Strategy recommendation method based on traffic situation map classification

Abstract

The invention provides a strategy recommendation method based on traffic situation map classification. The method comprises the following steps: creating a historical strategy information database; selecting a control area as a concerned area, acquiring traffic information in the concerned area, and constructing a traffic situation map; published strategies of the concerned area are obtained from the historical strategy information database, historical traffic situation maps are constructed, and the published strategies corresponding to the historical traffic situation maps are recorded; selecting a recommendation strategy, regarding each historical traffic situation map as a map category, carrying out map classification on the traffic situation maps to find the historical traffic situation maps of the same category as the traffic situation maps, and selecting a published strategy corresponding to the historical traffic situation map as the recommendation strategy; the action time period of the recommendation strategy is corrected and pushed to the user for use; and updating the historical strategy information database.

Description

Technical Field

[0001] The present invention relates to an air traffic flow management strategy recommendation method, in particular to a strategy recommendation method based on traffic situation graph classification. Background Art

[0002] This section merely provides background information related to the present disclosure and is not necessarily prior art.

[0003] The goal of air traffic flow management (ATFM) is to balance air traffic supply and demand by developing strategies. More precisely, ATFM involves traffic management personnel implementing effective scheduling strategies when air traffic approaches or reaches the available capacity of a control unit, optimizing the spatiotemporal distribution of flight traffic to ensure smooth and orderly air traffic.

[0004] As demand for air transport continues to grow, the research, development, and construction of traffic management systems have made tremendous progress. Unfortunately, currently widely used traffic management systems lack the ability to automatically recommend strategies. Strategy formulation is entirely the responsibility of traffic management personnel, who combine their own work experience with the real-time information provided by the traffic management system and forecasts for the future. This results in a high workload, high subjectivity, and low coordination. Saving past successful strategies in some way and automatically selecting and recommending them to traffic management personnel for use or reference at the appropriate time has become a pressing need for traffic management personnel.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the Invention

[0006] Purpose of the invention: The technical problem to be solved by the present invention is to provide a strategy recommendation method based on traffic situation map classification in response to the shortcomings of the existing technology.

[0007] In order to solve the above technical problems, the present invention discloses a strategy recommendation method based on traffic situation map classification, comprising the following steps:

[0008] Step 1: Create a historical policy information database to store the traffic information within the control area of ​​the issued policy and the unit that issued the policy when the policy was issued;

[0009] Step 2: Select a controlled area as the focus area, obtain traffic information within the focus area, and construct a traffic situation map;

[0010] Step 3: Obtain the published policies for the concerned area from the historical policy information database. For each published policy, obtain the traffic information of the concerned area when the policy was published from the historical policy information database, construct a historical traffic situation map, and record the published policies corresponding to each historical traffic situation map.

[0011] Step 4: Select a recommended strategy. Consider each historical traffic situation map as a map category. Classify the traffic situation map to find a historical traffic situation map of the same category as the traffic situation map. Select the published strategy corresponding to the historical traffic situation map as the recommended strategy.

[0012] Step 5: Modify the effective period of the recommendation strategy and push it to the user;

[0013] Step 6: Update the historical strategy information database to record the user's use of the recommended strategy. If the user does not use the recommended strategy but adopts a new strategy, the new strategy and the traffic information of the focus area when the new strategy is released will be stored in the historical strategy information database.

[0014] Furthermore, the traffic information described in step 1 includes: route information, airport predicted capacity information, and waypoint predicted capacity information.

[0015] Furthermore, the route information in step 1 includes temporary route information.

[0016] Furthermore, the historical policy information database described in step 1 includes:

[0017] Historical strategy information data table, historical route information data table and historical forecast information data table, among which,

[0018] The historical policy information data table stores published policy information, including: policy identifier, policy name, number of recommendations, policy effective period, policy release area, policy release time, and policy content;

[0019] The historical route information data table stores route information of the control area of ​​the unit that issues the policy when the policy is issued, including: policy identifier, policy issuance area, route information and temporary route information;

[0020] The historical prediction information data table stores the predicted traffic and capacity information of each airport and each waypoint in the control area of ​​the unit that issues the policy in the next k hours.

[0021] Furthermore, the historical forecast information data table in step 1 includes: a policy identifier, an object type, an object name, a policy release area, predicted traffic for the next k hours, and predicted capacity for the next k hours; wherein the object is an airport or a waypoint.

