CN109900276B - Station real-time emergency path planning method based on point-line-surface obstacle model construction - Google Patents

Abstract

The invention is a real-time emergency path planning method for a station constructed based on a point-line-plane obstacle model. On the premise that the position information of indoor fixed obstacles is not lost, the algorithm ensures that the algorithm can calculate the three-dimensional optimal path after the occurrence of a station emergency in real time. Its core idea is In order to: before the route planning, the overall architectural design plan of the station is processed in advance, the construction of the obstacle map in the whole area is completed, and the information is recorded in the server to shorten the system running time. At the same time, considering that some escalators in the station can go up and down, and some escalators from the waiting room to the platform can only go down, and only the stairs can go up, so the attribute information such as all the up and down stairs in the obstacle is used to add to the path cost. In order to solve the problem that the planned path conforms to the actual situation. Through accurate and real-time path planning, the station security personnel are guided to complete the emergency disposal work with high efficiency.

Description

A real-time emergency path planning method for stations based on the point-line-surface obstacle model

technical field

The invention belongs to the technical field of intelligent security, and in particular relates to a real-time emergency path planning method for a station constructed based on a point-line-plane obstacle model.

Background technique

In modern life, with the continuous deepening of the Internet and intelligent media, people's travel has been greatly facilitated. When we go out, we are more and more inseparable from the help of the navigation system. The technology of the outdoor navigation system is quite mature, but there is still a certain room for development of the indoor navigation. The key points are the selection of the best indoor positioning solution, the modeling of indoor maps and Research on performance optimization of indoor navigation algorithms. At present, indoor navigation still needs to be further developed, so far there is no unified standard in the industry, and it is only designed for specific environments. For the construction of indoor multi-dimensional maps, Liu Xiaoxiao from China University of Mining and Technology divides the spatial structure of the building to be studied from coarse-grained to fine-grained, and includes the required information in its D-K path. In the data structure used by the planning algorithm, it takes a lot of time to manually modify the hierarchical order of partitions, horizontal connections, obstacles, etc. in the process, the workload is large, and the content is complex; Huang Kejia of Beijing SuperMap Software et al. Design and Implementation of Cross-floor Path Planning Technology for Indoor Road Network" In the original data processing, the multi-storey road data is obtained, and then the bidirectional improved A* algorithm is used to plan the cross-floor path. , the complex road information brings complex logic and heavy calculation pressure to the algorithm, and the construction of the stair connection table cannot well contain equipment information such as railway stations, so it cannot bring accurate shortest planning paths that conform to the actual situation on site. ;Beijing University of Architecture's "Integrated 3D Navigation Path Planning for Indoor and Outdoor Digital Campus" is more about the reading of 3D models and UI design. The navigation planning engine just uses ArcGIS ready-made software to analyze the 3D navigation path network inside and outside the campus. Extraction and planning.

At present, indoor route planning is based on the complex environment information of indoor map data, which makes it impossible to select the road network like outdoor navigation. In addition, due to the complex roads inside the railway station, the connection between elevators is more complicated than that of elevators in other places. The problem of slow algorithm processing speed brought about by the slenderness of the platform layer makes the path planning more complicated. At present, the railway station has not really provided real-time emergency path planning to deal with emergencies.

SUMMARY OF THE INVENTION

For the current indoor path planning algorithms such as the D-K algorithm, the early data modification workload is large, the indoor road network needs to extract a large amount of information from the data used by the improved A* algorithm, and the genetic algorithm and ant colony algorithm of the intelligent bionic algorithm are applied in practice. The computational complexity in the traditional path planning method is too high, the algorithm structure is too simple, but the search efficiency is low, and as the number of obstacles increases or the environment becomes more complex, the complexity of the algorithm will increase. Rapidly increasing, it is impossible to construct maps for complex scenes such as large public places, especially railway stations, to complete path planning. Real-time path planning guides station security personnel to complete emergency handling work efficiently.

The technical scheme that realizes the object of the present invention is:

A real-time emergency path planning method for a station constructed based on a point-line-surface obstacle model, comprising the following steps:

Step 1: Create an environmental map of station security personnel:

1-1. Data collection: The point cloud data of the station to be planned collected by the laser scanner, the buildings and equipment related to the route planning are marked in advance whether it is traffic or the up and down attributes, floor information, and the storage of the map information is assisted. Work, generate multiple KML files according to the number of floors;

1-2. Map data analysis: Loop through the KML file obtained in step 1-1, read in batches, partition the building, define two types of structural units: closed and semi-open and closed, and complete the map data analysis ;

1-3. Storage of longitude, latitude and height information of structural units: in the analysis data of step 1-2, look for the contour inflection points of structural units, building and equipment attribute information related to path planning, and all the longitude and latitude information corresponding to the attribute information. Stored in the map initial data matrix [,]map;

