Method and system for automatic generation of an aircraft taxiway path

Abstract

The application discloses an automatic generation method and system of an aircraft taxiway, which comprises the following steps: acquiring airport GIS map data, and constructing a taxiway hierarchical graph based on an H3 spatial grid; after receiving an instruction containing a taxiway name sequence, automatically integrating the instruction with gate and target point information, and mapping starting and ending nodes in a graph structure; and then, automatically generating a continuous and reasonable optimal geometric taxiway by executing a search algorithm combining path name matching and turning angle punishment in the graph. The application realizes automatic mapping from a text instruction to a standardized geometric path, overcomes the defects of the prior art, such as dependence on manual operation, low efficiency and error-prone, and significantly improves the intelligent level and operation safety of airport ground taxiway guidance.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of air transportation, and in particular to a method and system for automatically generating an aircraft taxiway path. BACKGROUND

[0002] In the field of air transportation, the efficiency and safety of airport ground operations are of paramount importance. After leaving the runway and before reaching the parking stand (or vice versa), an aircraft must follow a specified taxiway path according to the taxiway instructions issued by air traffic controllers. Currently, such taxiway instructions are typically issued in the form of a sequence of taxiway names in text, such as "via taxiway B, A to runway 18".

[0003] Under current technical conditions, there are mainly two ways to execute such instructions:

[0004] Manual interpretation and guidance: After receiving the text instructions, the ground guidance vehicle driver or the apron command personnel manually searches for the taxiways corresponding to the names on paper or electronic maps, relying on their memory and experience of the airport layout, and then constructs a rough taxiway path in their minds or through simple mapping. This method is highly dependent on individual experience and has the risk of path identification errors and low guidance efficiency, especially in complex airports, low visibility conditions, or high-intensity operating environments, which can easily lead to ground taxiing conflicts, entry into restricted areas, and other unsafe incidents.

[0005] Semi-automatic system based on GIS maps: Some advanced airport operation management systems or electronic flight bags (EFB) integrate GIS vector map data of the airport. After receiving the taxiway instructions, the system operator still needs to manually select the taxiway segments corresponding to the name sequence on the map interface, and the system then connects these segments to form a visual path. Although this scheme realizes the visualization of the path, it has not essentially realized the automatic conversion from the name to the path.

[0006] However, the above existing technologies have the following outstanding technical defects:

[0007] Low degree of automation, relying on manual operation: The existing system cannot automatically identify taxiway objects according to taxiway path names, and needs to manually select each one on the map, which is tedious and prone to errors, and is difficult to adapt to rapid changes or multi-aircraft coordination requirements.

[0008] Lack of automatic association mechanism between names and spatial data: Taxiway path names only exist as text information, and no mapping relationship that can be automatically recognized by a computer is established between the underlying GIS taxiway geometric objects and their topological relationships, which leads to the system's inability to understand the semantics of the instructions and perform automatic path calculation.

[0009] Poor path topology continuity and insufficient accuracy: the path generated by the existing scheme is usually a simple connection of discrete taxiway line segments, lacking automatic identification and checking of taxiway connection points, directions and geometric continuity, which may generate a path with geometric discontinuity or unreasonable turning, and cannot meet the requirements of aircraft taxiing for path smoothness.

[0010] Weak path data standardization and sharing capability: the generated geometric path is usually stored in the original coordinate string or custom vector format, lacking a unified spatial index structure, which makes it difficult to efficiently and accurately share, compare and reuse path data between different systems.

[0011] In summary, the existing technology has inherent defects such as low automation, poor path generation accuracy and difficult data sharing, and there is an urgent need for a technical solution that can automatically, accurately and efficiently convert text-form taxi instructions into standardized geometric taxi paths. SUMMARY

[0012] To solve the technical problems in the background art, the present application provides an automatic generation method and system for an aircraft taxi path.

[0013] The automatic generation method for an aircraft taxi path provided by the present application comprises the following steps:

[0014] S1, obtaining GIS map data of a target airport, the GIS map data including geometric coordinates and attribute information of taxiway line segments, gate points and runway points;

[0015] S2, spatially encoding the geometric coordinates of the taxiway line segments, gate points and runway points based on a preset H3 grid resolution, generating corresponding H3 grid cell identifiers, and constructing a taxiway hierarchical graph according to the H3 grid cell identifiers and line segment connection relationships;

[0016] S3, obtaining a taxi instruction data to be executed, the taxi instruction data including a path name sequence, a specified gate number, an arrival and departure identification, and latitude and longitude coordinates of a target point;

[0017] S4, integrating the gate number to the head or tail of the path name sequence according to the arrival and departure identification to form an updated path name sequence, and mapping the gate number and the latitude and longitude coordinates of the target point to a starting node and an ending node respectively based on the taxiway hierarchical graph;

[0018] S5. Starting from the start node and ending at the end node, performing a search in the taxiway hierarchical graph, matching the name of a candidate edge with the updated path name sequence during the search process, and determining a turn penalty based on the azimuth angle difference of the candidate edge relative to the previous edge, and comprehensively evaluating the turn penalty and the path length to guide the search, to generate an optimal geometric path composed of a sequence of latitude and longitude coordinates with taxiway properties;

[0019] S6. Outputting the geometric coordinate sequence of the optimal geometric path and its corresponding taxiway number information.