[0022] Furthermore, the construction of the traffic situation map described in step 2 includes:

[0023] Step 2-1: Create a node set V, as follows:

[0024] The airports and waypoints in the area of ​​interest are sequentially placed into the node set V. Assuming the total number of airports and waypoints in the area of ​​interest is N, the node set is represented as follows:

[0025] V={v1,v2,v3,…v N}

[0026] Among them, v N Indicates the Nth node;

[0027] Step 2-2: Create edge set E, as follows:

[0028] Parse the route segment from the route information and temporary route information in the traffic information within the area of ​​interest. If there is a route segment specified by v i to v j The route segment is <v i , v j >Put it into the edge set E, where v i and v j Represents the i-th node and the j-th node in the node set V;

[0029] Step 2-3: Create an adjacency matrix A with a dimension of N×N, as follows:

[0030] Traverse the edge set E and set each element in the adjacency matrix A. The rules are as follows:

[0031]

[0032] Among them, A i，j Represents the element in row i and column j in the adjacency matrix A;

[0033] Step 2-4: Calculate the ratio of the predicted flow and predicted capacity of each node in the node set V in the next k hours, and use the obtained k-dimensional vector as the attribute vector of the node; let the predicted flow and predicted capacity of the i-th node in the t-th hour be f i，t and c i，t , where t = 1, 2, ..., k, then the attribute vector x of the i-th node i Expressed as:

[0034]

[0035] The node attribute matrix X composed of all nodes in the node set V is:

[0036] X=[x1 x2 … x N ]

[0037] Steps 2-5: Use the airports and waypoints in the area of ​​interest as nodes, the route segments consisting of routes and temporary routes as edges, and the ratio of node predicted traffic flow and node predicted capacity in the next k hours as node attribute vectors to construct the traffic situation graph G. pre , which is expressed as follows:

[0038] G pre =(V, E, A, X).

[0040] Furthermore, the steps of constructing a historical traffic situation map and recording the published policies corresponding to each historical traffic situation map described in step 3 include:

[0041] Step 3-1, obtain the published policies released in the focus area from the historical policy information data table, and set the number of published policies to M;

[0042] Step 3-2: Select the mth published policy whose policy identifier is y m ;

[0043] Step 3-3, through the policy identifier y m Obtain route information and temporary route information of the concerned area from the historical route information data table;

[0044] Step 3-4, through the policy identifier y m Obtain the predicted traffic and predicted capacity of all nodes for k hours when the strategy is released from the historical prediction information data table;

[0045] Step 3-5: Use the airports and waypoints in the focus area in the historical strategy information database as nodes, the route segments consisting of the routes and temporary routes in the focus area in the historical strategy information database as edges, and the ratio of the predicted flow and predicted capacity of the node k hours in the historical strategy information database as node attributes. Use the same method as described in step 2 to construct the traffic situation map to construct the historical traffic situation map G. m =(V m , E m , A m , X m );

[0046] Steps 3-6, record the historical traffic situation map G m With the policy identifier y m The corresponding relationship;

[0047] Step 3-7: Repeat steps 3-2 to 3-6 for all published strategies obtained in step 3-1 to obtain M historical traffic situation graphs G1, G2, G3, ..., G Mand their corresponding policy identifiers y1, y2, y3, …, yM.

[0048] Furthermore, the recommended strategy described in step 4 includes:

[0049] Step 4-1: treat the M historical traffic situation maps as M types of traffic situation map categories, use their corresponding strategy identifiers as category labels, and classify the traffic situation map G according to the M categories of the historical traffic situation map. pre Perform classification and obtain the classified category labels;

[0050] In step 4-2, the policy content of the recommended policy is obtained from the historical policy information data table through the category label, that is, the policy identifier.

[0051] Furthermore, the effective period of the revised recommendation strategy described in step 5 includes:

[0052] Step 5-1, let the release time of the recommended strategy be T pub , the strategy start time is T start , the strategy end time is T end , the current time is T now , for the strategy release time T pub and the current time T now Round up by hour, that is, only keep the hour information, and get the release time T′ of the rounding strategy pub and round the current time T′ now ;

[0053] Step 5-2, calculate the revised strategy start time T′ start and the revised strategy end time T′ end , as follows:

[0054]

[0055] Use the revised strategy to start time T′ start and the revised strategy end time T′ end Update the strategy start time T for replacing the recommended strategy start , strategy end time T end .

