A dynamic evacuation guidance system for cruise ship passengers

Abstract

The present invention belongs to the field of cruise ship accident evacuation guidance and discloses a cruise ship passenger dynamic evacuation guidance system. The system comprises three modules: monitoring and alarming, route planning, and dynamic identification. When a disaster occurs on a cruise ship or an abnormal ship posture triggers an alarm, the monitoring and alarming module collects relevant data and transmits it in real time. The route planning module reconstructs the route network and assigns path weights based on this data, using the Dijkstra algorithm with an improved search strategy to search for the global optimal route. The dynamic identification module maps route data to the directions indicated by the signs. The dynamic identification layout is optimized based on passenger perception during system deployment. During the evacuation process, the monitoring and alarming module monitors the crowd gathering situation on each road section in real time, and the relevant data is fed back to the route planning module for a new round of search. The present invention can plan the optimal route in real time based on the cruise ship accident scenario and dynamically adjust the directions indicated by the evacuation signs, thereby achieving dynamic guidance for crowd evacuation.

Description

Technical Field

[0001] The present invention belongs to the field of cruise ship accident evacuation guidance, and in particular relates to a cruise ship passenger dynamic evacuation guidance system. Background Art

[0002] Cruise ship accidents are often sudden, complex, and high-risk. These scenarios involve multiple factors, including structural damage, fire, explosion, and harsh environments, posing significant challenges to emergency evacuation and rescue efforts. In this context, how to quickly and effectively organize orderly evacuations in emergency situations has become a crucial issue for ensuring both life safety and public security.

[0003] Due to passengers' limited familiarity with the complex environment of cruise ships, emergency evacuation signs, as crucial information carriers during emergency evacuations, are often used to indicate the direction and location of destinations such as emergency exits. Their layout directly impacts whether evacuating passengers can quickly and accurately access route information. On the one hand, existing signs are often obscured or difficult to detect, severely impacting evacuation efficiency. Therefore, scientifically and rationally optimizing the layout of evacuation signs has important theoretical and practical application value. On the other hand, traditional sign systems rely on pre-planned escape routes to set directions. In the event of congestion or fire, these fixed directions no longer meet the needs of evacuation. Cruise ships also need to consider adverse sea conditions caused by extreme weather, collisions with offshore obstacles, or groundings in shallow waters, which can cause the ship to tilt. Therefore, enabling sign systems to dynamically guide evacuation in complex accident scenarios is crucial for improving evacuation efficiency and ensuring personnel safety. Summary of the Invention

[0004] In response to the existing technical defects in the background technology, the present invention proposes a dynamic evacuation guidance system for cruise ship passengers. When a disaster occurs in the cruise ship or the hull posture is abnormal, the alarm is triggered and the relevant data collected by the monitoring alarm module will be transmitted in real time. After receiving the data, the path planning module will use this as a basis to reconstruct the path network and assign path weights. After globally searching the new path network using the Dijkstra algorithm with an improved search strategy, the optimal route from each node to the exit will be obtained. The output route data will be mapped to the indication direction of the identification system by the dynamic identification module. The layout of the dynamic identification has been optimized by considering the pedestrian's perception behavior of the identification when the system is deployed. During the evacuation process, the monitoring alarm module will monitor the crowd gathering situation on each path in real time. The relevant data including crowd density will be sent to the path planning module again for a new round of search, thereby realizing dynamic guidance of crowd evacuation.

[0005] The technical solution adopted by the present invention is:

[0006] A dynamic evacuation guidance system for cruise ship passengers, characterized in that the guidance system includes a monitoring and alarm module, a path planning module, and a dynamic identification module, wherein the monitoring and alarm module dominates the perception layer of the system, the path planning module dominates the decision layer of the system, and the dynamic identification module dominates the application layer of the system;

[0007] The monitoring and alarm module is used for monitoring data and alarming;

[0008] The path planning module is used to reconstruct the path network based on the monitored data, assign path weights, and use the Dijkstra algorithm with an improved search strategy to search for the global optimal route;

[0009] The dynamic identification module is used to adjust the sign indication direction in real time according to the global optimal route based on the optimized sign layout.

[0010] Preferably, the monitoring and alarm module includes a fire alarm system, an inertial measurement unit and a video monitoring device, specifically as follows:

[0011] The fire alarm system is composed of sensors and alarms, wherein the sensors include temperature sensors, carbon monoxide detectors and smoke detectors, which are used to monitor the temperature, CO volume fraction and visibility of the road section;

[0012] The inertial measurement unit includes a speedometer, a gyroscope and a magnetometer, which are used to measure the hull attitude and tilt angle including roll and pitch;

[0013] When the temperature, CO2 volume fraction or hull tilt angle of the monitored section exceeds the set threshold, or the visibility is lower than the set threshold, the alarm will be activated. At the same time, the relevant data collected by the monitoring alarm module will be transmitted in real time;

[0014] The video surveillance equipment includes cameras deployed on various road sections, which are combined with wireless sensor networks and target detection and tracking technology to achieve the statistics and positioning of the number of pedestrians on each road section and the acquisition of crowd density data.

