A ground-air collaborative UAV dynamic game decision-making management planning system

Abstract

The present invention discloses a ground-to-air coordinated unmanned aerial vehicle (UAV) dynamic game decision-making management and planning system. The system comprises: an aerial UAV array state sensor group, a ground-based decision-making management center, and a planning executor. The ground-based decision-making management center includes a sensor database, a data preprocessing module, a spatiotemporal graph module, and a dynamic game decision-making module. Historical flight data is preprocessed and input into the spatiotemporal graph module. A graph convolutional neural network is used to analyze spatial position changes, and a gated recurrent unit is combined to capture temporal dynamics, thereby generating a pre-trained game decision-making model. Current flight data is preprocessed and input into the pre-trained game decision-making model. The generated state forecast is manually verified and adjusted to generate decision instructions. The execution module in the planning executor receives the decision instructions and sends corresponding operational instructions. The present invention innovatively introduces a spatiotemporal graph module into the UAV dynamic game decision-making process, accurately simulating the synergistic effects between individual UAVs and their temporal dynamic relationships. Ground-to-air coordination is achieved by combining the pre-trained model of the ground-based control decision center with real-time flight data from the aerial UAV array, completing adaptive adjustment and planning of dynamic game decisions.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a ground-to-air coordinated unmanned aerial vehicle (UAV) dynamic game decision-making management and planning system. Background Art

[0002] Drones, due to their flexibility, real-time data transmission, and autonomous mission execution, have become one of the fastest, most cost-effective, and most widely used unmanned means of carrying out missions worldwide. Drones are often deployed in arrays to collectively carry out missions. With the increasing prevalence of drone arrays, researchers have begun to pay attention to the changing patterns of spatial positional relationships and individual dynamic characteristics. Typically, drone arrays are equipped with integrated status sensors on their upper panels to monitor system status. By integrating sensor data from all drones, a database reflecting the dynamic game state of the drones can be established for remote ground control. Through ground-to-air collaboration, ground control centers can leverage historical flight databases to manage and plan the dynamic game of the drone array, achieving more efficient mission execution and system safety. Within an array, drones often collaborate with each other to carry out missions. When a drone serves as the primary mission performer, the speed and position relationships with other supporting drones become complex. These factors cause the spatial and temporal relationships within the drone array to fluctuate, become uncertain, and become unstable, impacting the system's dynamic game planning and mission execution efficiency, and even leading to safety incidents.

[0003] To address this challenge, it is crucial to scientifically and accurately capture the spatial position changes and individual temporal dynamics of drone arrays. This not only helps improve drone energy efficiency, but also ensures mission safety and increases the precision and adaptability of dynamic game decision-making. Artificial intelligence detection and planning methods can exploit more implicit features in drone sensor data and possess stronger learning capabilities. Recurrent Neural Networks (RNNs) and Gated Recurrent Units (GRUs), which leverage the properties of time series, are widely used to handle long-term temporal dependencies and time series data.

[0004] The GRU, a variant of recurrent neural networks, employs a unique gating structure and memory units to address the vanishing or exploding gradient issues of traditional networks when processing long-term time series data. Its hidden layer consists of input, output, and forget gates, with information selectively retained and unit state updates controlled by a system of gating units. For complex, high-dimensional drone array data, the GRU can exploit inherent hidden temporal dynamics and handle long-term temporal dependencies. However, in the dynamic game decision-making of drones, the GRU, relying solely on long-term temporal dependencies, struggles to detect the spatial coupling and mutual influence between drones in the array. Therefore, it is necessary to combine appropriate methods to exploit spatial features and obtain a more comprehensive and accurate understanding of the spatial changes in drone arrays.

[0005] Because drone sensors are distributed in a fixed pattern, the position information and other data they transmit back can represent the interrelationships between individual drones. Given the mutual influence and collaborative nature of individual drones in an array, the graph-learning-based GCN architecture offers a more universal advantage, capable of exploring the spatial coupling relationships between different drones in an array and more accurately determining their collaborative state. Furthermore, the recurrent neural network-based GRU architecture can capture the temporal dynamics of individual drones, allowing system detection to factor in the influence of historical flight data. However, currently, few methods and devices have incorporated graph learning architectures to assist in dynamic game decision-making, management, and planning for drones. Previous time-series models, such as the GRU, used for dynamic drone games have failed to incorporate the specific collaborative scenarios found in drone array applications through graph learning, resulting in low state detection accuracy and high error rates, leading to economic losses and reliability issues. Therefore, there is an urgent need to develop a better ground-to-air collaborative drone game decision-making, management, and planning system that can achieve adaptive, high-precision dynamic game decisions. Summary of the Invention

[0006] The purpose of the present invention is to address the deficiencies of the existing technology and provide a ground-to-air coordinated UAV dynamic game decision-making management planning system.