[0056] Furthermore, the update history policy information database described in step 6 includes:

[0057] Step 6-1: If the recommended strategy is used directly or modified by the user, the number of recommended times of the published strategy in the historical strategy information data table corresponding to the recommended strategy is increased by 1, and step 6-3 is executed; otherwise, step 6-2 is executed;

[0058] Step 6-2: The new strategy adopted by the user is stored in the historical strategy information data table and the number of recommendations is set to 0. The predicted traffic and predicted capacity information for each airport and waypoint in the current focus area for the next k hours is stored in the historical prediction information data table. The route information and temporary route information of the focus area are stored in the historical route information data table.

[0059] Step 6-3, traverse all published strategies in the historical strategy information data table, and regard published strategies whose recommendation times have been 0 within a preset period since the strategy was published as accidental strategies; delete the relevant information of accidental strategies in the historical strategy information data table, historical prediction information data table, and historical route information data table.

[0060] Beneficial effects:

[0061] (1) The method of the present invention fills the gap in the research of strategy recommendation function technology in the field of air traffic flow management, and provides technical support for technical breakthroughs in the development process of air traffic flow management system.

[0062] (2) The method of the present invention helps to reduce the workload of traffic management personnel and improve traffic management efficiency.

[0063] (3) The method of the present invention helps to reduce the subjectivity of traffic management personnel when formulating strategies, maintain the consistency of traffic control work, and improve collaboration efficiency.

[0064] (4) By counting the number of recommendations and regularly removing accidental strategies, it helps to reduce the probability of accidental strategies occurring again and improve the scientific nature of the strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more apparent.

[0066] Figure 1 It is a flow chart of a strategy recommendation method based on traffic situation graph classification.

[0067] Figure 2 The diagram is a schematic diagram of all published policy information for a certain period of time in the control area B in an embodiment.

[0068] Figure 3 This is a diagram of the graph classifier structure.

[0069] Figure 4 It is a schematic diagram of the strategy information of the recommended strategy. DETAILED DESCRIPTION

[0070] The present invention addresses the gap in the current air traffic flow management system in terms of automatic strategy recommendation, and provides a strategy recommendation method based on traffic situation map classification. The principle of the present invention is that when traffic management personnel formulate strategies, they often directly use or refer to strategies formulated in similar scenarios in their work experience. The present invention constructs the main information of concern to traffic management personnel when formulating strategies into a structured traffic situation map, finds historical traffic situation maps of the same category as the current traffic situation map through map classification, selects the published strategies corresponding to the historical traffic situation maps of the same category (with the highest degree of similarity to the traffic situation map) as recommended strategies, and recommends them to users after correcting the strategy effective period of the recommended strategies. The technical solution is as follows: A strategy recommendation method based on traffic situation map classification, such as Figure 1 As shown, the following steps are included:

[0071] Step 1: Create a historical policy information database. This database stores the published policy information, route information (including temporary routes) in the control area of ​​the policy issuing unit at the time of policy issuance, and predicted capacity information for each airport and waypoint in the control area of ​​the policy issuing unit for the next k hours. Specifically, it includes:

[0072] Step 1-1: Create a historical strategy information database;

[0073] Step 1-2: Create a historical policy information data table in the historical policy information database to store published policy information. The stored content must include the following information: policy identifier (the policy identifier is the unique identifier of the policy and can be regarded as the policy's "identity card"), policy name, number of recommendations (recording the number of times the policy was directly used or modified by users after being recommended), policy validity period (including policy start time and policy end time), policy release region, policy release time, and specific policy content;

[0074] Step 1-3: Create a historical route information data table in the historical policy information database to store route information (including temporary routes) for the area of ​​concern when the policy is issued. The stored content must include the following information: policy identifier, policy issuance area, route information, and temporary route information;

[0075] Steps 1-4: Create a historical forecast information table in the historical policy information database to store the predicted traffic and capacity information for each airport and waypoint in the focus area for the next k hours at the time the policy was issued. The stored information must include the following information: policy identifier, type (airport or waypoint), object name (four-character airport code or waypoint name), policy issuance area, predicted traffic for the next k hours, and predicted capacity for the next k hours.