[0020] Preferably, the taxiway hierarchical graph contains nodes and edges, and the node attributes include H3 grid cell identifier, latitude and longitude, and the edge attributes include name, distance and azimuth angle; the geometric coordinates of the taxiway line segment, the stand point and the runway point are spatially encoded based on the preset H3 grid resolution to generate the corresponding H3 grid cell identifier, specifically including:

[0021] Traversing the geometric objects of the taxiway line segment, the stand point and the runway point, for the taxiway line segment, extracting each coordinate point in its geometric object; for the stand point and the runway point, extracting its own coordinate point;

[0022] According to the preset H3 grid resolution, the latitude and longitude of each coordinate point is mapped to a unique H3 grid cell identifier.

[0023] Preferably, the taxiway hierarchical graph is constructed according to the H3 grid cell identifier and the line segment connection relationship, specifically including:

[0024] Initializing an empty graph structure;

[0025] Traversing each H3 grid cell identifier obtained after spatial encoding, adding each unique H3 grid cell identifier as a node to the graph structure, and recording the corresponding latitude and longitude coordinates;

[0026] According to the original connection relationship of the taxiway line segment, creating edges connecting the nodes in the graph structure, wherein for two consecutive different H3 grid cell identifier nodes on a taxiway path, creating bidirectional edges from the previous node to the next node and from the next node to the previous node;

[0027] Assigning attributes to each created edge, the attributes including at least name, distance and azimuth angle calculated based on node coordinates.

[0028] Preferably, the stand number is integrated into the head or tail of the path name sequence according to the airport identifier to form an updated path name sequence, specifically including:

[0029] If the port identification is a departure port identification, the stand number is inserted into the head of the path name sequence; if the port identification is an arrival port identification, the stand number is inserted into the tail of the path name sequence, to form an updated path name sequence.

[0030] Preferably, the stand number and the latitude and longitude coordinates of the target point are respectively mapped to a start node and an end node based on the taxiway hierarchical graph, and specifically comprising:

[0031] The H3 grid cell identifier of the latitude and longitude coordinates of the target point is calculated, and if the H3 grid cell identifier exists in the taxiway hierarchical graph, it is determined as the node corresponding to the target point, otherwise the corresponding node is determined in the taxiway hierarchical graph through nearest neighbor search;

[0032] The mapping dictionary of the stand number and the node generated when the taxiway hierarchical graph is constructed is queried to obtain the node corresponding to the stand number;

[0033] If the port identification is a departure port identification, the node corresponding to the stand number is taken as the start node, and the node corresponding to the target point is taken as the end node; if the port identification is an arrival port identification, the node corresponding to the target point is taken as the start node, and the node corresponding to the stand number is taken as the end node.

[0034] Preferably, step S5 specifically comprises:

[0035] S51, constructing a priority queue, and adding the start node as an initial state to the priority queue, the state including the current node, the formed path sequence, the current path length, the turn penalty count, the name of the previous edge, the azimuth angle of the previous edge, and the index of the currently matched path name segment;

[0036] S52, taking out the state with the optimal evaluation value from the priority queue for expansion, wherein the evaluation value is compared in priority with the turn penalty count, and when the turn penalty count is the same, the current path length is compared;

[0037] S53, if the current node of the current expansion state is the end node, and the index of the currently matched path name segment indicates that the updated path name sequence has been completely matched, the path sequence under the current state is converted into the optimal geometric path;

[0038] S54, for each adjacent node and connecting edge of the current node of the current expansion state, a preset processing operation is performed;

[0039] S55, repeating steps S52 to S54 until the optimal geometric path is found or the priority queue is empty.

[0040] Preferably, the calculation formula of the azimuth angle difference value is specifically:

[0041] Azimuth angle difference value = min(abs(current edge azimuth angle - previous edge azimuth angle), 360-abs(current edge azimuth angle - previous edge azimuth angle)).

[0042] Preferably, step S54 specifically comprises:

[0043] S541, calculate the azimuth angle difference value of the current connecting edge and the previous edge, if the azimuth angle difference value exceeds the maximum allowable turning angle, skip the candidate edge, if the azimuth angle difference value exceeds the normal turning angle threshold but does not exceed the maximum allowable turning angle, add one to the turning penalty count of the current state;

[0044] S542, match the name of the current connecting edge with the updated path name sequence, and update the path name segment index according to the matching result: if the current target segment name is matched, the path name segment index is advanced by one bit, if the previous target segment name or the preset unnamed road segment identifier is matched, the path name segment index remains unchanged, otherwise, skip the candidate edge;

[0045] S543, update the state based on the current connecting edge, add the adjacent node to the path sequence, accumulate the path length, and record the new turning penalty count, the previous edge name and the previous edge azimuth angle;

[0046] S544, compare the updated new state with the existing recorded state, if the turning penalty count of the new state is smaller, or the turning penalty counts are the same but the path length is shorter, add the new state to the priority queue.

