Use the pandas library to load the data and normalize the data to ensure that different feature data have the same scale;

[0026] 4.3) The evacuation time, evacuation route length and evacuation congestion time of each person are selected as feature variables, and the K-means mean clustering algorithm is used to classify the evacuated personnel, divide the crowd categories, and determine the classification of personnel groups under different life jacket storage scenarios;

[0027] 4.4) Further analyze the category characteristics and the distribution of each category in different life jacket storage scenarios, and calculate the average evacuation time, average evacuation path length and average evacuation congestion time of each category;

[0028] 4.5) Clarify the evacuation safety performance of various categories of personnel under various life jacket storage scenarios, determine the current optimal storage scenario for life jackets, provide a classification basis for the path planning optimization model, and help formulate different path planning strategies for different groups of personnel.

[0029] Preferably, in step 4.3), specifically:

[0030] 4.3.1) Define the number of crowd categories into which the evacuees are divided as K, and select j as the index of the crowd category. Each K corresponds to an index j data point as the initial cluster center, expressed as:

[0031] x i =[t i ,d i ,c i ]

[0032]

[0033] Among them, i represents the personnel number, x i represents the three-dimensional feature vector of the i-th person, t i is the evacuation time of the i-th person, d i is the length of the evacuation path of the i-th person, c i is the evacuation congestion time of the i-th person, μ j represents the center point of the jth category, μ jt represents the mean evacuation time of the jth category, μ jd represents the mean length of the evacuation path of the jth category, μ jc represents the mean evacuation congestion time of the jth category, and 0 represents the randomly selected initial category center point;

[0034] 4.3.2) Assign the data points of evacuees in different life jacket storage scenarios to the nearest cluster center;

[0035] y i =argmin j ||x i -μ j || 2

[0036] Among them, y i Indicates the index of the population category to which the i-th person belongs;

[0037] Euclidean distance is used to calculate the similarity between evacuation time, evacuation path length and evacuation congestion time:

[0038] x i -μ j || 2 =(t i -μ jt ) 2 +(d i -μ jd ) 2 +(c i -μ jc ) 2 ;

[0039] 4.3.3) For each category, calculate the mean of the evacuation time, evacuation path length and evacuation congestion time within each category, and determine the mean as the new cluster center to update the cluster center μ j :

[0040]

[0041] Among them, |C j | represents the number of data points in the jth category, C j represents the set of data points in the jth category;

[0042] Set the objective function E and repeat the above steps until the cluster center reaches the preset maximum number of iterations;

[0043]

[0044] 4.3.4) Calculate the category to which each person belongs.

[0045] Preferably, in step 5, a path planning model is established, the optimization objective function is clarified, and after the constraints are defined, the model is optimized in combination with the artificial fish swarm algorithm AFSA, so as to construct an optimized planning model for the evacuation path of personnel under different life jacket storage scenarios.

[0046] Preferably, in step 5, specifically:

[0047] 5.1) The optimization goal is to maximize the average evacuation speed and minimize the average evacuation congestion time of different life jacket storage scenarios. The optimization goal is converted into a mathematical expression and defined as the objective function f(x) of the optimization model:

[0048]

[0049]

[0050] Among them, x is the variable set of path optimization, including path length, evacuation time and congestion time, S is the number of evacuation scenarios, J is the number of crowd categories, d sj is the average evacuation path length of the jth group of people in the sth scenario, t sj is the average evacuation time of the jth group of people in the sth scenario, v sj is the average evacuation speed of the jth group of people in the sth scenario, c sj is the average congestion time of the jth group of people in the sth scenario, α and β are the weight coefficients of evacuation speed and congestion time;

[0051] 5.2) Determine the maximum evacuation path length, maximum evacuation time and the density of people in the stairwell of the evacuation passage as constraints, and convert the constraints into mathematical expressions as the constraints of the above optimization model:

[0052] d sj ≤D max t sj ≤T max

[0053]

[0054] Among them, D max is the maximum evacuation path length, T max is the maximum evacuation time, ρ max is the maximum permissible density of people in the evacuation stairway, ρ is the permissible density of people in the evacuation stairway, P represents the number of people going down the stairwell at a certain time, and A represents the area of ​​the stairwell;

[0055] 5.3) Select the artificial fish swarm algorithm AFSA, set the algorithm parameters and constraints, solve the path planning model, and determine the optimal evacuation path allocation plan;

[0056] 5.4) According to the specific position vectors shown in the optimal evacuation path allocation scheme, the optimal evacuation scenario and crowd category with the objective function of maximizing the average evacuation speed and minimizing the average evacuation congestion time are determined.