[0076] Step 2: Select a control area as the area of ​​interest. Obtain the airport predicted capacity information, waypoint predicted capacity information, and route information (including temporary routes) for the area of ​​interest, and construct a traffic situation map; specifically, including:

[0077] Step 2-1: Obtain the user's control area (i.e., area of ​​concern) through the user's login information;

[0078] Step 2-2: Obtain route and temporary route information for the area of ​​interest;

[0079] Step 2-3: Obtain the predicted traffic and capacity for each airport and waypoint in the area of ​​interest for the next k hours;

[0080] Step 2-4: Construct traffic situation graph G pre Steps 2-4 include:

[0081] Step 2-4-1: Create a node set V. Put the airports and waypoints in the area of ​​interest into the node set V in sequence. Assuming that the total number of airports and waypoints in the area of ​​interest is N, then V = {v1, v2, v3, ...v N}. In this case, v1 represents the first node, v2 represents the second node, and so on;

[0082] Step 2-4-2: Establish edge set E. Parse route segment information from the route and temporary route information of the area of ​​interest obtained in step 2-2. If there is a route segment information from v i to v j The route segment is <v i , v j >Put it into edge set E;

[0083] Step 2-4-3: Create an adjacency matrix A (dimension N×N). Traverse the edge set E and set each element in A according to the following rules:

[0084]

[0085] Among them, A i，j represents the element in row i and column j of the adjacency matrix A;

[0086] Step 2-4-4: Calculate the ratio of the predicted traffic and predicted capacity of the node in the next k hours (if the predicted capacity is 0, take the predicted capacity as 0.01 for calculation), and use the k-dimensional vector obtained after calculation as the attribute vector of the node. Assume that the predicted traffic and predicted capacity of the i-th node in the j-th (j is 1, 2, 3, ..., k) hour are f i，j and c i，j , then the attribute vector x of the i-th node i It can be expressed as:

[0087]

[0088] The node attribute matrix X of all nodes in the node set V is: X=[x1 x2 … x N ];

[0089] Step 2-4-5: Use the airports and waypoints in the area of ​​interest as nodes, the route segments consisting of routes and temporary routes as edges, and the ratio of node predicted traffic flow and node predicted capacity in the next k hours as node attribute vectors to construct the traffic situation graph G. pre =(V, E, A, X).

[0090] Step 3: Obtain all published policies for the area of ​​interest from the historical policy information database. For each published policy, obtain the airport predicted capacity information, waypoint predicted capacity information, and route information (including temporary routes) for the area of ​​interest at the time the policy was published from the historical policy information database, construct a historical traffic situation map, and record the published policies corresponding to each historical traffic situation map; specifically,

[0091] Step 3-1: Obtain all published policies for the area of ​​interest from the historical policy information data table (assuming the total number is M).

[0092] Step 3-2: Select a published policy (assuming the selected published policy has a serial number of i among all published policies), and record the policy identifier of the selected published policy as y i ;

[0093] Step 3-3: Pass the policy identifier y i Obtain route and temporary route information of the area of ​​interest from the historical route information data table;

[0094] Step 3-4: Pass the policy identifier y i Obtain the predicted traffic and predicted capacity of all nodes for k hours when the strategy is released from the historical prediction information data table;

[0095] Step 3-5: Use the airports and waypoints in the area of ​​interest in the historical strategy information database as nodes, the route segments consisting of the routes and temporary routes in the area of ​​interest in the historical strategy information database as edges, and the ratio of the predicted flow and predicted capacity of the node k hours in the historical strategy information database as node attributes to construct the historical traffic situation graph G. i =(V i , E i , A i , X i ). G i The construction process is the same as in steps 2-4 of G pre The construction process is basically the same, the only difference is that V i 、Ei 、A i and X i All information used in the construction process is taken from the historical strategy information database;

[0096] Step 3-6: Record historical traffic situation map G i With the policy identifier y i The corresponding relationship;

[0097] Step 3-7: Repeat steps 3-2 to 3-6 for all published strategies obtained in step 3-1 to obtain M historical traffic situation graphs G1, G2, G3, ..., G M And the strategy identifier y1, y2, y3, ..., y corresponding to each historical traffic situation graph M .