[0015] Preferably, the specific construction process of the path network in the path planning module is:

[0016] Treat each marker in the space as a node, connect each pair of adjacent nodes to form several paths, simplify markers that are close to each other into the same node, construct paths blocked by obstacles into broken lines, and when multiple paths intersect, only the shortest path is retained;

[0017] The path network is further simplified. Based on the fact that the bow and stern directions and the port and starboard directions represent the longitudinal and transverse directions of the ship, respectively, the paths along the two directions are called longitudinal paths and transverse paths, respectively. The longitudinal paths and transverse paths are called axial paths. The paths with smaller angles with the transverse and longitudinal axes in the path network are called axial paths, and the paths opposite to the axial paths are called oblique paths. When simplifying the path network, the path type is determined according to the following formula:

[0018]

[0019] in, Indicates that there are three paths: vertical, horizontal and diagonal. is the angle between the path and the bow and stern axis, is the threshold of the angle between the path and the horizontal and vertical axes, and its value range is .

[0020] Preferably, the path planning module uses the Dijkstra algorithm to obtain the global optimal route, as follows:

[0021] (1) Improve the search strategy of Dijkstra algorithm

[0022] Construct a virtual node connected to all exit nodes, set the weight of each path adjacent to the virtual node to zero, and use the virtual node as the starting node to search all nodes in the path network. Before each search, reconstruct the path network and assign path weights.

[0023] (2) Reconstructing the path network

[0024] When a hazard source is sensed or the temperature, visibility, or CO volume fraction on any road section deviates from the set monitoring threshold, the node corresponding to the road section in the path network is determined to be a dangerous node and the corresponding path is determined to be a dangerous path. All dangerous nodes and dangerous paths are eliminated to complete the reconstruction of the path network.

[0025] (3) Assigning path weights

[0026] The travel time of each road section is used as the weight of its corresponding path in the path network. The specific formula is:

[0027]

[0028] The travel time of the road section Travel time of axial routes by route type and the travel time of the diagonal path ;

[0029] The segment travel time of the axial path is calculated using the following formula:

[0030]

[0031] in, is the distance between the two ends of the road section, is the traffic speed of the road section;

[0032] The diagonal path is decomposed into two axial paths based on its projection on the longitudinal and transverse axes. The travel time of the longitudinal and transverse paths is then calculated and summed to obtain the total travel time. Finally, the total travel time is corrected to obtain the section travel time of the diagonal path. The specific formula is as follows:

[0033]

[0034]

[0035] in, is a correction factor used to avoid the path being "stretched".

[0036] Preferably, the calculation formula for the road section speed is:

[0037]

[0038] (1) When the ship is in an upright state, the travel time of a road section depends on its congestion level. The crowd density on the road section is an important indicator for measuring the congestion level of the road section.

[0039] The traffic speed of a road section under congestion is a piecewise function of the crowd density on the road section. The specific formula is:

[0040]

[0041]

[0042] in, is the speed before tilting, is the crowd density on the road section, is the total number of pedestrians gathered on the path, The available area of the path is the area of the area that pedestrians are actually allowed to reach after removing the space occupied by dynamic obstacles.

[0043] (2) When the hull tilts, the tilt state of the road section is divided into longitudinal tilt, transverse tilt and combined tilt according to the direction of the rotation axis. The speed attenuation coefficient is the ratio of the speed after tilt to the speed before tilt. The speed of the longitudinal and transverse tilt sections under congestion is calculated by the following formula:

[0044]

[0045]

[0046] in, is the speed of the longitudinal section, is the velocity attenuation coefficient of the longitudinal section, is the longitudinal inclination angle of the road section, is the speed of the inclined road section after tilting, is the velocity attenuation coefficient of the banked section, is the inclination angle of the road section;

[0047] The bidirectional inclination angles caused by compound inclination further increase the difficulty of traffic on the road section. Assuming that the effects of longitudinal inclination and transverse inclination on traffic speed are not independent of each other when the road section is compound inclination, the traffic speed on the compound inclination road section can be expressed as:

[0048]

[0049] in, is the traffic speed on the compound inclined road section, To obtain the speed of the longitudinal section and the speed of the inclined section The smaller value of is a penalty term, reflecting the further restriction of the passage speed caused by the combined effect of pitch and roll. The maximum permissible value of the total tilt angle is the sum of the absolute values of the pitch angle and the roll angle. The path will be difficult to pass. is the penalty coefficient, and its value range is , which indicates the degree of influence of the total inclination angle on the traffic speed. The larger the value, the more significant the penalty.

[0050] Preferably, the optimized layout of the signs in the dynamic sign module is based on an evacuation sign layout optimization method, and the specific process of the optimization method is as follows:

[0051] (1) The pedestrian action space and the sign layout space are set on the same rectangular coordinate plane, and the plane is discretized into a number of The walkable area and placement area are composed of rectangular units;

[0052] (2) Using the multi-exit spatial guidance demand assessment algorithm, the difference in the distance from the exit to the pedestrian after each step is measured by the difference in the distance between the pedestrian and the exit when there is a sign and when there is no sign guidance. The distribution of guidance demand in the walkable area is obtained. The calculation formula is:

[0053]

[0054] in, The guidance needs of any unit in the walkable area, is the pedestrian’s current unit, is the next unit in the optimal route from the current unit to its nearest exit, is the shortened value of the distance from the current unit to the target exit after the pedestrian moves from the current unit to the next unit when the sign is in effect. is the average shortening value when the pedestrian moves to each adjacent unit; is the exit closest to the current unit, For unit To the exit The shortest distance, For unit To the exit The shortest distance; is any neighboring unit of the current unit, For unit The set of neighboring cells, is the shortened value of the distance from the current unit to the target exit after the pedestrian moves from the current unit to the adjacent unit when there is no sign. For unit The number of neighboring units; is the nearest exit to the adjacent unit, For unit To the exit The shortest distance, For unit To the exit The shortest distance;