[0057] Preferably, in step 5.3), specifically:

[0058] 5.3.1) Initialize crowd position z i Randomly distributed in the cruise ship space in the form of vectors, each category of people represents an evacuation path allocation plan;

[0059] z i =(z s1 ,z s2 ,…,z sj )

[0060] Among them, z i represents the initial crowd position coordinate set, z sj represents the two-dimensional coordinates of the s-th path in the j-th category of people;

[0061] 5.3.2) Set the number of evacuation paths in the algorithm parameters to N, and define a three-dimensional matrix (N, S, J) to represent the position of each type of people in the evacuation path, where the position of each type of people is represented by a two-dimensional matrix of S×J;

[0062] 5.3.3) It is stipulated that during the evacuation process of each category of people, evacuation shall be carried out according to the foraging behavior, grouping behavior and tail-chasing behavior of the fish school;

[0063] 5.3.4) Repeat the calculation of the evacuation behavior of the above three categories of people until the maximum number of iterations is reached or the fitness value converges to a certain degree. After stopping the calculation, output the optimal evacuation path allocation plan with the fitness value, and at the same time ensure that the evacuation efficiency of the categories of people under different life jacket storage locations is the highest.

[0064] Preferably, in step 5.3.3), specifically:

[0065] Define foraging behavior: Set the category of people to randomly choose to move to the nearest exit in the initial stage, and calculate the fitness value of the new position after moving:

[0066]

[0067] Define crowd behavior: Under the condition of controlling the density of people in the evacuation stairwell and the flow rate at the exit, find the optimal evacuation path by collaborating with different categories of people around; the category crowd first finds the crowd category with the best fitness within the perception range and judges the crowding degree, where the initial crowding degree is set to 10%-20% of the crowd size, and then chooses to move to the category crowd with the best fitness value:

[0068]

[0069] Definition of tailgating behavior: When an evacuation leader with excellent performance appears during the evacuation process, the individual evacuees in the category group will judge the difference between their own fitness value and the leader, and then drive the category group to optimize the evacuation path through the tailing effect and radiation effect, thereby accelerating the category group to approach the optimal solution of the evacuation path throughout the process:

[0070]

[0071] Among them, Δz is the moving direction vector, visual is the crowd perception range, rand(-1,1) is a random number between -1 and 1, step is the maximum step length, z c is the center position of the crowd within the perception range, z best It is the position of the crowd category with the best fitness value within the perception range.

[0072] The beneficial effects of the present invention are as follows:

[0073] The present invention focuses on the evacuation time, the length of the evacuation path and the evacuation congestion time. It takes the selection of storage places and storage ratios of life jackets on cruise ships as scenario variables, determines the categories of people in evacuation conditions, and evaluates the safety of personnel evacuation under different life jacket storage scenarios. It uses an artificial fish swarm algorithm to construct an evacuation path optimization model, thereby providing strong support for optimizing the setting of life jacket storage scenarios, improving the evacuation efficiency of classified personnel, and formulating more effective safety evacuation strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 The figure is a flow chart of a method for planning the optimal evacuation route for cruise ship personnel taking into account the storage location of life jackets. DETAILED DESCRIPTION

[0075] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0076] like Figure 1 As shown, this embodiment provides an optimal evacuation path planning method for cruise ship personnel considering the storage position of life jackets, which includes the following steps:

[0077] Step 1: Construction of cruise ship 3D model;

[0078] In step 1, specifically:

[0079] 1.1) Collect the cruise ship plan, determine key information including cabin structure layout, emergency exit location, main vertical area, floor height, number of stairs, assembly point location, etc., and use 3D modeling software to build a 3D model of the cruise ship;

[0080] 1.2) Determine the evacuation routes for personnel in each main vertical zone, and plan the width and capacity of the evacuation passages from the personnel activity area to the safe exit;