[0047] Preferably, the comparison logic of the optimal evaluation value specifically comprises:

[0048] Compare the turning penalty counts of the two states, and prefer the state with smaller turning penalty count;

[0049] If the turning penalty counts of the two states are the same, compare their path lengths, and prefer the state with shorter path length.

[0050] The present application provides an aircraft taxiway path automatic generation system, comprising:

[0051] A data acquisition module is configured to acquire GIS map data of a target airport, wherein the GIS map data comprises geometric coordinates and attribute information of taxiway segments, gate points and runway points;

[0052] A graph construction module is configured to perform spatial coding on the geometric coordinates of the taxiway segments, gate points and runway points based on a preset H3 grid resolution, generate corresponding H3 grid cell identifiers, and construct a taxiway level graph according to the H3 grid cell identifiers and the segment connection relationship;

[0053] The first analysis module is used for obtaining a piece of to-be-executed taxi instruction data, and the taxi instruction data includes a path name sequence, a specified stand number, an arrival / departure airport identifier, and latitude and longitude coordinates of a target point.

[0054] The second analysis module is further used for integrating the stand number to the head or tail of the path name sequence according to the arrival / departure airport identifier, forming an updated path name sequence, and mapping the stand number and the latitude and longitude coordinates of the target point into a start node and an end node respectively based on a taxiway hierarchical graph.

[0055] The third analysis module is further used for taking the start node as a starting point and the end node as a terminal point to perform a search in the taxiway hierarchical graph, matching a name of a candidate edge with the updated path name sequence in the search process, determining a turning penalty based on an azimuth angle difference of the candidate edge relative to a previous edge, comprehensively evaluating the turning penalty and a path length to guide the search, and generating an optimal geometric path composed of a latitude and longitude coordinate sequence with a taxiway attribute.

[0056] The path output module is used for outputting a geometric coordinate sequence of the optimal geometric path and corresponding taxiway number information.

[0057] In the present application, the proposed automatic generation method and system of an aircraft taxi path realize full automatic generation from a text instruction to a continuous and reasonable geometric path by establishing an automatic mapping relationship between a taxi path name and spatial geometric data, and completely get rid of the dependence on manual interpretation and operation. The method not only significantly improves the efficiency and accuracy of path generation, effectively reduces the operation risk caused by human errors, but also realizes standardized expression and cross-system efficient sharing of taxi path data by introducing a unified spatial index structure, and improves the intelligentization, automation and safety level of airport ground operation. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 A work flow schematic diagram of the automatic generation method of an aircraft taxi path proposed in the present application;

[0059] Figure 2 A system architecture schematic diagram of the automatic generation system of an aircraft taxi path proposed in the present application. DETAILED DESCRIPTION

[0060] REFERENCE Figure 1 and Figure 2 The automatic generation method of an aircraft taxi path proposed in the present application includes the following steps:

[0061] S1, GIS map data of a target airport is obtained, and the GIS map data includes geometric coordinates and attribute information of taxiway line segments, stand points and runway points.

[0062] Specifically, by calling the GIS apron map service of a specific airport and inputting the airport three-letter code as a request parameter, the apron map data of the corresponding airport can be obtained, and the returned data set includes the spatial geometric information and attribute information of key objects such as taxiways, stands, and runways. Among them, the taxiway data contains taxiway number, taxiway line segment node coordinates (latitude and longitude), and line segment connection relationship; the stand data contains stand number, stand head and tail point latitude and longitude, and parking direction; the runway data contains runway number, runway axis segment latitude and longitude, and start and end point identifier.

[0063] S2, based on the preset H3 grid resolution, the geometric coordinates of the taxiway line segment, the stand point and the runway point are spatially encoded to generate the corresponding H3 grid cell identifier, and an apron hierarchical graph is constructed according to the H3 grid cell identifier and the line segment connection relationship.

[0064] In this embodiment, the apron hierarchical graph contains nodes and edges, the node attributes include H3 grid cell identifier, latitude and longitude, and the edge attributes include name, distance and azimuth; based on the preset H3 grid resolution, the geometric coordinates of the taxiway line segment, the stand point and the runway point are spatially encoded to generate the corresponding H3 grid cell identifier, which specifically includes:

[0065] Traverse the geometric objects of the taxiway line segment, the stand point and the runway point, for the taxiway line segment, extract each coordinate point in its geometric object; for the stand point and the runway point, extract its own coordinate point;

[0066] According to the preset H3 grid resolution, the latitude and longitude of each coordinate point is mapped to a unique H3 grid cell identifier.

[0067] In this embodiment, an apron hierarchical graph is constructed according to the H3 grid cell identifier and the line segment connection relationship, which specifically includes:

[0068] Initialize an empty graph structure;

[0069] Traverse each H3 grid cell identifier obtained after spatial encoding, add each unique H3 grid cell identifier as a node to the graph structure, and record the corresponding latitude and longitude coordinates;

[0070] According to the original connection relationship of the taxiway line segment, the edges connecting the nodes are created in the graph structure, wherein for the two different H3 grid cell identifier nodes on a taxiway path, a bidirectional edge is created from the previous node to the next node and from the next node to the previous node.