[0055] (3) Use genetic algorithm to solve the logo optimization layout, the formula is as follows:

[0056]

[0057]

[0058]

[0059] in, is the average matching degree between guidance demand and perception probability in the walkable area, that is, the guidance demand of all units and the corresponding perception probability The sum of products and the number of labels in the space The ratio, is the set of all units within the walkable area; is the actual size of the logo, is the horizontal distance from the pedestrian to the sign, is the horizontal observation angle, The height difference between the logo and the human eye is is the threshold for identifying the size of retinal imaging; is the set of all cells in the placeable area, is the set of all identifiers in the space, For any unit where the mark is located, is the angle of the marking direction, is the allowed number of markers in the space, To identify the number of units within the range of the system, it is determined by marking the legible area and the obstruction of obstacles;

[0060] The diameter of the label readable area Expressed as ,in is the readability factor; the judgment formula for identifying the readability area is:

[0061]

[0062] in, To identify the coordinates of the location, The radius of the legible area of the logo;

[0063] The ratio of the visible length of the logo to its actual length The calculation method is:

[0064]

[0065] in, is the vector pointing to the pedestrian. It is a vector pointing to the direction of the self.

[0066] Preferably, the dynamic identification module adjusts the identification direction in real time, and the specific process is as follows:

[0067] The global optimal route planned by the path planning module is used as the basis for the identification direction indication. The output result of the Dijkstra algorithm of the path planning module is a set of all nodes on the optimal route from each node to the exit. All nodes in the set share the same target exit. The next node in the direction of each node towards the exit is the guidance destination identified at the node.

[0068] Compared with the prior art, the present invention proposes a dynamic evacuation guidance system for cruise ship passengers, which has the following beneficial effects:

[0069] The present invention integrates three modules: monitoring and alarm, path planning, and dynamic identification, corresponding to the perception layer, decision layer, and application layer, respectively. These modules are supplemented by data storage and network communication modules to form a complete system. The monitoring and alarm module utilizes multi-sensor fusion technology. Temperature sensors, carbon monoxide detectors, smoke detectors, and an inertial measurement unit (speedometer, gyroscope, and magnetometer) collaboratively monitor environmental parameters (temperature, CO concentration, and visibility) and the ship's attitude (roll and pitch angles). When a monitoring parameter exceeds a threshold, an alarm activates and transmits data in real time. Video surveillance equipment, combined with target detection and tracking technology, enables pedestrian counting, location, and crowd density acquisition, providing the system with multi-dimensional dynamic perception data. The path planning module introduces virtual nodes and sets the weights of paths adjacent to the virtual nodes and exits to zero. Using the virtual node as the starting node, a single traversal can be performed to obtain the optimal route from all nodes to the exit, improving search efficiency. Sensor monitoring thresholds are also set to eliminate dangerous nodes and paths, enabling dynamic reconstruction of the path network. Path weights are assigned based on travel time, enabling path weight calculation in complex scenarios. The dynamic signage module discretizes the space into rectangular units and obtains the guidance demand distribution in the walkable area through a multi-exit spatial guidance demand assessment algorithm. With the matching degree between guidance demand and perceived probability as the goal, layout optimization is performed based on a genetic algorithm to achieve "node-multi-signage" collaborative guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 This is an architectural diagram of a cruise ship passenger dynamic evacuation guidance system proposed by the present invention;

[0071] Figure 2 This is a flow chart of the multi-exit space guidance demand assessment algorithm proposed by the present invention;

[0072] Figure 3 Schematic diagram of the path network of the path planning module in an embodiment of the present invention. DETAILED DESCRIPTION

[0073] The following will be combined with the drawings in the embodiments of this application to further clearly and completely describe the technical solutions in the embodiments of this application. It should be noted that the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making any creative work shall fall within the scope of protection of this application.

[0074] In order to make the invention objectives, technical solutions and advantages of this application clearer, the embodiments of this application are further described in detail in conjunction with the drawings in the specification: In order to more clearly understand the above-mentioned objectives, features and advantages of the present invention, the advantages of the present invention will be further illustrated by comparing the embodiments in conjunction with the drawings and specific implementation methods.

[0075] The present invention proposes a cruise passenger dynamic evacuation guidance system, the flow chart of which is as follows: Figure 1 The steps of this method are described in detail as shown below:

[0076] The guidance system includes a monitoring and alarm module, a path planning module, a dynamic identification module and a network communication module, wherein the monitoring and alarm module leads the perception layer of the system, the path planning module leads the decision layer of the system, and the dynamic identification module leads the application layer of the system;

[0077] The monitoring and alarm module includes a fire alarm system, an inertial measurement unit, and video monitoring equipment for monitoring data and alarms;

[0078] Specifically, the monitoring and alarm module includes a fire alarm system, an inertial measurement unit, and a video monitoring device, as follows:

[0079] The fire alarm system is composed of sensors and alarms, wherein the sensors include temperature sensors, carbon monoxide detectors and smoke detectors, which are used to monitor the temperature, CO volume fraction and visibility of the road section;

[0080] The inertial measurement unit includes a speedometer, a gyroscope and a magnetometer, which are used to measure the hull attitude and tilt angle including roll and pitch;

[0081] When the temperature, CO2 volume fraction, or hull tilt angle of the monitored section exceeds the set threshold, or the visibility falls below the set threshold, the alarm will be activated. At the same time, the relevant data collected by the monitoring alarm module will be transmitted in real time. During actual deployment, the sensor sets the monitoring thresholds for temperature, visibility, and CO2 volume fraction. The values of each monitoring threshold are based on the critical values of the indicators that ensure normal operation of personnel.