[0081] 1.3) According to the actual passenger capacity of the model, the number of people to be evacuated is set, and the evacuation parameters are constructed according to gender, age, movement speed and initial position distribution;

[0082] 1.4) Combined with the number, load and distribution of passengers inside the cruise ship stipulated in the advanced evacuation analysis in the IMO (MSC / Circ.1533) circular, it is assumed that the cruise ship is in daytime operation, 3 / 4 of the passengers are distributed in public places, and 1 / 4 of the passengers are distributed in the guest rooms. The passengers to be evacuated are imported into the cruise ship 3D model in proportion.

[0083] Step 2: Evacuation scenario construction and simulation;

[0084] In step 2, specifically:

[0085] 2.1) Determine the storage location and proportion of life jackets in the model, taking into account the IMO circular that life jackets should be easily accessible and quickly distributed inside cruise ships, such as near cabin passages or assembly points, and place life jackets at assembly points and inside cabins in different proportions;

[0086] 2.2) Use Pathfinder personnel evacuation simulation software to identify the target safe evacuation area and establish personnel evacuation scenarios when different life jackets are stored in different proportions at the assembly point and in the guest rooms;

[0087] 2.3) Observe the evacuation process during the simulation, analyze the changes in the evacuation time of the cruise ship, the length of the evacuation route and the congestion time parameters during the evacuation process, and provide basic data for subsequent analysis.

[0088] Step 3: Data collection;

[0089] In step 3, specifically:

[0090] 3.1) Collect data on personnel evacuation under different life jacket storage scenarios, and clarify the use of life jackets at assembly points and in guest rooms;

[0091] 3.2) Collect data such as the density of people in the evacuation stairwell, evacuation time, and length of the evacuation route to ensure that the research is based on real data and improve the credibility and practicality of the research.

[0092] Step 4: Classification of personnel evacuation characteristics;

[0093] In step 4, specifically:

[0094] 4.1) The data of evacuation time, evacuation route length and evacuation congestion time of all personnel under different life jacket storage scenarios are sorted into a data set to ensure that the correspondence between scenarios and personnel is complete and the values ​​are correct;

[0095] 4.2) Use the pandas library to load the data and normalize the data to ensure that different feature data have the same scale;

[0096] 4.3) The evacuation time, evacuation route length and evacuation congestion time of each person are selected as feature variables, and the K-means mean clustering algorithm is used to classify the evacuated personnel, divide the crowd categories, and determine the classification of personnel groups under different life jacket storage scenarios;

[0097] In step 4.3), specifically:

[0098] 4.3.1) Define the number of crowd categories into which the evacuees are divided as K, and select j as the index of the crowd category. Each K corresponds to an index j data point as the initial cluster center, expressed as:

[0099] x i =[t i ,d i ,c i ]

[0100]

[0101] Among them, i represents the personnel number, x i represents the three-dimensional feature vector of the i-th person, t i is the evacuation time of the i-th person, d i is the length of the evacuation path of the i-th person, c i is the evacuation congestion time of the i-th personnel, the center point of the j-th category, μ jt represents the mean evacuation time of the jth category, μ jd represents the mean length of the evacuation path of the jth category, μ jc represents the mean evacuation congestion time of the jth category, and 0 represents the randomly selected initial category center point;

[0102] 4.3.2) Assign the data points of evacuees in different life jacket storage scenarios to the nearest cluster center;

[0103] y i =argmin j ||x i -μ j || 2

[0104] Among them, y i Indicates the index of the population category to which the i-th person belongs;

[0105] Euclidean distance is used to calculate the similarity between evacuation time, evacuation path length and evacuation congestion time:

[0106] x i -μ j || 2 =(t i -μ jt ) 2 +(d i -μ jd ) 2 +(c i -μ jc ) 2 ;

[0107] 4.3.3) For each category, calculate the mean of the evacuation time, evacuation path length and evacuation congestion time within each category, and determine the mean as the new cluster center to update the cluster center μ j :

[0108]

[0109] Among them, |C j | represents the number of data points in the jth category, C j represents the set of data points in the jth category;

[0110] Set the objective function E and repeat the above steps until the cluster center reaches the preset maximum number of iterations;

[0111]

[0112] 4.3.4) Calculate the category to which each person belongs;

[0113] 4.4) Further analyze the category characteristics and the distribution of each category in different life jacket storage scenarios, and calculate the average evacuation time, average evacuation path length and average evacuation congestion time of each category;

[0114] 4.5) Clarify the evacuation safety performance of various categories of personnel under various life jacket storage scenarios, determine the current optimal storage scenario for life jackets, provide a classification basis for the path planning optimization model, and help formulate different path planning strategies for different groups of personnel.