[0098] Step 4: Consider each historical traffic situation map as a map category. By classifying the traffic situation map (see the supplementary description for an introduction to map classification), find the historical traffic situation map of the same category as the traffic situation map and select the published strategy corresponding to the historical traffic situation map of the same category as the recommended strategy; specifically, the following steps are involved:

[0099] Step 4-1: Consider each historical traffic situation map as a type of traffic situation map, use the strategy identifier corresponding to each historical traffic situation map as a category label of the traffic situation map, classify the traffic situation map according to the M categories of the historical traffic situation map, and obtain the classified category labels;

[0100] Step 4-2: Obtain the recommended strategy content information from the historical strategy information data table using the category label (actually a strategy identifier) ​​after classification of the traffic situation map.

[0101] Step 5: Modify the recommended policy's effective period and then push it to the user. This includes:

[0102] Step 5-1: Assume that the release time of the recommended strategy is T pub , the strategy start time is T start , the strategy end time is T end , the current time is T now , first of all, T pub and T now Round each hour (retain only hour information) to get T' pub and T′ now ;

[0103] Step 5-2: Calculate the revised strategy start time T′ using the following formula: start and the revised strategy end time T′ endThe strategy start time for updating the recommended strategy is T′ start , the strategy end time is T′ end :

[0104]

[0105] Step 6: Record the user's use of the recommended strategy. If the user does not use the recommended strategy but develops a new strategy, then when the new strategy is released, the new strategy and the airport predicted flow information, waypoint predicted flow information, and route information (including temporary routes) of the focus area at the time of the new strategy release must be stored in the historical strategy information database. Regularly check and clear the accidental strategies in the historical strategy information database as well as the airport predicted flow information, waypoint predicted flow information, and route information (including temporary routes) of the focus area at the time of the accidental strategy release. Specifically including:

[0106] Step 6-1: If the recommended strategy is used directly or modified by the user, the number of recommended times of the published strategy in the historical strategy information data table corresponding to the recommended strategy is increased by 1, and step 6-2 is skipped; otherwise, step 6-2 is executed;

[0107] Step 6-2: When the user publishes a new strategy, the new strategy information (the number of recommendations is set to 0 at this time) is stored in the historical strategy information data table, the predicted traffic and predicted capacity of each airport and waypoint in the focus area for the next k hours are stored in the historical prediction information data table, and the route information (including temporary routes) of the focus area is stored in the historical route information data table;

[0108] Step 6-3: Traverse all published strategies in the historical strategy information data table and deem any strategy whose recommendation count has been zero for a certain period (e.g., three years) since its release as a contingency strategy. Delete the contingency strategy information from the historical strategy information data table and, using the strategy identifier of the contingency strategy, delete the relevant information from the historical prediction information data table and the historical route information data table.

[0109] Supplementary explanation: Graph classification is a learning task at the graph level. In layman's terms, it is to comprehensively analyze the structural information and node information of the graph to compare the similarity between the graph of unknown category and the graph of known category, and find the category of the graph of unknown category from a limited number of graph categories. The graph classification function is mostly implemented using graph neural network technology, which is mainly divided into two steps: (1) using global pooling or hierarchical pooling to extract and aggregate the structural features and node features of the graph to obtain the vector space representation of the graph; (2) using the vector space classification method to classify the vector space representation of the graph. The specific method is: (1) using graph convolution to learn a score for each node in the graph; (2) using graph pooling to eliminate low-scoring nodes; (3) using the readout mechanism to aggregate the remaining node information to obtain the vector space representation of the graph; (4) using the vector space representation of the graph to implement the graph classification function using the vector space classification method.

[0110] Example:

[0111] The following is a specific implementation of the strategy recommendation method based on traffic situation map classification in a traffic management system A that is actually in operation, with a portion of the control area B as the focus area, and with reference to the accompanying drawings. Parameter description: k = 6, T now The time is 2021-09-17 19:01. The specific implementation steps are as follows:

[0112] 1. Create a historical strategy information database, and create a historical strategy information data table, a historical route information data table, and a historical prediction information data table in the historical strategy information database.