[0071] Assign attributes to each edge created, and the attributes at least include name, distance calculated based on node coordinates, and azimuth.

[0072] Specifically, the operation process of constructing the taxiway level graph is as follows:

[0073] (1) Initialize the graph structure Graph;

[0074] (2) Traverse each record in the read-in taxiway geographic data set, extract the MultiLineString type of the geometric object from each record. For each MultiLineString object, traverse each single line string in turn, and for each single line string, traverse each point in the order of the point sequence and obtain the longitude and latitude coordinates (longitude lon, latitude lat) of each point;

[0075] (3) For the current traversed point, according to the longitude and latitude coordinates and the preset H3 resolution, the point is mapped to a unique H3 grid cell identifier, and the corresponding grid cell identifier (H3 ID) is obtained. Add the point to the graph structure Graph, and set the attributes of the node as follows:

[0076] 1) Node identifier: H3 ID;

[0077] 2) Latitude: the latitude of the current point;

[0078] 3) Longitude: the longitude of the current point;

[0079] (4) When the H3 grid cell identifier of the current point is obtained, it is judged whether the previous node identifier exists and is different from the current H3 identifier:

[0080] 1) If the condition is met, create a directed edge from the previous node to the current node in the graph structure Graph;

[0081] 2) At the same time, create a reverse directed edge from the current node to the previous node to realize bidirectional communication.

[0082] S3, obtain a to-be-executed taxi instruction data, the taxi instruction data including a path name sequence, a specified stand number, an arrival and departure airport identifier, and longitude and latitude coordinates of a target point.

[0083] S4, integrate the stand number into the head or tail of the path name sequence according to the arrival and departure airport identifier, form an updated path name sequence, and map the stand number and the longitude and latitude coordinates of the target point to a starting node and an ending node respectively based on the taxiway level graph.

[0084] In this embodiment, the stand number is integrated into the head or tail of the path name sequence according to the arrival and departure airport identifier to form an updated path name sequence, which specifically includes:

[0085] If the arrival and departure airport identifier is a departure airport identifier, the gate number is inserted into the head of the path name sequence; if it is an arrival airport identifier, the gate number is inserted into the tail of the path name sequence, forming an updated path name sequence.

[0086] In this embodiment, based on the taxiway hierarchical graph, the gate number and the latitude and longitude coordinates of the target point are respectively mapped to the start node and the end node, specifically including:

[0087] The H3 grid cell identifier of the latitude and longitude coordinates of the target point is calculated, and if the H3 grid cell identifier exists in the taxiway hierarchical graph, it is determined as the node corresponding to the target point, otherwise the corresponding node is determined in the taxiway hierarchical graph through nearest neighbor search;

[0088] The mapping dictionary of the gate number and the node generated when the taxiway hierarchical graph is constructed is queried to obtain the node corresponding to the gate number;

[0089] If the arrival and departure airport identifier is a departure airport identifier, the node corresponding to the gate number is taken as the start node, and the node corresponding to the target point is taken as the end node; if the arrival and departure airport identifier is an arrival airport identifier, the node corresponding to the target point is taken as the start node, and the node corresponding to the gate number is taken as the end node.

[0090] S5, taking the start node as the starting point and the end node as the terminal point, performing search in the taxiway hierarchical graph, matching the name of the candidate edge with the updated path name sequence in the search process, and determining the turning penalty based on the azimuth angle difference of the candidate edge relative to the previous edge, and comprehensively evaluating the turning penalty and the path length to guide the search, generating an optimal geometric path composed of latitude and longitude coordinate sequences with taxiway properties.

[0091] In this embodiment, step S5 specifically includes:

[0092] S51, constructing a priority queue, and adding the start node as the initial state to the priority queue, the state including the current node, the path sequence formed, the current path length, the turning penalty count, the name of the previous edge, the azimuth angle of the previous edge, and the index of the path name segment matched currently;

[0093] S52, taking out the state with the optimal evaluation value from the priority queue for expansion, wherein the evaluation value is compared in priority with the turning penalty count, and when the turning penalty count is the same, the current path length is compared;

[0094] S53, if the current node of the current expansion state is the end node, and the index of the path name segment matched currently indicates that the updated path name sequence has been completely matched, the path sequence under the current state is converted into the optimal geometric path;

[0095] S54, for each adjacent node and connecting edge of the current node of the current extended state, performing a preset processing operation;

[0096] Specifically, step S54 specifically comprises:

[0097] S541, calculating the azimuth angle difference value of the current connecting edge and the previous edge, if the azimuth angle difference value exceeds the maximum allowed turning angle, skipping the candidate edge, if the azimuth angle difference value exceeds the regular turning angle threshold but does not exceed the maximum allowed turning angle, adding one to the turning penalty count of the current state;

[0098] S542, matching the name of the current connecting edge with the updated path name sequence, and updating the path name segment index according to the matching result: if the current target segment name is matched, the path name segment index is pushed by one bit, if the previous target segment name or the preset unnamed road segment identifier is matched, the path name segment index remains unchanged, otherwise, the candidate edge is skipped;

[0099] S543, updating the state based on the current connecting edge, adding the adjacent node to the path sequence, accumulating the path length, and recording the new turning penalty count, the previous edge name and the previous edge azimuth angle;

[0100] S544, comparing the updated new state with the existing recorded state, if the turning penalty count of the new state is smaller, or the turning penalty counts are the same but the path length is shorter, the new state is added to the priority queue.