[0115] Step 5: Construct and solve the personnel evacuation path optimization planning model.

[0116] In step 5, a path planning model is established, the optimization objective function is clarified, and the constraints are defined before combining the artificial fish swarm algorithm (AFSA) to optimize the model, thereby constructing an optimization planning model for the evacuation path under different life jacket storage scenarios.

[0117] In step 5, specifically:

[0118] 5.1) The optimization goal is to maximize the average evacuation speed and minimize the average evacuation congestion time of different life jacket storage scenarios. The optimization goal is converted into a mathematical expression and defined as the objective function f(x) of the optimization model:

[0119]

[0120] Among them, x is the variable set of path optimization, including path length, evacuation time and congestion time, S is the evacuation scene, J is the personnel category, d sj is the average evacuation path length of the jth group of people in the sth scenario, t sj is the average evacuation time of the jth group of people in the sth scenario, v sj is the average evacuation speed of the jth group of people in the sth scenario, c sj is the average congestion time of the jth group of people in the sth scenario, α and β are the weight coefficients of evacuation speed and congestion time;

[0121] 5.2) Determine the maximum evacuation path length, maximum evacuation time and the density of people in the stairwell of the evacuation passage as constraints, and convert the constraints into mathematical expressions as the constraints of the above optimization model:

[0122] d sj ≤D max t sj ≤T max

[0123]

[0124] Among them, D max is the maximum evacuation path length, T max is the maximum evacuation time, ρ max is the maximum permissible density of people in the evacuation stairway, ρ is the permissible density of people in the evacuation stairway, P represents the number of people going down the stairwell at a certain time, and A represents the area of ​​the stairwell;

[0125] 5.3) Select the artificial fish swarm algorithm (AFSA), set the algorithm parameters and constraints, solve the path planning model, and determine the optimal evacuation path allocation plan; the artificial fish swarm algorithm is an optimization algorithm that simulates the foraging behavior of fish. In the study of the internal evacuation scenario of the cruise ship, it is defined that the crowd will have foraging behavior, clustering behavior and tail-chasing behavior similar to that of fish during the evacuation process. The crowd in the scene searches for the optimal solution through adaptive behavior;

[0126] In step 5.3), specifically:

[0127] 5.3.1) Initialize crowd position z i Randomly distributed in the cruise ship space in the form of vectors, each category of people represents an evacuation path allocation plan;

[0128] z i =(z s1 ,z s2 ,…,z sj )

[0129] Among them, z i represents the initial crowd position coordinate set, z sj represents the two-dimensional coordinates of the s-th path in the j-th category of people;

[0130] 5.3.2) Set the number of evacuation paths in the algorithm parameters to N, and define a three-dimensional matrix (N, S, J) to represent the position of each type of people in the evacuation path, where the position of each type of people is represented by a two-dimensional matrix of S×J;

[0131] 5.3.3) It is stipulated that during the evacuation process of each category of people, evacuation shall be carried out according to the foraging behavior, grouping behavior and tail-chasing behavior of the fish school;

[0132] Preferably, in step 5.3.3), specifically:

[0133] Define foraging behavior: Set the category of people to randomly choose to move to the nearest exit in the initial stage, and calculate the fitness value of the new position after moving:

[0134]

[0135] Define crowd behavior: Under the condition of controlling the density of people in the evacuation stairwell and the flow rate at the exit, find the optimal evacuation path by collaborating with different categories of people around; the category crowd first finds the crowd category with the best fitness within the perception range and judges the crowding degree, where the initial crowding degree is set to 10%-20% of the crowd size, and then chooses to move to the category crowd with the best fitness value:

[0136]