[0082] The video surveillance equipment includes cameras deployed on various road sections, combined with wireless sensor networks and target detection and tracking technology, to achieve the statistics and positioning of the number of pedestrians on each path and the acquisition of crowd density data. The crowd density and other data obtained on this basis will be provided to the system for re-route planning.

[0083] The path planning module is used to reconstruct the path network based on the monitored data, assign path weights, and use the Dijkstra algorithm with an improved search strategy to search for the global optimal route;

[0084] Specifically, the specific construction process of the path network in the path planning module is as follows:

[0085] Treat each marker in the space as a node, connect each pair of adjacent nodes to form several paths, simplify markers that are close to each other into the same node, construct paths blocked by obstacles into broken lines, and when multiple paths intersect, only the shortest path is retained;

[0086] The path network is further simplified. Based on the fact that the bow and stern directions and the port and starboard directions represent the longitudinal and transverse directions of the ship, respectively, the paths along the two directions are called longitudinal paths and transverse paths, respectively. The longitudinal paths and transverse paths are called axial paths. The paths with smaller angles with the transverse and longitudinal axes in the path network are called axial paths, and the paths opposite to the axial paths are called oblique paths. The formula for determining the path type when simplifying the path network is:

[0087]

[0088] in, Indicates that there are three paths: vertical, horizontal and diagonal. is the angle between the path and the bow and stern axis, is the threshold of the angle between the path and the horizontal and vertical axes, and its value range is .

[0089] Specifically, the path planning module uses the Dijkstra algorithm to obtain the global optimal route, as follows:

[0090] (1) Improve the search strategy of Dijkstra algorithm

[0091] Construct a virtual node connected to all exit nodes, set the weight of each path adjacent to the virtual node to zero, and use the virtual node as the starting node to search all nodes in the path network. Before each search, reconstruct the path network and assign path weights.

[0092] (2) Reconstructing the path network

[0093] When a hazard source is sensed or the temperature, visibility, or CO volume fraction on any road section deviates from the set monitoring threshold, the node corresponding to the road section in the path network is determined to be a dangerous node and the corresponding path is determined to be a dangerous path. All dangerous nodes and dangerous paths are eliminated to complete the reconstruction of the path network.

[0094] (3) Assigning path weights

[0095] The travel time of each road section is used as the weight of its corresponding path in the path network. The specific formula is:

[0096]

[0097] The travel time of the road section in the formula According to the route type, the travel time of the axial route can be divided into and the travel time of the diagonal path ;

[0098] The segment travel time of the axial path is calculated using the following formula:

[0099]

[0100] in, is the distance between the two ends of the road section, is the traffic speed of the road section;

[0101] The diagonal path is decomposed into two axial paths based on its projection on the longitudinal and transverse axes. The travel time of the longitudinal and transverse paths is then calculated and summed to obtain the total travel time. Finally, the total travel time is corrected to obtain the section travel time of the diagonal path. The specific formula is as follows:

[0102]

[0103] in, is a correction factor used to avoid the path being "stretched".

[0104] Specifically, the calculation formula for the road section speed is:

[0105]

[0106] (1) When the ship is in an upright state, the travel time of a road section depends on its congestion level. The crowd density on the road section is an important indicator for measuring the congestion level of the road section.

[0107] The traffic speed of a road section under congestion is a piecewise function of the crowd density on the road section. The specific formula is:

[0108]

[0109]

[0110] in, is the speed before tilting, is the crowd density on the road section, is the total number of pedestrians gathered on the path, The available area of the path is the area of the area that pedestrians are actually allowed to reach after removing the space occupied by dynamic obstacles.

[0111] (2) When the hull tilts, the tilt state of the road section is divided into longitudinal tilt, transverse tilt and combined tilt according to the direction of the rotation axis. The speed attenuation coefficient is the ratio of the speed after tilt to the speed before tilt. The speed of the longitudinal and transverse tilt sections under congestion is calculated by the following formula:

[0112]

[0113]

[0114] in, is the speed of the longitudinal section, is the velocity attenuation coefficient of the longitudinal section, is the longitudinal inclination angle of the road section, is the speed of the inclined road section after tilting, is the velocity attenuation coefficient of the banked section, is the inclination angle of the road section;

[0115] The bidirectional inclination angles caused by compound inclination further increase the difficulty of traffic on the road section. Assuming that the effects of longitudinal inclination and transverse inclination on traffic speed are not independent of each other when the road section is compound inclination, the traffic speed on the compound inclination road section can be expressed as:

[0116]

[0117] in, is the traffic speed on the compound inclined road section, To obtain the speed of the longitudinal section and the speed of the inclined section The smaller value of is a penalty term, reflecting the further restriction of the passage speed caused by the combined effect of pitch and roll. The maximum permissible value of the total tilt angle is the sum of the absolute values of the pitch angle and the roll angle. The path will be difficult to pass. is the penalty coefficient, and its value range is , which indicates the degree of influence of the total inclination angle on the traffic speed. The larger the value, the more significant the penalty.

[0118] Specifically, the optimized layout of the signs in the dynamic sign module is based on an evacuation sign layout optimization method, and the specific process of the optimization method is as follows:

[0119] (1) The pedestrian action space and the sign layout space are considered to be the same rectangular coordinate plane, and the plane is discretized into several The walkable area and placement area are composed of rectangular units. Assuming that the actual position of the sign or pedestrian is the center point of the unit, each unit represents both the potential position of the sign in the placement area and the moving target of the pedestrian in the walkable area. In addition, it is used to represent obstacles that prevent pedestrian movement or areas where signs cannot be placed.