1-4. Use the point, line and surface obstacle construction method to complete the transformation of the initial map data matrix [,]map to the map data matrix [lon_error,lat_error]map and then to the obstacle model matrix [b_lon_error,b_lat_error]b_map, and realize the transformation from world coordinates to Mapping of pixel coordinates:

(1) From the map initial data matrix [,]map information stored in steps 1-3, find the boundary value of the obstacle, and record the maximum and minimum values of the longitude and latitude coordinates of the corresponding obstacle, which are recorded as lon_max, lat_max and lon_min, lat_min;

(2) Determine the scale of the map data matrix according to the difference between the maximum and minimum values of longitude and the difference between the maximum and minimum values of latitude, and denote the map data matrix as [lon_error,lat_error]map; take the point (lon_min, lat_min) as the origin of world coordinates,

Where lon_error=lon_max-lon_min;

lat_error=lat_max-lat_min;

The map data is extracted by the straight-line scanning method of point, line and surface, the latitude and longitude coordinates of each obstacle vertex are extracted in turn, the mapping relationship between the world coordinates and the pixel coordinates is established, and the original longitude and latitude information is further processed, respectively recorded as ( b_lon_1,b_lat_1),(b_lon_2,b_lat_2),...,(b_lon_n,b_lat_n), through b_lon_i=k*b_lat_i+b, traverse to get the position of all coordinate points passing through the line, set all points on the line to 0, representing The area is impassable; when the area is passable, the corresponding position pixel is 0, and the initial obstacle model matrix is obtained, and the map data matrix is all converted to the pixel coordinate system and the corresponding information is stored in the established initial obstacle model matrix;

where i represents the storage order of vertices, n is the total number of storage points, 0<i≤n; k is the slope of the line connecting two adjacent endpoints, and b is the straight line intercept;

(3) Double solid line processing is performed on the pixel coordinate mutation position in step (2) to obtain the final obstacle model matrix [b_lon_error, b_lat_error] b_map, and the construction of the station security personnel environment map is completed;

Step 2: Complete the construction of the station security personnel location information data link

Station security personnel update location information in real time to the base station server on the corresponding floor through handheld or wearable devices, and the base station server exchanges information with the route planning server to complete the data preparation before emergencies;

Step 3: Use the improved A* algorithm to calculate the path to the emergency scene for all security personnel and store it, and screen and display the paths that meet the conditions by setting constraints for security personnel to the emergency scene location;

3-1. Data transmission after emergencies: The path planning server stores the obstacle model matrix [b_lon_error,b_lat_error]b_map constructed in step 1, and the client brought by any station security personnel uploads the nearby data in time after an emergency occurs. Location information to the route planning server;

3-2. Uploading the location information of multiple security personnel: The path planning server obtains the location information of all security personnel through the positioning base station, and uses the improved A* algorithm to calculate accurately in real time; the location information includes accurate latitude and longitude and floor height, according to the building. The scale is used to construct obstacles with variable steps; the precise position information is converted into the relative position information relative to the origin of the pixel coordinates in the pixel coordinates;

3-3. The multi-path calculation of the improved A* algorithm: the path planning server uses the location information of all security personnel at the station and the location information of emergencies uploaded by a security personnel to the path planning server, and comprehensively considers the elevator up and down of the planned path. , underground passage paths and the actual situation of some elevators dedicated to security personnel, the improved A* algorithm is used many times to calculate and store all the routes of security personnel to the emergency scene in real time;

3-4. Realization of cross-floor algorithm: The path stored in the previous step is set by setting the path distance and the number of spanning floors. When the security emergency path on the same floor is smaller than any building and equipment related to path planning on the floor When the path of the site is completed, the path planning ends; otherwise, record the buildings and equipment related to path planning that are the closest to the emergency site, and accurately determine the coordinates of the buildings and equipment related to path planning. Near the floor, and then take the buildings and equipment related to path planning as nodes, find the buildings and equipment exits related to path planning corresponding to the next target floor, transfer to the calculation of the next floor, and carry out the total distance after the navigation path is stored. Comparing the cost, when the distance between the two paths across the floors is greater than the distance between the floors, the shortest path on the same floor is directly output, otherwise the sum of the paths across the floors is output, and the path planning is over; Emergency planning paths for personnel, and at the same time the path is output, instructions are sent to the station security personnel through the base station server on the corresponding floor to complete high-speed and high-accuracy emergency handling.