[0101] S55, repeating steps S52 to S54 until the optimal geometric path is found or the priority queue is empty.

[0102] Specifically, the calculation formula of the azimuth angle difference value is specifically:

[0103] Azimuth angle difference value = min(abs(current edge azimuth angle - previous edge azimuth angle), 360-abs(current edge azimuth angle - previous edge azimuth angle)).

[0104] It should be noted that the comparison logic of the optimal evaluation value is specifically:

[0105] Comparing the turning penalty counts of two states, and preferentially selecting the state with smaller turning penalty count;

[0106] If the turning penalty counts of two states are the same, comparing the path lengths, and preferentially selecting the state with shorter path length.

[0107] It should be noted that in the GIS modeling of the airport taxiway network, due to factors such as the accuracy of the base map and the line segment splitting method, some adjacent line segments have a large turning angle in geometry, which seems to not meet the aircraft turning limit, but the corresponding actual taxiway is still feasible. At the same time, there are a large number of long-distance road segments with the same name in the taxiway topology. If only a simple matching is performed according to the order of the path name, it is easy to turn at the wrong node, resulting in a generated geometric path that does not match the actual taxiway, or even a path containing unreasonable sharp turns, affecting the safety of aircraft operation.

[0108] Therefore, the embodiment introduces azimuth angle calculation and turning penalty mechanism on the basis of geometric path search based on path name sequence, combines path name constraint and turning angle control, and sorts and filters the candidate paths through a comprehensive evaluation function. On the one hand, it ensures that the turning only occurs at the expected position on the long-distance taxiway with the same name; on the other hand, under the premise of meeting the path name requirement, the geometric path with fewer large-angle turning times and shorter overall path is preferentially selected.

[0109] Specifically, the input data for geometric path generation is preprocessed, and the input data includes a path name sequence path_names, a stand number stand, an in-or-out identification in_or_out, and a latitude and longitude coordinate point target_point of a target position. The specific processing steps are as follows:

[0110] 1) path name sequence path_names list update

[0111] If in_or_out=D, it is an outport path generation, and the stand number stand is inserted into the head of the path name sequence path_names; if in_or_out=A, it is an inport path generation, and the stand number stand is inserted into the tail of the list path_names.

[0112] 2) Calculate the start node start_node and the end node end_node

[0113] First, the latitude and longitude information of the latitude and longitude coordinate point target_point of the target position is extracted, which is mapped into H3ID, and it is judged whether the H3ID is in the graph structure Graph. If it is, the H3ID is the node target_node corresponding to the target point; if not, the KD tree structure is constructed to find the node closest to the target point from the graph structure Graph, and the node is the node target_node corresponding to the target point.

[0114] Then, the stand number stand is input into the mapping dictionary between stand number and node generated in the stand layer hierarchical grid construction process, to obtain the node data stand_node of the stand.

[0115] Finally, the starting node start_node and the ending node end_node are confirmed by judging the in-or-out identifier in_or_out, as follows:

[0116] If in_or_out=D, start_node=stand_node and end_node=target_node.

[0117] If in_or_out=A, start_node=target_node and end_node=stand_node.

[0118] In the embodiment, the process of generating the corresponding geometric path according to the input path name sequence and the pre-determined starting node and ending node includes the following steps:

[0119] (1) Initialization of state and priority queue

[0120] 1) On the premise that the starting node and the ending node of the determined path search are determined, the starting node start_node is set as the initial current node, and the path search state is initialized accordingly;

[0121] 2) A minimum heap priority queue is constructed, and the queue elements include: turn penalty count penalty_count, current path length path_len, current node current_node, formed path sequence path, previous edge name prev_edge_name, previous edge bearing prev_bearing, and the current matched path name segment index seg_idx.

[0122] The above queue elements are initialized as follows:

[0123] current_node=start_node, path_len=0;

[0124] penalty_count=0, prev_edge_name=None, prev_bearing=None, seg_idx=0;

[0125] 3) Establish the state record structure best_state to record the minimum number of large turns and the shortest path length corresponding to each state when it is reached, in order to prune and optimize subsequent searches, and the initial state is float('inf').

[0126] (2) Search process and turn penalty strategy

[0127] 1) Take out the state with the minimum current evaluation value (turn penalty count penalty_count first, and path length path_len second) from the priority queue as the current expansion state;

[0128] 2) If the current node is the end node, i.e. current_node = end_node, then convert the path sequence under the current state into a geometric path, which is calculated as follows:

[0129] Traverse the triplets node1, edge_name, node2 in the path sequence path that has been formed, obtain the edge attribute data edge_data, the latitude and longitude data start_coords of node1, and the latitude and longitude data end_coords of node2 from the graph structure Graph, and add [start_coords, end_coords] to the line segment list. Extract taxiway_id from edge_data and add it to the list for storing taxiway IDs. Finally, convert the line segment list into a multiline format that can be recognized by a GIS map.