[0120] (2) When there are no obstacles around the pedestrian, the next moving target is the adjacent 8 units. Assuming that the pedestrian is not familiar with the space he is in, he will randomly choose a direction to move with the same probability when there is no sign. When there is a sign, the pedestrian will be guided to the adjacent unit closest to the target exit.

[0121] If the exits closest to the pedestrian's current unit and the adjacent unit are different, the pedestrian's target exit is the exit closest to the next unit when there is no sign to guide them, while the pedestrian's target exit is the exit closest to the unit they are in when there is a sign to guide them;

[0122] The multi-exit spatial guidance demand assessment algorithm is used to obtain the distribution of guidance demand in the walkable area. The algorithm uses the difference in the distance shortened from the exit after each step of the pedestrian with and without sign guidance to measure the difference in guidance demand at different locations in the walkable area. The calculation formula is:

[0123]

[0124] in, The guidance needs of any unit in the walkable area, is the pedestrian’s current unit, is the next unit in the optimal route from the current unit to its nearest exit, is the shortened value of the distance from the current unit to the target exit after the pedestrian moves from the current unit to the next unit when the sign is in effect. is the average shortening value when the pedestrian moves to each adjacent unit; is the exit closest to the current unit, For unit To the exit The shortest distance, For unit To the exit The shortest distance; is any neighboring unit of the current unit, For unit The set of neighboring cells, is the shortened value of the distance from the current unit to the target exit after the pedestrian moves from the current unit to the adjacent unit when there is no sign. For unit The number of neighboring units; is the nearest exit to the adjacent unit, For unit To the exit The shortest distance, For unit To the exit The shortest distance;

[0125] (3) Use genetic algorithm to solve the logo optimization layout, the formula is as follows:

[0126]

[0127]

[0128]

[0129] in, is the average matching degree between guidance demand and perception probability in the walkable area, that is, the guidance demand of all units and the corresponding perception probability The sum of products and the number of labels in the space The ratio, is the set of all units within the walkable area; is the actual size of the logo, is the horizontal distance from the pedestrian to the sign, is the horizontal observation angle, To mark the height difference between the logo and the human eye, is the threshold for identifying the size of retinal imaging; is the set of all cells in the placeable area, is the set of all identifiers in the space, For any unit where the mark is located, is the angle of the marking direction, is the allowed number of markers in the space, To identify the number of units within the range of the system, it is determined by marking the legible area and the obstruction of obstacles;

[0130] The diameter of the label readable area Expressed as ,in is the readability factor; the judgment formula for identifying the readability area is:

[0131]

[0132] in, To identify the coordinates of the location, The radius of the legible area of the logo;

[0133] The ratio of the visible length of the logo to its actual length The calculation method is:

[0134]

[0135] in, is the vector pointing to the pedestrian. It is a vector pointing to the direction of the self.

[0136] Specifically, the dynamic identification module adjusts the identification direction in real time, and the specific process is as follows:

[0137] The global optimal route planned by the path planning module is used as the basis for the direction indication. The output result of the Dijkstra algorithm of the path planning module is the set of all nodes on the optimal route from each node to the exit. All nodes in the set share the same target exit. Each node is guided to the next node in the direction of the exit, which is the destination marked at the node.

[0138] Since the paths are modeled by simplifying the nearby signs into the same node, the positional relationship between the node and the next node needs to be mapped to the directions indicated by different signs at the same time.

[0139] The change of the direction indicated by the dynamic sign depends on the transformation of the direction arrow, and the transformation of the direction arrow is achieved by controlling the light emitting diode.

[0140] The system further comprises a data storage module composed of a database and a network communication module composed of Ethernet and a wireless sensor network, and different modules are interconnected in series.

[0141] The specific experiments of the present invention are described in detail with reference to the technical solutions and the accompanying drawings as follows:

[0142] This example uses an evacuation sign layout optimization method to determine the distribution of signs on a cruise ship. The optimization method is implemented in the order of target space modeling, guidance demand assessment, key parameter setting, and optimization model solution.

[0143] When modeling the target space, to ensure that each unit can accommodate a pedestrian or sign, its single-side size is close to the width of the human body or the length of the sign. Taking into account the average shoulder width of an adult and the length of the sign set by the user, the plane representing the action space and layout space is discretized into a certain unit size.

[0144] When guiding demand assessment, the nearest exit of the unit and the distance from the unit to the exit are searched in combination with the Dijkstra algorithm, such as Figure 2 As shown;

[0145] The Dijkstra algorithm is used to take the end point as the starting node and all possible starting points of pedestrians as the target nodes. The Dijkstra algorithm has no early exit mechanism and must traverse all nodes. The search for the "target" node is completed through all nodes in one traversal.

[0146] Specifically, any cell representing an exit is taken as the starting node, and all cells adjacent to this cell in the walkable area (no more than ) together form the first layer of neighboring nodes, and all cells adjacent to these cells together form the second layer of neighboring nodes, and so on. All cells in the walkable area eventually form a complete weighted undirected graph;

[0147] After multiple cells representing the same exit have traversed the weighted undirected graph starting with them, for each cell in the walkable area, the distance from it to different cells representing the same exit is compared to obtain the cell that actually represents the exit location among the different cells.

[0148] The number of walkable areas that need to be traversed is equal to the number of units that make up an exit. The total number of units that make up all exits is the number of times the algorithm needs to be executed. Furthermore, the virtual node connected to all exits (units) can be used as the starting node to complete the search of the walkable area for each exit through one traversal.