[0137] Definition of tailgating behavior: When an evacuation leader with excellent performance appears during the evacuation process, the individual evacuees in the category group will judge the difference between their own fitness value and the leader, and then drive the category group to optimize the evacuation path through the tailing effect and radiation effect, thereby accelerating the category group to approach the optimal solution of the evacuation path throughout the process:

[0138]

[0139] Among them, Δz is the moving direction vector, visual is the crowd perception range, rand(-1,1) is a random number between -1 and 1, step is the maximum step length, z c is the center position of the crowd within the perception range, z best The position of the crowd category with the best fitness value within the sensing range;

[0140] 5.3.4) Repeat the calculation of the evacuation behavior of the above three categories of people until the maximum number of iterations is reached or the fitness value converges to a certain degree, stop the calculation and output the optimal evacuation path allocation plan with the fitness value, while ensuring the highest evacuation efficiency of the categories of people under different life jacket storage locations;

[0141] 5.4) According to the specific position vectors shown in the optimal evacuation path allocation scheme, the optimal evacuation scenario and crowd category with the objective function of maximizing the average evacuation speed and minimizing the average evacuation congestion time are determined.

[0142] The optimal evacuation path allocation scheme can provide a specific evacuation strategy, clarify the evacuation effects of different life jacket storage positions and categories of people, and provide a reference for subsequent cruise ship structure design and evacuation strategies. Under the determined crowd categories, how should personnel allocate paths to maximize evacuation speed and minimize congestion time. The above results can provide a scientific basis for the internal structure design of cruise ships, the formulation of emergency evacuation plans, evacuation drills, etc., to improve the efficiency and safety of personnel evacuation. Through subsequent research, the storage scene setting of life jackets inside cruise ships and the design of evacuation facilities can be further optimized, and more targeted emergency evacuation plans can be formulated, thereby improving the effectiveness of the overall evacuation system.

[0143] This method focuses on the evacuation time, the length of the evacuation path and the evacuation congestion time. The selection and storage ratio of the cruise ship life jacket storage location are used as scenario variables to determine the categories of people in evacuation conditions and evaluate the evacuation safety of people under different life jacket storage scenarios. The artificial fish swarm algorithm is used to construct an evacuation path optimization model, thereby providing strong support for optimizing the setting of life jacket storage scenarios, improving the evacuation efficiency of different categories of people and formulating more effective safety evacuation strategies.

[0144] The present invention and its embodiments are described schematically above, and the description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by it and designs a structural method and an embodiment similar to the technical solution without creativity without departing from the purpose of the invention, they shall all fall within the protection scope of the present invention.

Claims

1. An optimal evacuation path planning method for cruise ship personnel considering the storage location of life jackets, characterized by: The following steps are involved: Step 1: Construction of cruise ship 3D model; Step 2: Evacuation scenario construction and simulation; Step 3: Data collection; Step 4: Classification of personnel evacuation characteristics; Step 5: Construct and solve the personnel evacuation path optimization planning model.

2. The optimal evacuation path planning method for cruise ship personnel considering the storage location of life jackets according to claim 1 is characterized by: In step 1, specifically: 1.1) Collect the cruise ship plan, determine the cabin structure layout, emergency exit location, main vertical area, floor height, number of stairs, assembly point location information, and use 3D modeling software to build a 3D model of the cruise ship; 1.2) Determine the evacuation routes for personnel in each main vertical zone, and plan the width and capacity of the evacuation passages from the personnel activity area to the safe exit; 1.3) According to the actual passenger capacity of the model, the number of people to be evacuated is set, and the evacuation parameters are constructed according to gender, age, movement speed and initial position distribution; 1.4) Combined with the number, load and distribution of passengers inside the cruise ship stipulated in the advanced evacuation analysis in the IMO circular, it is assumed that the cruise ship is in daytime operation, 3 / 4 of the passengers are distributed in public places, and 1 / 4 of the passengers are distributed in the guest rooms. The passengers to be evacuated are imported into the cruise ship 3D model in proportion.