[0130] Finally, output the corresponding path ID information and geometric representation, and terminate the search at this time;

[0131] 3) Construct the state identifier:

[0132] state_key = (current_node, seg_idx, prev_edge_name);

[0133] Extract the current evaluation value: current_state_value = (penalty_count, path_len). If best_state[state_key] ≤ current_state_value, proceed to the next loop, otherwise update best_state[state_key] = current_state_value, and continue with step 4);

[0134] 4) For the current node, enumerate all its neighboring nodes and for each candidate edge neighbor between the current node and the neighboring node, do the following:

[0135] Get the name of the candidate edge, current_edge_name and calculate the bearing between the current node and the candidate edge, current_bearing, and the bearing of the edge, current_bearing;

[0136] If there is a previous edge bearing, prev_bearing, then calculate the turn angle, turn_angle, as follows:

[0137] turn_angle = min(abs(current_bearing - prev_bearing),

[0138] 360 - abs(current_bearing - prev_bearing));

[0139] If turn_angle > max_turn_angle (max_turn_angle = 150) then skip this candidate edge and do not expand it further. Otherwise, if turn_angle > 60, then add 1 to the turn penalty count for the current state, i.e. penalty_count = penalty_count + 1. If turn_angle < 60, then do not increase the penalty.

[0140] (3) Path name sequence matching and segment index updating

[0141] 1) For each candidate edge, match it against the input path name sequence based on its name:

[0142] If the candidate edge name matches the current target path name (the name indicated by the path name segment index), i.e. current_edge_name == path_names[seg_idx], then consider that the edge completes the matching of the current target segment and increment the path name segment index, seg_idx = seg_idx + 1.

[0143] If the candidate edge name matches the previous target path name, i.e. current_edge_name == path_names[seg_idx - 1], then consider it as a continuous extension on the same name segment and keep the path name segment index unchanged.

[0144] If the candidate edge name is empty or marked as "unnamed" (indicating that the segment is a turning arc), i.e., current_edge_name == "unnamed", the edge is allowed as a transition segment, and the path name segment index is kept unchanged.

[0145] If the candidate edge name does not match the current target segment name, nor the previous target segment name, and does not belong to the allowed unnamed segments, the candidate edge is discarded.

[0146] With the above rules, in the presence of large segments of the same name, the system can still continuously travel along the same named segments, and only switch names and turn at the expected node locations, provided that the turning angle constraints are met, thus ensuring that the path turns at the correct locations.

[0147] 2) Update the current path, path length, state key, and state value as follows:

[0148] Update the current path:

[0149] new_path = path + [(current_node, current_edge_name, neighbor)];

[0150] Update the path length:

[0151] new_path_len = path_len + 1;

[0152] Update the state key:

[0153] new_state_key = (neighbor, seg_idx, current_edge_name);

[0154] Update the state value:

[0155] new_state_value = (penalty_count, new_path_len);

[0156] If new_state_value < best_state[new_state_key], i.e., if the new state has not been visited or has a better path (fewer turns or the same number of turns but a shorter path), push the new state into the priority queue for further search, otherwise prune the state and do not expand it. Repeat this process by taking the state with the smallest evaluation value from the priority queue for expansion until the priority queue is empty or a terminal path that meets the conditions is found.

[0157] S6, output the geometric coordinate sequence of the optimal geometric path and the corresponding taxiway number information.

[0158] With reference to Figure 1 and Figure 2 The application provides an automatic generation system for an aircraft taxiway path, comprising:

[0159] A data acquisition module is configured to acquire GIS map data of a target airport, wherein the GIS map data comprises geometric coordinates and attribute information of taxiway line segments, gate points and runway points.

[0160] A graph construction module is configured to perform spatial coding on the geometric coordinates of the taxiway line segments, gate points and runway points based on a preset H3 grid resolution, to generate corresponding H3 grid cell identifiers, and to construct a taxiway level graph according to the H3 grid cell identifiers and line segment connection relationships.

[0161] A first analysis module is configured to acquire a piece of to-be-executed taxi instruction data, wherein the taxi instruction data comprises a path name sequence, a specified gate number, an identification of an arrival / departure airport and latitude / longitude coordinates of a target point.

[0162] A second analysis module is further configured to integrate the gate number into a head or tail of the path name sequence according to the identification of the arrival / departure airport, to form an updated path name sequence, and to map the gate number and the latitude / longitude coordinates of the target point into a starting node and an ending node respectively based on the taxiway level graph.

[0163] A third analysis module is further configured to perform a search in the taxiway level graph with the starting node as a starting point and the ending node as a terminal point, to match a name of a candidate edge with the updated path name sequence in the search process, and to determine a turning penalty based on an azimuth angle difference of the candidate edge relative to a previous edge, to comprehensively evaluate the turning penalty and a path length to guide the search, and to generate an optimal geometric path composed of a sequence of latitude / longitude coordinates with taxiway attributes.