[0149] When setting key parameters, if multiple signs cannot be placed in the same unit at the same time, the sign capacity in the space is the total number of units in the placement area, and the allowed number of signs is an empirical estimate based on the size of the space;

[0150] The text height of the sign is an empirical value based on the set sign size; the eye level height is close to the average height of wheelchair users to meet the guidance needs of the vast majority of pedestrians.

[0151] When solving the optimization model, the inherent uncertainty of the genetic algorithm will lead to the randomness of the optimal result of the objective function. Therefore, it is necessary to perform multiple model solutions to select the logo layout with the largest average matching degree.

[0152] Any layout of markers in the target space is a feasible solution to the optimization model. Each marker in the layout has 、 and The three types of variables correspond to them. According to the allowed number of labels in the space, the total number of variables in any feasible solution is ;

[0153] The components of the feasible solution are as follows:

[0154]

[0155] in, is a decision variable with a value of 0 or 1, used to determine whether the logo is adopted in the layout. The index of the unit to be identified, its value should not be greater than the total number of units in the placement area ;

[0156] In order to improve the efficiency of model solving, The original value range is limited to integers; considering the pedestrians’ path-finding habits, The value of is restricted to the set among;

[0157] The solution process requires obtaining the effective range of the sign system under any layout. The sign's recognizable range is related to the degree to which obstacles block pedestrians' line of sight. If the sign is considered a light source, the actual recognizable range is the area within the legible area where its light reaches.

[0158] Each unit in the walkable area can only be occupied by one pedestrian at a time. Theoretically, the pedestrian capacity in the space is the total number of units in the walkable area. ;

[0159] After solving the optimized layout of the signs, the path network in the path planning module is modeled. The construction method is as follows:

[0160] Treat each identifier in the space as a node, connect each pair of adjacent nodes to form several edges, and the whole composed of all nodes and edges is the path network of the space;

[0161] Some signs that are close to each other are simplified to the same node during modeling, and the path connecting the two subspaces needs to be constructed as a broken line due to obstacles.

[0162] The path network is further simplified. Considering that the bow and stern directions and the port and starboard directions represent the longitudinal and transverse directions of the hull, respectively, the paths along the above two directions can be called longitudinal paths and transverse paths, respectively (collectively referred to as axial paths). The paths in the path network with smaller angles with the longitudinal and transverse axes can be approximately regarded as axial paths.

[0163] Paths 8 to 9, 14 to 15, and 3 are considered longitudinal paths, while path 6 is considered a transverse path. Paths 7 and 16 are partially considered longitudinal paths and partially considered transverse paths. Figure 3 shown.

[0164] The Dijkstra algorithm is used as the underlying algorithm of the guidance system path planning module. In order to realize the search of the "target" node by all nodes in one traversal, the node where each exit is located is used as the starting node, and all nodes in the path network are searched to obtain the optimal route from each node to the corresponding exit.

[0165] In order to further improve the search efficiency, a virtual node is constructed that is connected to all exit nodes, and the weight of each path adjacent to the virtual node is set to zero. Since the cost from the virtual node to each exit node is zero, the positional relationship of each node in the path network will not be destroyed after the virtual node is constructed. Figure 3 shown.

[0166] When the virtual node is used as the starting node, the path network only needs to be traversed once to complete the search of all nodes at each exit node. The next node of the virtual node in the node set of the optimal route to any node represents the target exit of the node.

[0167] To ensure that pedestrians are not led to the fire scene by signs during the evacuation process, dangerous areas that are inaccessible during fires are excluded through path network reconstruction during the system's path planning stage.

[0168] During actual deployment, monitoring thresholds for temperature, visibility, and CO volume fraction will be set for sensors in the perception layer. The values of each threshold will be based on the critical values of each indicator that can ensure the normal operation of personnel.

[0169] When the system senses a hazard source or detects that the temperature, visibility, or CO volume fraction on any road section deviates from the normal range, the node or path corresponding to that road section in the path network will be determined as a dangerous node (path). The decision-making layer will then remove all dangerous nodes and dangerous paths from the path network to complete the reconstruction of the path network.

[0170] The path network is regarded as a weighted undirected graph in the path planning module. Before traversing the Dijkstra algorithm with an improved search strategy, the module will recalculate the weights of each path based on real-time data.

[0171] If the path weight is set to the length between adjacent nodes, the total cost considered in the path search is the distance from the starting node to the target node. The target exit of any node is the exit closest to that node. The global optimal route obtained by path planning with path length as the weight can be used as the basis for static sign direction indication. That is, when there are no hazards on the cruise ship, the default indication direction of the sign system will eventually guide pedestrians to their nearest exit.

[0172] Considering that the ultimate goal of emergency evacuation is to ensure the safe evacuation of people in the shortest possible time, dynamic signs will always guide pedestrians to evacuate from the exit as quickly as possible, but this exit may not be the closest to the pedestrians. The system decision-making layer uses the travel time of each section as the weight of each path when planning the global path.

[0173] When calculating the travel time of a road section, it is necessary to determine the distance between the two ends of the road section and the travel speed of the road section. The travel speed of the road section is affected by its congestion level and inclination state.

[0174] The inclination state of the section will change with the change of the hull posture. For the axial path in the path network, when the hull tilts longitudinally, the section corresponding to the longitudinal path will tilt longitudinally, and the section corresponding to the transverse path will tilt transversely. When the hull tilts transversely, the section corresponding to the transverse path will tilt longitudinally, and the section corresponding to the longitudinal path will tilt transversely.