The beneficial effects of the present invention are:

On the one hand, the method of the present invention uses the centralized control method, the path planning server obtains the global information of the working environment of the station security personnel, and provides traffic information for emergency emergency response; The path information calculated by the layer base station server and the access path planning server. The route planning server with fast computing speed performs complex data calculation work, and sends the obtained route results to the client where the security personnel are located, which shortens the system running time. In addition, comprehensive information is obtained after the construction of the point, line and surface obstacle model, and variable step size data processing is used according to the actual situation of the building, which greatly controls the size of the model information matrix, and the improved "double solid line" obstacle construction method solves the problem. The low precision of the map data matrix leads to the problem of "penetrating the wall" of the improved A* algorithm, which solves the problem of excessive calculation load caused by the large scale of the server-side map. The method can ensure the effectiveness and real-time performance of the improved A* algorithm under the accuracy of keeping the information such as obstacles as undistorted as possible. About 4 seconds after the response, the path planning is carried out immediately, and the final output three-dimensional path accuracy is at the meter level instead of the simple distribution of some nodes, thus ensuring the real-time emergency path planning of the station, greatly improving the office efficiency of emergency rescue of security personnel, and more It greatly ensures the safety of passengers, and adds practicality and innovation to the intelligent monitoring of the station.

Description of drawings

Figure 1 3D map of a railway station map data.

Figure 2. The 2.5D KML map on the first floor of a railway station and the matrix simulation diagram after analysis.

Figure 3. The 2.5D KML map on the second floor of a railway station and the matrix simulation diagram after analysis.

Figure 4. The 2.5D KML map on the third floor of a railway station and the matrix simulation diagram after analysis.

Figure 5 is a block diagram of the on-site data link structure.

Figure 6. Flow chart of three-dimensional emergency path planning of a railway station.

Figure 7 is a flowchart of the improved A* pathfinding algorithm.

Figure 8 shows the simulation and simulation display of the host computer after the method of the present invention is encapsulated.

Figure 9 shows the effect diagram of the integrated simulation display of the path planning module of the upper computer in the station monitoring room after the method of the present invention is encapsulated.

Detailed ways

The present invention will be further discussed in detail below with reference to the accompanying drawings and specific embodiments.

The present invention is a real-time emergency path planning method for a station constructed based on a point-line-plane obstacle model, comprising the following steps:

Step 1: Create an environmental map of station security personnel:

1-1. Data collection: The point cloud data of the station to be planned collected by the laser scanner is converted into 3ds format and imported into 3dmax; the localspace software is used to draw a rough outline, and the buildings and equipment related to path planning are Mark in advance whether to pass or go up and down attributes, floor information, the buildings and equipment related to path planning refer to elevators, artificial escalators, outbound passages, escalators for security personnel, etc.), to assist in completing the storage of map information, such as For some elevators, set the flag to 10 in the properties, indicating that only the downlink can be used, and the flag is 01, which means that the elevator can only go up and not down; generate multiple KML files according to the number of floors;

1-2. Map data analysis: Loop through the KML file obtained in step 1-1, read in batches, partition the building, define two types of structural units: closed and semi-open and closed, and complete the map data analysis ;

1-3. Storage of longitude, latitude and height information of structural units: in the analysis data of step 1-2, look for the contour inflection points of structural units, building and equipment attribute information related to path planning, and all the longitude and latitude information corresponding to the attribute information. Stored in the map initial data matrix [,]map, which is convenient for subsequent modules to retrieve data and use;

1-4. Use the point, line and surface obstacle construction method to complete the transformation of the initial map data matrix [,]map to the map data matrix [lon_error,lat_error]map and then to the obstacle model matrix [b_lon_error,b_lat_error]b_map, and realize the transformation from world coordinates to Mapping of pixel coordinates:

(4) From the map initial data matrix [,]map information stored in steps 1-3, find the boundary value of the obstacle, and record the maximum and minimum values of the longitude and latitude coordinates of the corresponding obstacle, which are recorded as lon_max, lat_max and lon_min, lat_min;

(5) Determine the scale of the map data matrix according to the difference between the maximum and minimum values of longitude and the difference between the maximum and minimum values of latitude, and record the map data matrix as [lon_error,lat_error]map; take the point (lon_min, lat_min) as the origin of world coordinates,

Where lon_error=lon_max-lon_min;

lat_error=lat_max-lat_min;

(6) Extract the map data by using the point-line-surface line scanning method, and complete the extraction of the latitude and longitude coordinates of the vertices of each obstacle in turn, denoted as (lon_1, lat_1), (lon_2, lat_2), ..., (lon_n, lat_n), establish the mapping relationship between world coordinates and pixel coordinates, and further process the original latitude and longitude information, which are recorded as (b_lon_1, b_lat_1), (b_lon_2, b_lat_2), ..., (b_lon_n, b_lat_n), through b_lon_i=k* b_lat_i+b, traverse to get the positions of all coordinate points passing through the line, set all points on the line to 0, indicating that the area is impassable; when the area is passable, the corresponding position pixel is 0, and the initial obstacle model matrix is obtained. The data matrix is all converted to the pixel coordinate system and the corresponding information is stored in the established initial obstacle model matrix;

b_lon_i=(lon_i-lon_min)*1000000

b_lat_i=(lat_i-lat_min)*1000000

b=b_lon_i-k*b_lat_i

where i represents the storage order of vertices, n is the total number of storage points, 0<i≤n;

(b_lon_i, b_lat_i), (b_lon_i+1, b_lat_i+1) are the deviation information of the longitude and latitude of the adjacent endpoints of the obstacles in this layer of buildings relative to the origin of the pixel coordinate system, and k is the distance between the two adjacent endpoints. slope.