[0164] A path output module is configured to output the geometric coordinate sequence of the optimal geometric path and the corresponding taxiway number information.

[0165] The above merely describes a preferred specific embodiment of the application, but the protection scope of the application is not limited thereto, and any person skilled in the art can make equivalent replacements or changes to the technical solution and the inventive concept of the application within the technical range disclosed by the application, which should be covered within the protection scope of the application.

Claims

1. A method for automatic generation of an aircraft taxiway path, characterized in that, The method comprises the following steps: S1, acquiring GIS map data of a target airport, the GIS map data comprising geometric coordinates and attribute information of taxiway line segments, stand points and runway points; S2, spatially encoding the geometric coordinates of the taxiway line segments, stand points and runway points based on a preset H3 grid resolution, generating corresponding H3 grid cell identifiers, and constructing a taxiway level graph according to the H3 grid cell identifiers and line segment connection relationships; S3, acquiring a taxi instruction data to be executed, the taxi instruction data comprising a path name sequence, a specified stand number, an arrival and departure airport identifier, and latitude and longitude coordinates of a target point; S4, integrating the stand number into the head or tail of the path name sequence according to the arrival and departure airport identifier, forming an updated path name sequence, and mapping the stand number and the latitude and longitude coordinates of the target point into a start node and an end node respectively based on the taxiway level graph; S5, taking the start node as a starting point and the end node as an ending point, performing a search in the taxiway level graph, matching a name of a candidate edge with the updated path name sequence in the search process, determining a turning penalty based on an azimuth angle difference of the candidate edge relative to a previous edge, and comprehensively evaluating the turning penalty and a path length to guide the search, to generate an optimal geometric path composed of a sequence of latitude and longitude coordinates with taxiway properties; S6, outputting a geometric coordinate sequence of the optimal geometric path and corresponding taxiway number information thereof; Step S5 specifically comprises: S51, constructing a priority queue, adding the start node as an initial state into the priority queue, and the state comprising a current node, a formed path sequence, a current path length, a turning penalty count, a name of a previous edge, an azimuth angle of the previous edge, and a currently matched path name segment index; S52, taking out a state with an optimal evaluation value from the priority queue for expansion, wherein the evaluation value is compared in priority with the turning penalty count, and the current path length is compared when the turning penalty counts are the same; S53, if the current node of the current expansion state is the end node, and the currently matched path name segment index indicates that the updated path name sequence has been completely matched, converting a path sequence in the current state into the optimal geometric path; S54, performing a preset processing operation on each adjacent node and connecting edge of the current node of the current expansion state; S55, repeating steps S52 to S54 until the optimal geometric path is found or the priority queue is empty; Step S54 specifically comprises: S541, calculating an azimuth angle difference of the current connecting edge and the previous edge, skipping the candidate edge if the azimuth angle difference exceeds a maximum allowed turning angle, and adding one to the turning penalty count of the current state if the azimuth angle difference exceeds a regular turning angle threshold but does not exceed the maximum allowed turning angle. S542, match the name of the current connection edge with the updated path name sequence, and update the path name segment index according to the matching result: if the current target segment name is matched, the path name segment index is advanced by one bit, if the previous target segment name or the preset unnamed road segment identifier is matched, the path name segment index remains unchanged, otherwise the candidate edge is skipped; S543, based on the current connection edge update state, add the adjacent node to the path sequence, accumulate the path length, and record the new turn penalty count, the previous edge name and the previous edge azimuth angle; S544, compare the updated new state with the existing recorded state, if the new state has smaller turn penalty count, or the turn penalty count is the same but the path length is shorter, the new state is added to the priority queue.

2. The method of claim 1, wherein, The taxiway hierarchical graph includes nodes and edges, and the node attributes include H3 grid cell identifier, latitude and longitude, and the edge attributes include name, distance and azimuth angle; the geometric coordinates of the taxiway line segment, the stand point and the runway point are spatially encoded based on the preset H3 grid resolution, and the corresponding H3 grid cell identifier is generated, which specifically includes: Traverse the geometric objects of the taxiway line segment, the stand point and the runway point, for the taxiway line segment, extract each coordinate point in the geometric object; for the stand point and the runway point, extract the coordinate point itself; According to the preset H3 grid resolution, the latitude and longitude of each coordinate point is mapped to a unique H3 grid cell identifier.

3. The method of claim 2, wherein, The taxiway hierarchical graph is constructed according to the H3 grid cell identifier and the line segment connection relationship, specifically including: Initialize an empty graph structure; Traverse each H3 grid cell identifier obtained after spatial encoding, add each unique H3 grid cell identifier as a node to the graph structure, and record the corresponding latitude and longitude coordinates; According to the original connection relationship of the taxiway line segment, edges connecting the nodes are created in the graph structure, wherein for two consecutive different H3 grid cell identifier nodes on a taxiway path, bidirectional edges are created from the previous node to the next node and from the next node to the previous node; Each edge created is assigned attributes, and the attributes at least include name, distance and azimuth angle calculated based on node coordinates.