3. The optimal evacuation path planning method for cruise ship personnel considering the storage location of life jackets according to claim 2 is characterized by: In step 2, specifically: 2.1) Determine the storage location and proportion of life jackets in the model, taking into account the IMO circular that life jackets should be easily accessible and quickly distributed inside cruise ships, and place life jackets at different proportions at the assembly point and inside the guest rooms; 2.2) Use Pathfinder personnel evacuation simulation software to identify the target safe evacuation area and establish personnel evacuation scenarios when different life jackets are stored in different proportions at the assembly point and in the guest rooms; 2.3) Observe the evacuation process during the simulation, analyze the changes in the evacuation time of the cruise ship, the length of the evacuation route and the congestion time parameters during the evacuation process, and provide basic data for subsequent analysis.

4. The method for planning the optimal evacuation path for cruise ship personnel considering the storage location of life jackets according to claim 3, characterized in that: In step 3, specifically: 3.1) Collect data on personnel evacuation under different life jacket storage scenarios, and clarify the use of life jackets at assembly points and in guest rooms; 3.2) Collect data on the density of people in the evacuation stairwells, evacuation time and the length of the evacuation route to ensure that the research is based on real data.

5. The method for planning the optimal evacuation path for cruise ship personnel considering the storage position of life jackets according to claim 4, characterized in that: In step 4, specifically: 4.1) The data of evacuation time, evacuation route length and evacuation congestion time of all personnel under different life jacket storage scenarios are sorted into a data set to ensure that the correspondence between scenarios and personnel is complete and the values ​​are correct; 4.2) Use the pandas library to load the data and normalize the data to ensure that different feature data have the same scale; 4.3) The evacuation time, evacuation route length and evacuation congestion time of each person are selected as feature variables, and the K-means mean clustering algorithm is used to classify the evacuated personnel, divide the crowd categories, and determine the classification of personnel groups under different life jacket storage scenarios; 4.4) Further analyze the category characteristics and the distribution of each category in different life jacket storage scenarios, and calculate the average evacuation time, average evacuation path length and average evacuation congestion time of each category; 4.5) Clarify the evacuation safety performance of various categories of personnel under various life jacket storage scenarios, determine the current optimal storage scenario for life jackets, provide a classification basis for the path planning optimization model, and help formulate different path planning strategies for different groups of personnel.

6. The method for planning the optimal evacuation path for cruise ship personnel considering the storage location of life jackets according to claim 5, characterized in that: In step 4.3), specifically: 4.3.1) Define the number of crowd categories into which the evacuees are divided as K, and select j as the index of the crowd category. Each K corresponds to an index j data point as the initial cluster center, expressed as: x i =[t i ,d i ,c i ] Among them, i represents the personnel number, x i represents the three-dimensional feature vector of the i-th person, t i is the evacuation time of the i-th person, d i is the length of the evacuation path of the i-th person, c i is the evacuation congestion time of the i-th person, μ j represents the center point of the jth category, μ jt represents the mean evacuation time of the jth category, μ jd represents the mean length of the evacuation path of the jth category, μ jc represents the mean evacuation congestion time of the jth category, and 0 represents the randomly selected initial category center point; 4.3.2) Assign the data points of evacuees in different life jacket storage scenarios to the nearest cluster center; y i =argmin j ||x i -m j || 2 Among them, y i Indicates the index of the population category to which the i-th person belongs; Euclidean distance is used to calculate the similarity between evacuation time, evacuation path length and evacuation congestion time: ||x i -m j || 2 =(t i -m jt ) 2 +(d i -m jd ) 2 +(c i -m jc ) 2 ; 4.3.3) For each category, calculate the mean of the evacuation time, evacuation path length and evacuation congestion time within each category, and determine the mean as the new cluster center to update the cluster center μ j : Among them, |C j | represents the number of data points in the jth category, C j represents the set of data points in the jth category; Set the objective function E and repeat the above steps until the cluster center reaches the preset maximum number of iterations; 4.3.4) Calculate the category to which each person belongs.

7. The method for planning the optimal evacuation path for cruise ship personnel considering the storage position of life jackets according to claim 6, characterized in that: In step 5, a path planning model is established, the optimization objective function is clarified, and after defining the constraints, the model is optimized in combination with the artificial fish swarm algorithm AFSA to construct an optimization planning model for the evacuation path of personnel under different life jacket storage scenarios.