[0113] 2. Build a traffic situation map

[0114] 2-1. Obtain information on airports (24 in total) and waypoints (267 in total) in control area B through the environmental data interface of traffic management system A. N = 24 + 267 = 291;

[0115] 2-2. 961 domestic air routes (including 254 temporary routes) and corridor information for all airports in control area B are obtained through traffic management system A. The information on all air routes in control area B at the current moment (a total of 285 segments) is then parsed.

[0116] 2-3. Obtain the predicted traffic volume for each airport and waypoint in the next k hours in the B control area through the A traffic management system, and obtain the predicted capacity for each airport and waypoint in the B control area through the A traffic management system. Since there are a large number of airports and waypoints in the B control area, in order to simplify the description, this implementation method will only be explained using Airport C as an example. The predicted arrival traffic volume and predicted arrival capacity of Airport C in the next k hours (starting from 2021-09-17 20:00) are [48, 51, 43, 40, 31, 29] and [45, 45, 45, 45, 45, 45] respectively. It is not difficult to find that Airport C will be in a state of capacity and flow imbalance in the next 2 hours, and traffic management personnel need to formulate and publish strategies to achieve capacity and flow balance;

[0117] 2-4. Use the airport and waypoint information of control area B obtained in 2-1, the route information of control area B obtained in 2-2, and the predicted capacity flow information of each airport and waypoint in control area B for the next k hours obtained in 2-3 to construct a traffic situation map.

[0118] 3. Build a historical traffic situation map

[0119] 3-1. Obtain all historical policies issued by control area B from the historical policy information data table, such as Figure 2 As shown;

[0120] 3-2. Select a policy and record the policy identifier (in Traffic Management System A, the policy identifier is generated when the policy is formulated and is usually used as the primary key field in the database and is generally not displayed in the interface);

[0121] 3-3. Using the policy identifier of the selected published policy, obtain the route and temporary route information at the time of policy release from the historical route information data table, and obtain the predicted traffic flow and predicted capacity for each airport and each waypoint in the B control area for k hours at the time of policy release from the historical forecast information data table, and construct a historical traffic situation map. Record the correspondence between the historical traffic situation map and the policy identifier;

[0122] 3-4. Repeat 3-2 and 3-3 for all historical strategies obtained in 3-1 to obtain the historical traffic situation map corresponding to all published strategies.

[0123] 4. Obtain recommendation strategy through graph classification

[0124] 4-1. Build a graph classifier, the structure diagram is as follows Figure 3As shown in the figure, the graph classifier in this implementation adopts a layered pooling method. Layered pooling can effectively solve the problem of large amounts of information loss in the early stages of the graph pooling process. By connecting the output of each pooling layer through a fully connected layer, a vector space representation of the graph is obtained. The vector space representation of the graph is then used as the input of the vector space convolutional neural network classifier to achieve graph classification.

[0125] 4-2. Use the historical traffic situation map and the strategy identifier corresponding to each historical traffic situation map as training samples to train the map classifier;

[0126] 4-3. Input the traffic situation map into the map classifier to obtain the category label of the traffic situation map, and obtain the recommended strategy content information from the historical strategy information data table through the category label (actually a strategy identifier of a published strategy). Figure 4 As shown;

[0127] 5. Modify the effective period of the recommended strategy. Figure 4 As shown, the recommended policy was published at 2021-06-25 19:12, its start time is 2021-06-25 20:30, and its end time is 2021-06-25 23:59. The revised policy start time is calculated to be 2021-09-17 20:30, and the revised policy end time is 2021-09-17 23:59. After updating the recommended policy start and end times, the policy content is pushed to users for use.

[0128] 6. Using traffic management system A to simulate and analyze the recommended strategy, the predicted traffic flow at Airport C is [45, 47, 46, 43, 32, 29], greatly alleviating the need for capacity-flow balance. The user adds the flow control settings "Strategy start time: 2021-09-17 20:30; Strategy end time: 2021-09-17 23:59; Flow control content: 3 aircraft will land at Airport C via ENLAB point / 60 minutes, minimum 10 minutes." Simulation analysis is then performed again, resulting in a predicted traffic flow at Airport C of [45, 45, 44, 44, 34, 30], which meets the need for capacity-flow balance.

[0129] 7. This recommended strategy is a user-modified recommended strategy. After the strategy is published, the recommended count of the corresponding published strategy in the historical strategy information data table will be increased by 1.