[0175] Taking the bow trim of the hull as an example, the sections corresponding to path 6 and part of path 7 and path 16 are heel sections, and the sections corresponding to path 3, paths 8 to 9, paths 14 to 15, and another part of path 7 and path 16 are longitudinal sections. Figure 3 shown.

[0176] The direction of the road section's longitudinal inclination angle is relative to the pedestrian's direction. When making judgments, the pedestrian's direction of travel is always taken as the positive direction of the road section's longitudinal axis. When the longitudinal axis rotates counterclockwise, the longitudinal inclination angle is positive, indicating that the pedestrian is on an uphill section. When the longitudinal axis rotates clockwise, the longitudinal inclination angle is negative, indicating that the pedestrian is on a downhill section.

[0177] As the algorithm traverses the path network, the direction of pedestrian movement in each section can be determined based on the principle that the search order is opposite to the actual route direction. The longitudinal inclination angle of the corresponding section of the axial path can be obtained by combining the hull posture and path type.

[0178] Taking the bow trim of a ship as an example, the inclination angle of the heel section is consistent with the longitudinal trim angle of the ship. The following is an example of the rules for determining the inclination direction of the longitudinal trim section:

[0179] If node 2 / 3 is the current node and node 6 is the next node, the pedestrian is moving toward the stern of the ship, that is, uphill, on the corresponding section of path 3, so the pitch angle of the section is positive; if node 6 is the current node and node 2 / 3 is the next node, the pedestrian is moving toward the bow of the ship, that is, downhill, on the corresponding section of path 3, so the pitch angle of the section is positive, such as Figure 3 shown.

[0180] The output result of the Dijkstra algorithm in the path planning module is a set of all nodes on the optimal route from each node to the exit. All nodes in the set share the same target exit, and the next node in the direction of each node toward the exit is the basis for the direction indication marked at the node.

[0181] Since the identifiers with close distances are simplified as the same node when modeling the path network, the positional relationship between the node and the next node needs to be mapped to the indicated directions of different identifiers at the same time.

[0182] The change of the direction indicated by the dynamic sign depends on the transformation of the direction arrow, and the transformation of the direction arrow can be achieved by controlling the light emitting diode.

[0183] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0184] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A cruise passenger dynamic evacuation guidance system, characterized in that: The guidance system includes a monitoring and alarm module, a path planning module and a dynamic identification module, wherein the monitoring and alarm module leads the perception layer of the system, the path planning module leads the decision layer of the system, and the dynamic identification module leads the application layer of the system; The monitoring and alarm module is used for monitoring data and alarming; The path planning module is used to reconstruct the path network based on the monitored data, assign path weights, and use the Dijkstra algorithm with an improved search strategy to search for the global optimal route; Treat each marker in the space as a node, connect each pair of adjacent nodes to form several paths, simplify markers that are close to each other into the same node, construct paths blocked by obstacles into broken lines, and when multiple paths intersect, only the shortest path is retained; The path network is further simplified. Based on the fact that the bow and stern directions and the port and starboard directions represent the longitudinal and transverse directions of the ship, respectively, the paths along the two directions are called longitudinal paths and transverse paths, respectively. The longitudinal paths and transverse paths are called axial paths. The paths with smaller angles with the transverse and longitudinal axes in the path network are called axial paths, and the paths opposite to the axial paths are called oblique paths. When simplifying the path network, the path type is determined according to the following formula: ； in, Indicates that there are three paths: vertical, horizontal and diagonal. is the angle between the path and the bow and stern axis, is the threshold of the angle between the path and the horizontal and vertical axes, and its value range is ; The dynamic sign module is used to adjust the sign indication direction in real time according to the global optimal route based on the sign optimization layout; The optimized layout of the signs in the dynamic sign module is based on an evacuation sign layout optimization method, and the specific process of the optimization method is as follows: (1) The pedestrian action space and the sign layout space are set on the same rectangular coordinate plane, and the plane is discretized into a number of The walkable area and placement area are composed of rectangular units; (2) Using the multi-exit spatial guidance demand assessment algorithm, the difference in the distance from the exit to the pedestrian after each step is measured by the difference in the distance between the pedestrian and the exit when there is a sign and when there is no sign guidance. The distribution of guidance demand in the walkable area is obtained. The calculation formula is: ; in, The guidance needs of any unit in the walkable area, is the pedestrian’s current unit, is the next unit in the optimal route from the current unit to its nearest exit, is the shortened value of the distance from the current unit to the target exit after the pedestrian moves from the current unit to the next unit when the sign is in effect. is the average shortening value when the pedestrian moves to each adjacent unit; is the exit closest to the current unit, For unit To the exit The shortest distance, For unit To the exit The shortest distance; is any neighboring unit of the current unit, For unit The set of neighboring cells, is the shortened value of the distance from the current unit to the target exit after the pedestrian moves from the current unit to the adjacent unit when there is no sign. For unit The number of neighboring units; is the nearest exit to the adjacent unit, For unit To the exit The shortest distance, For unit To the exit The shortest distance; (3) Use genetic algorithm to solve the logo optimization layout, the formula is as follows: ； ; ; in, is the average matching degree between guidance demand and perception probability in the walkable area, that is, the guidance demand of all units and the corresponding perception probability The sum of products and the number of labels in the space The ratio, is the set of all units within the walkable area; is the actual size of the logo, is the horizontal distance from the pedestrian to the sign, is the horizontal observation angle, The height difference between the logo and the human eye is is the threshold for identifying the size of retinal imaging; is the set of all cells in the placeable area, is the set of all identifiers in the space, For any unit where the mark is located, is the angle of the marking direction, is the allowed number of markers in the space, To identify the number of units within the range of the system, it is determined by marking the legible area and the obstruction of obstacles; The diameter of the label readable area Expressed as ,in is the readability factor; the judgment formula for identifying the readability area is: ; in, To identify the coordinates of the location, The radius of the legible area of the logo; The ratio of the visible length of the logo to its actual length The calculation method is: ； in, is the vector pointing to the pedestrian. It is a vector pointing to the direction of the self.