8. The method for planning the optimal evacuation path for cruise ship personnel considering the storage position of life jackets according to claim 7, characterized in that: In step 5, specifically: 5.1) The optimization goal is to maximize the average evacuation speed and minimize the average evacuation congestion time of different life jacket storage scenarios. The optimization goal is converted into a mathematical expression and defined as the objective function f(x) of the optimization model: Among them, x is the variable set of path optimization, including path length, evacuation time and congestion time, S is the number of evacuation scenarios, J is the number of crowd categories, d sj is the average evacuation path length of the jth group of people in the sth scenario, t sj is the average evacuation time of the jth group of people in the sth scenario, v sj is the average evacuation speed of the jth group of people in the sth scenario, c sj is the average congestion time of the jth group of people in the sth scenario, α and β are the weight coefficients of evacuation speed and congestion time; 5.2) Determine the maximum evacuation path length, maximum evacuation time and the density of people in the stairwell of the evacuation passage as constraints, and convert the constraints into mathematical expressions as the constraints of the above optimization model: d sj ≤D max t sj ≤T max Among them, D max is the maximum evacuation path length, T max is the maximum evacuation time, P max is the maximum permissible density of people in the evacuation stairway, ρ is the permissible density of people in the evacuation stairway, P represents the number of people going down the stairwell at a certain time, and A represents the area of ​​the stairwell; 5.3) Select the artificial fish swarm algorithm AFSA, set the algorithm parameters and constraints, solve the path planning model, and determine the optimal evacuation path allocation plan; 5.4) According to the specific position vectors shown in the optimal evacuation path allocation scheme, the optimal evacuation scenario and crowd category with the objective function of maximizing the average evacuation speed and minimizing the average evacuation congestion time are determined.

9. The method for planning the optimal evacuation path for cruise ship personnel considering the storage position of life jackets according to claim 8, characterized in that: In step 5.3), specifically: 5.3.1) Initialize crowd position z i Randomly distributed in the cruise ship space in the form of vectors, each category of people represents an evacuation path allocation plan; With i =(from s1 ,With s2 ,…,With sj ) Among them, z i represents the initial crowd position coordinate set, z sj represents the two-dimensional coordinates of the s-th path in the j-th category of people; 5.3.2) Set the number of evacuation paths in the algorithm parameters to N, and define a three-dimensional matrix (N, S, J) to represent the position of each type of people in the evacuation path, where the position of each type of people is represented by a two-dimensional matrix of S×J; 5.3.3) It is stipulated that during the evacuation process of each category of people, evacuation shall be carried out according to the foraging behavior, grouping behavior and tail-chasing behavior of the fish school; 5.3.4) Repeat the calculation of the evacuation behavior of the above three categories of people until the maximum number of iterations is reached or the fitness value converges to a certain degree. After stopping the calculation, output the optimal evacuation path allocation plan with the fitness value, and at the same time ensure that the evacuation efficiency of the categories of people under different life jacket storage locations is the highest.

10. The method for planning the optimal evacuation path for cruise ship personnel considering the storage position of life jackets according to claim 9, characterized in that: In step 5.3.3), specifically: Define foraging behavior: Set the category of people to randomly choose to move to the nearest exit in the initial stage, and calculate the fitness value of the new position after moving: Define crowd behavior: Under the condition of controlling the density of people in the evacuation stairwell and the flow rate at the exit, find the optimal evacuation path by collaborating with different categories of people around; the category crowd first finds the crowd category with the best fitness within the perception range and judges the crowding degree, where the initial crowding degree is set to 10%-20% of the crowd size, and then chooses to move to the category crowd with the best fitness value: Definition of tailgating behavior: When an evacuation leader with excellent performance appears during the evacuation process, the individual evacuees in the category group will judge the difference between their own fitness value and the leader, and then drive the category group to optimize the evacuation path through the tailing effect and radiation effect, thereby accelerating the category group to approach the optimal solution of the evacuation path throughout the process: where Δz is the movement direction vector, visual is the crowd perception range, rand(−1, 1) is a random number between -1 and 1, step is the maximum step size, z c is the central position of the crowd within the perception range, and z best is the position of the crowd category with the optimal fitness value within the perception range.

Images