[0130] In a specific implementation, the present application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium is capable of storing a computer program that, when executed by the data processing unit, executes the invention of a strategy recommendation method based on traffic situation map classification provided by the present invention and some or all of the steps in each embodiment. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0131] Those skilled in the art will clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of computer programs and their corresponding general hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, in essence or in other words, the part that contributes to the prior art, can be embodied in the form of a computer program, i.e., a software product. The computer program software product can be stored in a storage medium and includes a number of instructions for enabling a device including a data processing unit (which can be a personal computer, server, single-chip microcomputer, MCU, or network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0132] The present invention provides a method and approach for policy recommendation based on traffic situation map classification. There are numerous methods and approaches for implementing this technical solution. The above is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art may make various improvements and modifications without departing from the principles of the present invention, and such improvements and modifications are also within the scope of protection of the present invention. Any components not specified in this embodiment may be implemented using existing technologies.

Claims

1. A strategy recommendation method based on traffic situation map classification, characterized in that: The steps include: Step 1, create a historical policy information database to store the issued policies and the traffic information in the control area of ​​the unit that issued the policy when the policy was issued; Step 2: Select a controlled area as a focus area, obtain traffic information in the focus area, and construct a traffic situation map; Step 3, obtaining the published policies of the concerned area from the historical policy information database, and for each published policy, obtaining the traffic information of the concerned area when the policy was published from the historical policy information database, constructing a historical traffic situation map and recording the published policies corresponding to each historical traffic situation map; Step 4, select a recommended strategy, regard each historical traffic situation map as a map category, classify the traffic situation map, find a historical traffic situation map of the same category as the traffic situation map, and select the published strategy corresponding to the historical traffic situation map as the recommended strategy; Step 5: Modify the effective period of the recommended strategy and push it to the user; Step 6, update the historical strategy information database, record the user's use of the recommended strategy, if the user does not use the recommended strategy but adopts a new strategy, then store the new strategy and the traffic information of the concerned area when the new strategy is released in the historical strategy information database.

2. The strategy recommendation method based on traffic situation map classification according to claim 1 is characterized in that: The traffic information described in step 1 includes: route information, airport predicted capacity information and waypoint predicted capacity information.

3. The strategy recommendation method based on traffic situation map classification according to claim 2 is characterized in that: The route information in step 1 includes temporary route information.

4. The strategy recommendation method based on traffic situation map classification according to claim 3 is characterized in that: The historical strategy information database described in step 1 includes: Historical strategy information data table, historical route information data table and historical forecast information data table, among which, The historical policy information data table stores the released policy information, including: policy identifier, policy name, number of recommendations, policy effective period, policy release area, policy release time and policy content; The historical route information data table stores the route information of the control area of ​​the unit that issues the policy when the policy is issued, including: policy identifier, policy issuance area, route information and temporary route information; The historical prediction information data table stores the predicted traffic and predicted capacity information of each airport and each waypoint in the control area of ​​the unit that issues the strategy when the strategy is issued for the next k hours.

5. The strategy recommendation method based on traffic situation map classification according to claim 4 is characterized in that: The historical forecast information data table in step 1 includes: a policy identifier, an object type, an object name, a policy release area, a predicted flow rate for the next k hours, and a predicted capacity for the next k hours; wherein the object is an airport or a waypoint.