(7) When using the improved A* algorithm for the model constructed above, due to the processing of variable step size, the planned route will pass through obstacles at the pixel coordinate mutation position. This scheme doubles the pixel coordinate mutation position. The solid line processing makes the planned path more accurate, and the final obstacle model matrix [b_lon_error, b_lat_error]b_map is obtained.

Step 2: Complete the construction of the station security personnel location information data link

Station security personnel update location information in real time to the base station server on the corresponding floor through handheld or wearable devices, and the base station server exchanges information with the route planning server to complete the data preparation before emergencies;

Step 3: Use the improved A* algorithm to calculate the path to the emergency scene for all security personnel and store it, and screen and display the paths that meet the conditions by setting constraints for security personnel to the emergency scene location;

3-1. Data transmission after emergencies: The path planning server stores the obstacle model matrix [b_lon_error,b_lat_error]b_map constructed in step 1, and the client brought by any station security personnel uploads the nearby data in time after an emergency occurs. location information to the route planning server,

3-2. Uploading the location information of multiple security personnel: The path planning server obtains the location information of m security personnel through the positioning base station, and uses the improved A* algorithm for accurate real-time calculation; the location information includes accurate longitude, latitude and floor height, according to the building The object scale is used to construct the obstacle with variable step size; the precise position information is converted into the relative position information relative to the pixel coordinate origin under the pixel coordinates, which can be used by the improved A* algorithm.

list<h> ^j =list<h> ^j

In the formula, 0<j≤m, list<lon,lat> ^j.lon is the longitude information in the combination of longitude and latitude information of security personnel, list<lon,lat> ^j.lat is the latitude information in the combination of longitude and latitude information of security personnel, and Step_size is Custom step value. list<lon,lat> ^j is the latitude and longitude information of the jth security personnel, list<x> ^j is the relative position of the jth security personnel in the longitude relative to the origin, list<y> ^j is the jth security personnel relative to the longitude The relative position of the origin in latitude, list<h> ^j is the actual height information of the jth security personnel; x and y represent the relative position of the security personnel in latitude and longitude relative to the origin, h is the actual height of the floor .

3-3. The multi-path calculation of the improved A* algorithm: the path planning server uses the location information of all security personnel at the station and the location information of emergencies uploaded by a security personnel to the path planning server, and comprehensively considers the elevator up and down of the planned path. , underground passage paths, and some elevators for security personnel, etc., the improved A* algorithm is used many times to calculate and store all the routes of security personnel to the emergency scene in real time.

3-4. Implementation of the cross-floor algorithm: The path stored in the previous step is set by setting the constraints of the path distance and the number of spanning floors. When the security emergency path on the same floor is smaller than the path of any elevator on the floor to the scene, the path planning ends. ; Otherwise, record the elevator closest to the emergency scene, accurately determine the coordinate information of the elevator's longitude and latitude through the mapping relationship, search for the adjacent floors, and then use the elevator as a node to find the elevator exit corresponding to the next target floor, and transfer to the next floor. After the navigation path is stored, the total distance cost is compared. When the distance between the two paths across the floor is greater than the distance of the floor, the shortest path on the same floor is directly output, otherwise the sum of the paths across the floor is output, and the path planning ends; screening The fastest emergency planning path for security personnel to reach the emergency scene, and at the same time when the path is output, instructions are sent to the station security personnel through the base station server on the corresponding floor to complete high-speed and high-accuracy emergency handling work.

The algorithm flow of the specific implementation of the improved A* algorithm belongs to the existing technology, and the details are as follows:

(1) Create two linked lists to store the nodes to be detected and the detected nodes: respectively create two empty linked lists: OPEN linked list and CLOSED linked list, and put the starting point in the OPEN list;

(2). If there is no node in the OPEN table, it means that the path planning fails and the algorithm ends; if there is a node, go to step (3);

(3). Select the node N with the smallest evaluation function F value in the OPEN table and put it in the CLOSED table;

(4). Determine whether node N is the target node: if the Manhattan distance between the current node N and the target node is 0, then the target point is reached; if the target node is reached, the path planning is successful, go to step (10), otherwise, it has not yet arrived target node, go to step (5);

(5). Expanding node N: Select the one that is not in the CLOSED table, and use the adjacent intersection grid in front of the road where node N is currently located as a connected child node, and calculate its F value;

The formula for calculating the F value is:

F(n)=G'(n)+H'(n) (3)

Among them, F(n) is the evaluation function of node n, G'(n) is the actual cost function of node N, which represents the movement cost from the starting node to the current node, H'(n) is the heuristic function of node N, Represents the estimated movement cost from the current node to the target node.