4. The method of claim 1, wherein, The airport identifier is inserted into the head or tail of the path name sequence to form an updated path name sequence, specifically including: If the airport identifier is a departure identifier, the stand number is inserted into the head of the path name sequence; if it is an arrival identifier, the stand number is inserted into the tail of the path name sequence to form an updated path name sequence.

5. The method of claim 1, wherein, The stand number and the latitude and longitude coordinates of the target point are respectively mapped to the starting node and the ending node based on the taxiway hierarchical graph, specifically including: Calculate the H3 grid cell identifier of the latitude and longitude coordinates of the target point, if the H3 grid cell identifier exists in the taxiway hierarchical graph, it is determined as the node corresponding to the target point, otherwise the corresponding node is determined in the taxiway hierarchical graph through nearest neighbor search; The mapping dictionary of the gate number and the node generated when the taxiway level graph is constructed is queried to obtain the node corresponding to the gate number; If the airport identification is the departure identification, the node corresponding to the gate number is taken as the starting node and the node corresponding to the target point is taken as the ending node; if the airport identification is the arrival identification, the node corresponding to the target point is taken as the starting node and the node corresponding to the gate number is taken as the ending node.

6. The method of claim 1, wherein, The calculation formula of the azimuth difference value is specifically: Azimuth difference value = min(abs(current edge azimuth - previous edge azimuth), 360-abs(current edge azimuth - previous edge azimuth)).

7. The method of claim 1, wherein, The comparison logic of the optimal evaluation value is specifically: The turning penalty counts of two states are compared, and the state with the smaller turning penalty count is preferentially selected; If the turning penalty counts of two states are the same, the path lengths thereof are compared, and the state with the shorter path length is preferentially selected.

8. An automatic generation system of an aircraft taxiway path, characterized by, It comprises: A data acquisition module is configured to acquire GIS map data of a target airport, wherein the GIS map data comprises geometric coordinates and attribute information of taxiway line segments, gate points and runway points; A graph construction module is configured to perform spatial coding on the geometric coordinates of the taxiway line segments, gate points and runway points based on a preset H3 grid resolution, to generate corresponding H3 grid cell identifications, and to construct a taxiway level graph according to the H3 grid cell identifications and line segment connection relationships; A first analysis module is configured to acquire a piece of to-be-executed taxi instruction data, wherein the taxi instruction data comprises a path name sequence, a specified gate number, an airport identification and latitude and longitude coordinates of a target point; A second analysis module is further configured to integrate the gate number to the head or tail of the path name sequence according to the airport identification, to form an updated path name sequence, and to map the gate number and the latitude and longitude coordinates of the target point to a starting node and an ending node respectively based on the taxiway level graph; A third analysis module is further configured to perform search in the taxiway level graph with the starting node as the starting point and the ending node as the terminal point, to match the name of a candidate edge with the updated path name sequence in the search process, to determine a turning penalty based on an azimuth difference value of the candidate edge relative to a previous edge, to comprehensively evaluate the turning penalty and a path length to guide the search, and to generate an optimal geometric path composed of a sequence of latitude and longitude coordinates with taxiway properties. A path output module is configured to output a sequence of geometric coordinates of the optimal geometric path and corresponding taxiway number information thereof. The third analysis module specifically comprises: A priority queue is constructed, and the starting node is added to the priority queue as an initial state, wherein a state comprises a current node, a formed path sequence, a current path length, a turning penalty count, a name of a previous edge, an azimuth of the previous edge and an index of a matched path name segment; An optimal state in terms of evaluation value is taken out from the priority queue for expansion, wherein the evaluation value is preferentially compared in terms of the turning penalty count, and the current path length is compared when the turning penalty counts are the same. If the current node of the current expansion state is an end node, and the current matched path name segment index indicates that the updated path name sequence has been completely matched, the path sequence in the current state is converted into an optimal geometric path; For each adjacent node and connecting edge of the current node of the current expansion state, a preset processing operation is performed; The above process is repeated until an optimal geometric path is found or the priority queue is empty; For each adjacent node and connecting edge of the current node of the current expansion state, a preset processing operation is performed, specifically including: The azimuth angle difference between the current connecting edge and the previous edge is calculated, if the azimuth angle difference exceeds the maximum allowable turning angle, the candidate edge is skipped, if the azimuth angle difference exceeds the regular turning angle threshold but does not exceed the maximum allowable turning angle, the turning penalty count of the current state is increased by one; The name of the current connecting edge is matched with the updated path name sequence, and the path name segment index is updated according to the matching result: if the current target segment name is matched, the path name segment index is pushed forward by one bit, if the previous target segment name or the preset unnamed road segment identifier is matched, the path name segment index remains unchanged, otherwise, the candidate edge is skipped; Based on the current connecting edge, the adjacent node is added to the path sequence, the path length is accumulated, and the new turning penalty count, the previous edge name and the previous edge azimuth angle are recorded; The updated new state is compared with the existing recorded state, if the turning penalty count of the new state is smaller, or the turning penalty counts are the same but the path length is shorter, the new state is added to the priority queue.

Images