[0007] The objective of the present invention is achieved through the following technical solutions: a ground-to-air coordinated UAV dynamic game decision-making management planning system, the system comprising: a UAV array state sensor group located in the air, a decision-making management center located on the ground, and a planning executor; the decision-making management center located on the ground comprises a sensor database, a data preprocessing module, a spatiotemporal graph module, and a dynamic game decision module;

[0008] The drone array status sensor group located in the air is used to detect the real-time position and individual status of the drone array, and transmit the newly monitored current flight data to the sensor database in the decision-making management center located on the ground;

[0009] The drone array consists of N drones, each of which is equipped with a corresponding state sensor;

[0010] The sensor database is used to receive the current flight data obtained by new monitoring and store historical flight data, and then transfer the current flight data and historical flight data to the data preprocessing module;

[0011] The data preprocessing module is used to perform preprocessing operations on the current flight data and the historical flight data respectively, wherein the preprocessing operations include missing value filling and data cleaning, to obtain preprocessed current flight data and historical flight data, and then transfer the preprocessed historical flight data to the spatiotemporal graph module, and transfer the preprocessed current flight data to the dynamic game decision module;

[0012] The spatiotemporal graph module is used to perform mutual information evaluation on the pre-processed historical flight data to screen variables, obtain the adjacency matrix using the maximum nearest neighbor method, train the spatiotemporal graph model using the GCN graph learning module and the GRU time series learning module, and save the optimal model as the pre-trained game decision model and save it in the dynamic game decision module;

[0013] The dynamic game decision module is used to use the pre-trained game decision model to predict the pre-processed current flight data, obtain the predicted flight data for the next moment, and generate a state forecast based on it; and manually confirm and adjust the state forecast to generate corresponding decision instructions and send them to the planning executor;

[0014] The planning executor is used to use the internal execution module to receive decision instructions and send corresponding operation instructions to achieve the synergy effect and adaptive adjustment planning of the drone cluster.

[0015] Furthermore, the sensor database receives the current flight data obtained through new monitoring and stores historical flight data, specifically:

[0016] The sensor database receives the current flight data obtained by new monitoring, and the current flight data is in, is the flight data sent back by the state sensor on the j-th UAV in the UAV array at the m-th moment, j = 1, 2, …, j, …, N, and m represents the current moment;

[0017] The sensor database stores historical flight data. The historical flight data is

[0018]

[0019] in, is the flight data set sent back by the state sensor on the jth UAV in the UAV array at time 0 to m-1, is the flight data sent back by the j-th UAV at the i-th moment.

[0020] Furthermore, the data preprocessing module performs preprocessing operations on the current flight data and the historical flight data respectively. The preprocessing operations include missing value filling and data cleaning to obtain the preprocessed current flight data and historical flight data, specifically:

[0021] The data preprocessing module converts historical flight data Perform preprocessing operations, including missing value filling and data cleaning, to obtain preprocessed historical flight data

[0022]

[0023] in, is the historical flight data set after preprocessing operation of the j-th UAV; The historical flight data of the j-th UAV after preprocessing operation at the i-th moment;

[0024] The missing value filling is specifically as follows: when the flight data returned by the j-th UAV at the i-th moment Any variable in If there is a missing value, the missing value is filled by using the variables of the k moments before and after through mathematical calculation to obtain the variable after the missing value is filled. The formula is as follows:

[0025]

[0026] The data cleaning is specifically to delete the flight data of each drone at every moment. Complete data cleaning based on the erroneous data in the data;

[0027] The data preprocessing module converts the current flight data into Perform preprocessing operations to obtain the preprocessed current flight data in, is the current flight data of the j-th UAV after the preprocessing operation at the m-th moment.