(6) Determine the extended connected child node: determine whether the extended connected child node is in the OPEN table, if the child node is in the OPEN table, go to step (8), otherwise go to step (7);

(7). Add the expanded connected child node to the OPEN table: add the expanded connected child node to the OPEN table as the child node of the node, point the parent node pointer of the node to node N, and then go to step (2);

(8) Determine the F value of the expansion node: calculate the F value of the node passing through the node N, and the calculation method of the F value is the same as the calculation method of the F value in step (5); if it is greater than its own F value, go to step ( 9); otherwise, go to step (2);

(9) Update the value of the extension node Use the value of the node itself to update the value of the node in the table, and go to step (2);

(10). Finally, the path obtained by extending each node from the target point to its parent node is the path planned by this method.

The real-time emergency path planning method of the station constructed based on the point-line-plane obstacle model of the present invention is encapsulated, and the test simulation route is carried out; the final environment and data deployment are combined with the actual situation of the specific station, and the relevant functions of path planning are placed in the system. The real-time three-dimensional path planning of this application is an auxiliary tool for the station security personnel after responding to emergencies, and can effectively improve the safety and efficiency of the station.

The above path planning server is installed in the monitoring room of the station, and a base station server is placed on each floor of the station to realize the transmission of security personnel location information data.

In the present invention, the origin of the world coordinates corresponds to the origin of the pixel coordinates, and both can be considered as (lon_min, lat_min).

Example 1

The real-time emergency path planning method for a station constructed based on a point-line-surface obstacle model in this embodiment includes the following steps:

Step 1: Create an environmental map of station security personnel:

1-1. Data collection: Taking a three-story train station in China as an example, the first floor is mainly a ticket hall, a second-floor platform, and a third-floor waiting room. The three-dimensional software displays the building as shown in Figure 1, which is collected by a laser scanner. The point cloud data of the station to be planned is converted into 3ds format and imported into 3dmax; the localspace software is used to draw a rough outline, and the buildings and equipment related to route planning are marked in advance whether it is traffic or up and down attributes, and floor information. The buildings and equipment refer to paths such as elevators, artificial escalators, exit passages, and escalators dedicated to security personnel, which assist in the storage of map information. For example, for some elevators, set the flag to 10 in the attribute, indicating that only downhill, not upbound, is allowed. The representative whose flag is 01 can only go up and down; one KML file is generated for each floor, and three different KML files are generated for three floors;

1-2. Map data analysis: loop through the three KML files obtained in step 1-1, read in batches, divide the building, define two types of structural units, closed and semi-open and closed, to complete the map data analysis;

1-3. Storage of longitude, latitude and height information of structural units: in the analysis data of step 1-2, look for the contour inflection points of structural units, building and equipment attribute information related to path planning, and all the longitude and latitude information corresponding to the attribute information. Stored in the map initial data matrix [,]map, which is convenient for subsequent modules to retrieve data and use;

1-4. Use the point, line and surface obstacle construction method to complete the transformation of the initial map data matrix [,]map to the map data matrix [lon_error,lat_error]map and then to the obstacle model matrix [b_lon_error,b_lat_error]b_map, and realize the transformation from world coordinates to The mapping of pixel coordinates, the obstacle model matrix with the matrix scale of [193,182], [189,493] and [119,142], the initial map data matrix of the three floors and the simulation map of the obstacle model matrix are shown in Figure 2 and Figure 3 respectively. and Figure 4:

(1) From the map initial data matrix [,]map information stored in steps 1-3, find the boundary value of the obstacle, and record the maximum and minimum values of the longitude and latitude coordinates of the obstacle, which are recorded as lon_max, lat_max respectively and lon_min, lat_min;

(2) Determine the scale of the map data matrix according to the difference between the above-mentioned longitude maximum value and minimum value, and the latitude maximum value and minimum value difference, and record the map data matrix as [lon_error,lat_error]map;

Where lon_error=lon_max-lon_min;

lat_error=lat_max-lat_min;

(3) Extract the map data by using the line scanning method of the point, line and surface, and complete the extraction of the latitude and longitude coordinates of the vertices of each obstacle in turn, which are recorded as (lon_1, lat_1), (lon_2, lat_2), ..., (lon_n, lat_n), to establish a mapping relationship between world coordinates and pixel coordinates. Further processing of the original latitude and longitude information, respectively recorded as (b_lon_1, b_lat_1), (b_lon_2, b_lat_2), ..., (b_lon_n, b_lat_n), through b_lon_i=k*b_lat_i+b, traverse to get the position of all coordinate points passing through the line , set all the points on the line to 0, which means the area is impassable; when the area is passable, the corresponding position pixel is 0, and the initial obstacle model matrix is obtained, and the map data matrix is all converted to the pixel coordinate system and the established initial obstacle Material model matrix for corresponding information storage;

b_lon_i=(lon_i-lon_min)*1000000

b_lat_i=(lat_i-lat_min)*1000000

b=b_lon_i-k*b_lat_i

where i represents the storage order of vertices, n is the total number of storage points, 0<i≤n;

(b_lon_i, b_lat_i), (b_lon_i+1, b_lat_i+1) are the deviation information of the longitude and latitude of the adjacent endpoints of all fixed obstacles in this layer of buildings relative to the origin of the pixel coordinate system, and k is the connection between the two adjacent endpoints. the slope of the line.