6. The strategy recommendation method based on traffic situation diagram classification according to claim 5 is characterized in that: The construction of the traffic situation map described in step 2 includes: Step 2-1, establish a node set V, as follows: The airports and waypoints in the area of ​​interest are sequentially put into the node set V. Assuming the total number of airports and waypoints in the area of ​​interest is N, the node set is expressed as follows: V={v1,v2,v3,…v N } Among them, v N Represents the Nth node; Step 2-2, establish the edge set E, as follows: Parse the route segment from the route information and temporary route information in the traffic information within the area of ​​interest, if there is a route segment specified by v i to v j If the route segment <v i ,v j >Put it into the edge set E, where v i and v j Represents the i-th node and the j-th node in the node set V; Step 2-3, establish an adjacency matrix A with a dimension of N×N, as follows: Traverse the edge set E and set each element in the adjacency matrix A. The rules are as follows: Among them, A i,j Represents the element in the i-th row and j-th column of the adjacency matrix A; Step 2-4, calculate the ratio of the predicted flow and predicted capacity of each node in the node set V in the next k hours, and use the obtained k-dimensional vector as the attribute vector of the node; let the predicted flow and predicted capacity of the i-th node in the t-th hour be f i,t and c i,t , where t = 1, 2, ..., k, then the attribute vector x of the i-th node i It is expressed as: The node attribute matrix X composed of all nodes in the node set V is: X=[x1 x2 … x N ] Step 2-5: Take the airports and waypoints in the area of ​​interest as nodes, the route segments consisting of routes and temporary routes as edges, and the node attribute vector with the ratio of node predicted flow and node predicted capacity in the next k hours. pre , which is expressed as follows: G pre =(V,E,A,X)。 7. The strategy recommendation method based on traffic situation diagram classification according to claim 6 is characterized in that: The construction of the historical traffic situation map and recording of the published policies corresponding to each historical traffic situation map described in step 3 include: Step 3-1, obtain the published strategies released in the concerned area from the historical strategy information data table, and set the number of published strategies to M; Step 3-2, select the mth published policy, the policy identifier of which is y m ; Step 3-3, through the policy identifier y m Obtain route information and temporary route information of the concerned area from the historical route information data table; Step 3-4, through the policy identifier y m Obtain the predicted traffic and predicted capacity of all nodes for k hours when the strategy is released from the historical prediction information data table; Step 3-5, using the airports and waypoints in the concerned area in the historical strategy information database as nodes, the route segments consisting of the routes and temporary routes in the concerned area in the historical strategy information database as edges, and the predicted flow and predicted capacity ratio of the node k hour in the historical strategy information database as node attributes, the same method as described in step 2 is used to construct the historical traffic situation graph G m =(V m ,E m ,A m ,X m ); Step 3-6, record the historical traffic situation map G m With the policy identifier y m The corresponding relationship; Step 3-7: Repeat steps 3-2 to 3-6 for all published strategies obtained in step 3-1 to obtain M historical traffic situation graphs G1, G2, G3, ..., G M and their corresponding policy identifiers y1, y2, y3, …, y M .

8. The strategy recommendation method based on traffic situation diagram classification according to claim 7 is characterized in that: Recommended strategies as described in Step 4, including: Step 4-1: regard the M historical traffic situation graphs as M types of traffic situation graphs, use their corresponding strategy identifiers as category labels, and classify the traffic situation graph G according to the M categories of the historical traffic situation graphs. pre Perform classification and obtain the classified category labels; Step 4-2, obtain the policy content of the recommended policy from the historical policy information data table through the category label, that is, the policy identifier.

9. The strategy recommendation method based on traffic situation diagram classification according to claim 8 is characterized in that: The revised recommended strategy described in step 5 will be effective during the following period: Step 5-1, let the release time of the recommended strategy be T pub , the strategy start time is T start , the strategy end time is T end , the current time is T now , for the strategy release time T pub and the current time T now Round up by hour, that is, only keep the hour information, and get the release time of the rounding strategy T p ′ ub and round the current time T n ′ ow ; Step 5-2, calculate the revised strategy start time T s ′ tart and the revised strategy end time T e ′ nd , as follows: Use the modified strategy starting time T s ′ tart and the revised strategy end time T e ′ nd Update the strategy start time T of the recommended strategy start , strategy end time T end .

10. The strategy recommendation method based on traffic situation diagram classification according to claim 9 is characterized in that: The updating history strategy information database described in step 6 includes: Step 6-1, if the recommended strategy is used directly or modified by the user, the number of recommendations of the published strategy in the historical strategy information data table corresponding to the recommended strategy is increased by 1, and step 6-3 is executed; otherwise, step 6-2 is executed; Step 6-2, storing the new strategy adopted by the user into the historical strategy information data table and setting the number of recommendations to 0, storing the predicted traffic and predicted capacity information of each airport and each waypoint in the current focus area for the next k hours into the historical prediction information data table, and storing the route information and temporary route information of the focus area into the historical route information data table; Step 6-3, traverse all published strategies in the historical strategy information data table, and regard the published strategies whose recommendation times have been 0 within the preset period starting from the strategy release time as accidental strategies; delete the relevant information of the accidental strategies in the historical strategy information data table, the historical prediction information data table and the historical route information data table.

Images