2. A cruise ship passenger dynamic evacuation guidance system according to claim 1, characterized in that: The monitoring and alarm module includes a fire alarm system, an inertial measurement unit, and video monitoring equipment, as follows: The fire alarm system is composed of sensors and alarms, wherein the sensors include temperature sensors, carbon monoxide detectors and smoke detectors, which are used to monitor the temperature, CO volume fraction and visibility of the road section; The inertial measurement unit includes a speedometer, a gyroscope and a magnetometer, which are used to measure the hull attitude and tilt angle including roll and pitch; When the temperature, CO2 volume fraction or hull tilt angle of the monitored section exceeds the set threshold, or the visibility is lower than the set threshold, the alarm will be activated. At the same time, the relevant data collected by the monitoring alarm module will be transmitted in real time; The video surveillance equipment includes cameras deployed on various road sections, which are combined with wireless sensor networks and target detection and tracking technology to achieve the statistics and positioning of the number of pedestrians on each road section and the acquisition of crowd density data.

3. A cruise ship passenger dynamic evacuation guidance system according to claim 1, characterized in that: The path planning module uses the Dijkstra algorithm to search for the global optimal route, as follows: (1) Improve the search strategy of Dijkstra algorithm Construct a virtual node connected to all exit nodes, set the weight of each path adjacent to the virtual node to zero, and use the virtual node as the starting node to search all nodes in the path network. Before each search, reconstruct the path network and assign path weights. (2) Reconstructing the path network When a hazard source is sensed or the temperature, visibility, or CO volume fraction on any road section deviates from the set monitoring threshold, the node corresponding to the road section in the path network is determined to be a dangerous node and the corresponding path is determined to be a dangerous path. All dangerous nodes and dangerous paths are eliminated to complete the reconstruction of the path network. (3) Assigning path weights The travel time of each road section is used as the weight of its corresponding path in the path network. The specific formula is: ； The travel time of the road section Travel time of axial routes by route type and the travel time of the diagonal path ; The segment travel time of the axial path is calculated using the following formula: ; in, is the distance between the two ends of the road section, is the traffic speed of the road section; The diagonal path is decomposed into two axial paths based on its projection on the longitudinal and transverse axes. The travel time of the longitudinal and transverse paths is then calculated and summed to obtain the total travel time. Finally, the total travel time is corrected to obtain the section travel time of the diagonal path. The specific formula is as follows: ； ； in, is a correction factor used to avoid "stretching" the path.

4. A cruise ship passenger dynamic evacuation guidance system according to claim 3, characterized in that: The calculation formula for the traffic speed of the road section is: ; (1) When the ship is in an upright state, the travel time of a road section depends on its congestion level. The crowd density on the road section is an important indicator for measuring the congestion level of the road section. The traffic speed of a road section under congestion is a piecewise function of the crowd density on the road section. The specific formula is: ; ； in, is the speed before tilting, is the crowd density on the road section, is the total number of pedestrians gathered on the path, The available area of the path is the area of the area that pedestrians are actually allowed to reach after removing the space occupied by dynamic obstacles. (2) When the hull tilts, the tilt state of the road section is divided into longitudinal tilt, transverse tilt and combined tilt according to the direction of the rotation axis. The speed attenuation coefficient is the ratio of the speed after tilt to the speed before tilt. The speed of the longitudinal and transverse tilt sections under congestion is calculated by the following formula: ; ； in, is the speed of the longitudinal section, is the velocity attenuation coefficient of the longitudinal section, is the longitudinal inclination angle of the road section, is the speed of the inclined road section after tilting, is the velocity attenuation coefficient of the banked section, is the inclination angle of the road section; The bidirectional inclination angles caused by compound inclination further increase the difficulty of traffic on the road section. Assuming that the effects of longitudinal inclination and transverse inclination on traffic speed are not independent of each other when the road section is compound inclination, the traffic speed on the compound inclination road section can be expressed as: ； ； in, is the travel speed on the compound inclined road section, To obtain the speed of the longitudinal section and the speed of the inclined section The smaller value of is a penalty term, reflecting the further restriction of the passage speed caused by the combined effect of pitch and roll. The maximum permissible value of the total tilt angle is the sum of the absolute values of the pitch angle and the roll angle. The path will be difficult to pass. is the penalty coefficient, and its value range is , which indicates the degree of influence of the total inclination angle on the traffic speed. The larger the value, the more significant the penalty.

5. A cruise ship passenger dynamic evacuation guidance system according to claim 1, characterized in that: The dynamic identification module adjusts the identification direction in real time. The specific process is as follows: The global optimal route planned by the path planning module is used as the basis for the identification direction indication. The output result of the Dijkstra algorithm of the path planning module is a set of all nodes on the optimal route from each node to the exit. All nodes in the set share the same target exit. The next node in the direction of each node towards the exit is the guidance destination identified at the node.

Images