(4) When using the improved A* algorithm for the model constructed above, due to the processing of variable step size, the planned route will pass through obstacles at the pixel coordinate mutation position. This scheme doubles the pixel coordinate mutation position. The solid line processing makes the planned path more accurate, and the final obstacle model matrix [b_lon_error, b_lat_error]b_map is obtained.

Step 2: Complete the construction of the station security personnel location information data link

Station security personnel update location information in real time to the base station server on the corresponding floor through handheld or wearable devices, and the base station server exchanges information with the route planning server to complete the data preparation before emergencies;

Step 3: Use the improved A* algorithm to calculate the path to the emergency scene for all security personnel and store it, and screen and display the paths that meet the conditions by setting constraints for security personnel to the emergency scene location;

3-1. Data transmission after emergencies: The path planning server stores the obstacle model matrix [b_lon_error,b_lat_error]b_map constructed in step 1, and the client brought by any station security personnel uploads the nearby data in time after an emergency occurs. The location information is sent to the route planning server. During the test, the location where one of the security personnel clicked to upload (112.581745804, 37.860034226, 2) is the emergency location.

3-2. Uploading the location information of multiple security personnel: The path planning server obtains the location information of m security personnel through the positioning base station, and uses the improved A* algorithm for accurate real-time calculation; the location information includes accurate longitude, latitude and floor height, according to the building The object scale is used to construct the obstacle with variable step size; the precise position information is converted into the relative position information relative to the pixel coordinate origin under the pixel coordinates, which can be used by the improved A* algorithm.

list<h> ^j =list<h> ^j

In the formula, 0<j≤m, list<lon,lat> ^j.lon is the longitude information in the combination of longitude and latitude information of security personnel, list<lon,lat> ^j.lat is the latitude information in the combination of longitude and latitude information of security personnel, and Step_size is Custom step value. list<lon,lat> ^j is the latitude and longitude information of the jth security personnel, list<x> ^j is the relative position of the jth security personnel in the longitude relative to the origin, list<y> ^j is the jth security personnel relative to the longitude The relative position of the origin in the latitude, list<h> ^j is the actual height information of the jth security personnel; x and y respectively represent the relative position of the security personnel in latitude and longitude relative to the origin, and h is the actual floor. high.

In actual work, the simulation m=4, that is, four security personnel upload the route planning server, and the location coordinates are (112.581394226, 37.859811784, 2), (112.581372551, 37.860340320, 2), (112.581760254, 37.860281382, 3) and (112.59596862000000) ,3), the train station set the three floor heights as 5.2m, 17.5m and 26.5m respectively, and set the cost of the underground passage to 10.0m, looking for a security personnel closest to the emergency scene.

3-3. The multi-path calculation of the improved A* algorithm: the path planning server uses the location information of all security personnel at the station and the location information of emergencies uploaded by a security personnel to the path planning server, and comprehensively considers the elevator up and down of the planned path. , underground passage paths, and some elevators dedicated to security personnel. The improved A* algorithm was used many times to calculate and store all routes from security personnel to the emergency scene in real time. During the algorithm test, a total of 4 routes to the security personnel were planned. The paths of personnel, of which 2 belong to the same floor and 2 belong to the cross-floor path, respectively record the data of the latitude, longitude and altitude information of each step.

3-4. Implementation of the cross-floor algorithm: The path stored in the previous step is set by setting the constraints of the path distance and the number of spanning floors. When the security emergency path on the same floor is smaller than the path of any elevator on the floor to the scene, the path planning ends. ; Otherwise, record the elevator closest to the emergency scene, accurately determine the coordinate information of the elevator's longitude and latitude through the mapping relationship, search for the adjacent floors, and then use the elevator as a node to find the elevator exit corresponding to the next target floor, and transfer to the next floor. After the navigation path is stored, the total distance cost is compared. When the distance between the two paths across the floor is greater than the distance of the floor, the shortest path on the same floor is directly output, otherwise the sum of the paths across the floor is output, and the path planning ends; screening The fastest emergency planning path for security personnel to arrive at the emergency scene. When the path is output, instructions are sent to the station security personnel through the base station server on the corresponding floor to complete the high-speed and high-accuracy emergency handling work, and finally find the distance to the emergency. For the path of the nearest security personnel on site, the path planning server outputs corresponding signals to notify the corresponding personnel to go to the site for processing. The flow chart of the cross-floor path algorithm is shown in Figure 6.

Among them, the algorithm flow of the specific implementation of the improved A* algorithm belongs to the existing technology. The algorithm flow chart is shown in Figure 7, and the details are as follows:

(1) Create two linked lists to store the nodes to be detected and the detected nodes: respectively create two empty linked lists: OPEN linked list and CLOSED linked list, and put the starting point in the OPEN list;

(2). If there is no node in the OPEN table, it means that the path planning fails and the algorithm ends; if there is a node, go to step (3);

(3). Select the node N with the smallest evaluation function F value in the OPEN table and put it in the CLOSED table;

(4). Determine whether node N is the target node: if the Manhattan distance between the current node N and the target node is 0, then the target point is reached; if the target node is reached, the path planning is successful, go to step (10), otherwise, it has not yet arrived target node, go to step (5);

(5). Expanding node N: Select the one that is not in the CLOSED table, and use the adjacent intersection grid in front of the road where node N is currently located as a connected child node, and calculate its F value;

The formula for calculating the F value is:

F(n)=G'(n)+H'(n) (3)

Among them, F(n) is the evaluation function of node n, G'(n) is the actual cost function of node N, which represents the movement cost from the starting node to the current node, H'(n) is the heuristic function of node N, Represents the estimated movement cost from the current node to the target node.

(6) Determine the extended connected child node: determine whether the extended connected child node is in the OPEN table, if the child node is in the OPEN table, go to step (8), otherwise go to step (7);

(7). Add the expanded connected child node to the OPEN table: add the expanded connected child node to the OPEN table as the child node of the node, point the parent node pointer of the node to node N, and then go to step (2);

(8) Determine the F value of the expansion node: calculate the F value of the node passing through the node N, and the calculation method of the F value is the same as the calculation method of the F value in step (5); if it is greater than its own F value, go to step ( 9); otherwise, go to step (2);

(9) Update the value of the extension node Use the value of the node itself to update the value of the node in the table, and go to step (2);

(10). Finally, the path obtained by extending each node from the target point to its parent node is the path planned by this method.

The real-time emergency path planning method of the station constructed based on the point-line-plane obstacle model of the present invention is encapsulated, and a test simulation is carried out to simulate the route. The route simulation diagram of part of the calculated two-story platform is shown in Figure 8; the final environment and data deployment and specific station Combined with the actual situation, the relevant functions of path planning are placed in the station monitoring room and tested and run. The screenshots of the specific partial function modules are shown in Figure 9. In Figure 9, there are five people in total: one of the security personnel is a simulated emergency scene. He went to upload the location information to start the program to find the way. Road King planned to find all their route maps from the remaining four people on different floors. Finally, through comparison, he found the path with the least cost, that is, the distance from the emergency scene. The closest one. The real-time three-dimensional path planning of the present application is an auxiliary tool for station security personnel after responding to emergencies, and can effectively improve the safety and work efficiency of the station.

The above path planning server is installed in the monitoring room of the station, and a base station server is placed on each floor of the station to realize the transmission of security personnel location information data.

The present invention is a real-time emergency path planning method for a station constructed based on a point-line-plane obstacle model. On the premise of not losing the position information of fixed indoor obstacles, the algorithm ensures that the algorithm can calculate the three-dimensional optimal path after the occurrence of a station emergency in real time. The idea is: before the route planning, the overall architectural design plan of the station is processed in advance, the construction of the obstacle map in the entire area is completed, and the information is recorded in the server to shorten the system running time. At the same time, considering that some escalators in the station can go up and down, and some escalators from the waiting room to the platform can only go down, and only the stairs can go up, so the attribute information such as all the up and down stairs in the obstacle is used to add to the path cost. In order to solve the problem that the planned path conforms to the actual situation. The emergency path planning method is suitable for large public places such as railway stations, shopping malls, and airports.

Matters not addressed in the present invention are known in the art.

Claims (4)

1. A real-time emergency path planning method for a station constructed based on a point-line-plane obstacle model comprises the following steps:

the method comprises the following steps: establishing an environment map of station security personnel:

1-1, acquiring data, namely marking whether the buildings and equipment related to path planning pass or uplink and downlink attributes and floor information in advance through point cloud data of a station to be planned acquired by a laser scanner, assisting in finishing the storage work of map information, and generating a plurality of KM L files according to the number of floors;

1-2, analyzing map data, namely circularly traversing and reading in batch from the KM L file obtained in the step 1-1, partitioning the interior of a building, respectively defining two types of closed structural units and semi-open structural units, and completing map data analysis;

1-3, storing longitude and latitude height information of the structural unit: in the analysis data in the step 1-2, searching the contour inflection point of the structural unit, and building and equipment attribute information related to path planning, wherein longitude and latitude information corresponding to the attribute information is all stored in a map initial data matrix [, ] map;

1-4, completing the conversion from the map initial data matrix [, ] map to the map data matrix [ lon _ error, lat _ error ] map and then to the obstacle model matrix [ b _ lon _ error, b _ lat _ error ] b _ map by using a point-line-plane obstacle construction method, and realizing the mapping from world coordinates to pixel coordinates:

(1) finding the boundary value of the obstacle from the map initial data matrix [, ] map information stored in the step 1-3, recording the maximum value and the minimum value of longitude and latitude coordinates of the corresponding obstacle, and respectively recording the maximum value and the minimum value as lon _ max, lat _ max, lon _ min and lat _ min;

(2) determining the scale of a map data matrix according to the difference value of the maximum value and the minimum value of the longitude coordinate and the difference value of the maximum value and the minimum value of the latitude coordinate, and recording the map data matrix as [ lon _ error, lat _ error ] map; using the point (lon _ min, lat _ min) as world coordinate origin, wherein

Extracting data of map data by adopting a point-line-surface linear scanning method, sequentially completing the extraction of longitude and latitude coordinates of each barrier vertex, establishing a mapping relation between world coordinates and pixel coordinates, further processing original longitude and latitude information, and respectively recording the information as

、

、…、

Traversal is obtained by

Setting all points on a straight line to be 0 according to the positions of all coordinate points of the line, and representing that the area is not accessible; when the area can pass through, the pixel of the corresponding position is 0 to obtain an initial obstacle model matrix, and the map data matrix is integratedConverting the initial obstacle model matrix into a pixel coordinate system and storing corresponding information in the established initial obstacle model matrix;

where i represents the storage order of the vertices, n is the total number of storage points, 0<i

n; k is the slope of the connecting line of two adjacent end points, and b is the straight line intercept;

(3) performing double-solid-line processing on the pixel point coordinate mutation position in the step (2) to obtain a final obstacle model matrix [ b _ lon _ error, b _ lat _ error ] b _ map, and completing construction of a station security personnel environment map;

step two: finishing the construction of a station security personnel position information data chain

Station security personnel update position information to a corresponding floor base station server in real time through handheld or wearable equipment, and the base station server exchanges information with a path planning server to complete data preparation work before an emergency;

step three: calculating paths reaching the emergency scene for all security personnel by using an improved A-algorithm, storing the paths, and screening and displaying paths meeting conditions by setting constraint conditions from the security personnel to the emergency scene;

3-1, data transmission after emergency: the path planning server stores the barrier model matrix [ b _ lon _ error, b _ lat _ error ] b _ map constructed in the step one, and a client carried by any station security personnel uploads nearby position information to the path planning server in time after an emergency occurs;

3-2, uploading position information of multiple security personnel: the path planning server acquires the position information of all security personnel through the positioning base station, and the improved A-star algorithm is used for accurately calculating in real time; the position information comprises accurate longitude and latitude and floor height, and the step length-variable barrier construction is carried out according to the building scale; converting the accurate position information into relative position information relative to the origin of the pixel coordinate under the pixel coordinate;

3-3, multipath calculation for the improved a-algorithm: the path planning server uses the position information of all security personnel in a station and the position information of an emergency uploaded to the path planning server by a certain security personnel, comprehensively considers the practical conditions of the up-down elevator path and the underground passage path of the planned path and the special elevator of part of the security personnel, uses the improved A algorithm for many times, and calculates and stores all paths from the security personnel to the emergency site in real time;

3-4, implementation of a cross-floor algorithm: when the security emergency path of the same floor is smaller than the path of any building and equipment related to path planning to the site of the floor through setting the path distance and the constraint conditions of the number of the spanning floors, the path planning is finished; otherwise, recording the building and the equipment which are closest to the site of the emergency and related to the path planning, accurately determining coordinate information of the longitude and latitude of the building and the equipment related to the path planning and converted through a mapping relation, searching for an adjacent floor, further taking the building and the equipment related to the path planning as nodes, finding a building and an equipment outlet which correspond to a next target floor and are related to the path planning, converting the building and the equipment outlet into calculation of the next floor, comparing total path costs after storing navigation paths, directly outputting shortest paths of the same floor when two paths of cross-floors are greater than the distance of the floor, otherwise, outputting path summation of the cross-floors, and finishing the path planning; and screening out the emergency planning path of the security personnel which reaches the emergency scene fastest, outputting the path and sending an instruction to the station security personnel through the corresponding floor base station server to complete the emergency processing work with high speed and high accuracy.

2. A planning method according to claim 1, wherein the mapping relationship between world coordinates and pixel coordinates is:

wherein,

deviation information of the longitude and latitude of the ith vertex of the obstacle in the current floor building relative to the origin of the pixel coordinate system.

3. The planning method according to claim 1, characterized in that the buildings and equipment related to path planning in step 1-1 are elevators, escalators, outbound tunnels or escalators dedicated to security personnel.

4. The planning method according to claim 1, wherein the method is applied to train stations, shopping malls, airports.

Images