[0028] Furthermore, the space-time graph module performs mutual information evaluation on the pre-processed historical flight data to screen variables, obtains the adjacency matrix using the maximum nearest neighbor method, and trains the space-time graph model using the GCN graph learning module and the GRU time series learning module. The optimal model is saved as the pre-trained game decision model and saved in the dynamic game decision module. Specifically, the process includes the following sub-steps:

[0029] (a.1) The key component of the spatiotemporal graph model is the spatiotemporal graph neural network, which is divided into three key elements: the time domain, the spatial domain, and the graph structure. Graph structure: The construction of the graph includes two steps: screening UAVs that are highly relevant to the flight mission, and determining the edge relationships between these UAVs and generating the adjacency matrix required by the GCN. Spatial domain: The graph convolutional neural network is a neural network specifically designed to process graph-structured data. It learns the interrelationships between different UAV flight data from a spatial perspective. It captures the spatial dependencies between UAVs by propagating information between each UAV and its neighbors in the graph. Temporal domain: The gated recurrent unit is a recurrent neural network used for time series analysis. It introduces a gating mechanism to control the flow of information from the perspective of the time domain to capture the dynamic changes in the flight data of each UAV over time.

[0030] First, the preprocessed historical flight data The mutual information value is obtained by evaluating the historical flight data set after preprocessing operations for all arbitrary two UAVs A and B. The calculation formula is as follows:

[0031]

[0032] Where A=1,2,…,A,…,N,B=1,2,…,B,…,N,A≠B;αis the historical flight data set after preprocessing operation The number of grids divided in the direction, β is the historical flight data set after preprocessing operation The number of grids divided in the direction; for and The joint probability of for The marginal probability of for The marginal probability of ;λ is the variable control parameter;

[0033] Arrange the mutual information values ​​corresponding to all two arbitrary drones from large to small, select the historical flight data set after preprocessing corresponding to the first S drones with the largest mutual information values ​​as the key historical flight data set, and obtain the key historical flight data

[0034]

[0035] in, represents the g-th key historical flight data set, g=1,2,...,g,…,S;

[0036] (a.2) Use the maximum neighbor method to analyze key historical flight data Construct edge relationships and obtain edge weight matrix

[0037]

[0038] Among them, edge f,g Represents a collection of key historical flight data and key historical flight data sets The edge weights between them are f=1,2,…,f,…,S; when f≠g, Where ξ represents the bandwidth of the Gaussian kernel function; when f = g, edge f,g =0;

[0039] Key historical flight data The similarity is obtained with each key historical flight data set The K other key historical flight data sets with the greatest similarity are in the edge weight matrix Only keep key historical flight data sets The edge weights of the corresponding K other key historical flight data sets are modified to 0, and the graph structure is obtained.

[0040] Then the graph structure Converted into adjacency matrix structure It is best to get key historical flight data Image in, Shows key historical flight data The time step relationship in

[0041] (a.3) Then the figure Input into the GCN graph learning module, use the graph convolution network to perform convolution calculation, and get the output of the GCN graph learning module

[0042] Among them, σ(·) is the activation function; U is given by Calculated by Laplace:

[0043]

[0044] Where D is the graph The node degree matrix of

[0045] g θ Is the convolution kernel function:

[0046] Among them, θ k is the polynomial parameter, T k are the Chebyshev polynomial parameters, For key historical flight data The calculated diagonal matrix;

[0047] (a.4) Output It consists of sub-outputs from time 0 to time t-1: in, Represents the output at the i-th moment;

[0048] and key historical flight data Divide by time 0 to time t-1, divided into in, represents the key historical flight data set at the i-th moment;

[0049] It will then output and key historical flight data Input to the GRU timing learning module to capture temporal dependencies based on the gated recurrent unit;

[0050]

[0051] Where Θ is the Hadamard symbol; Represents the output of the update gate at the i-th moment; represents the output of the reset gate at the i-th moment; is the hidden state at the i-1th moment; is the hidden state at the i-th moment; is the candidate hidden state at the i-th moment; W xr 、W xz 、W xh 、W hr 、W hz and W hh are learnable parameters; b r 、b z and b h are bias parameters respectively;

[0052] After the above calculation, the time domain convolution output at the m-1th moment is obtained Among them, is a learnable parameter, b γ is the bias parameter;

[0053] (a.5) Then according to the time domain convolution output at the m-1th moment The key historical flight data set at the m-1th moment Calculate the mean square error to get the total loss; based on the total loss, the learnable parameter W xr 、W xz 、Wxh 、W hr 、W hz 、W hh and W γ And the bias parameter b r 、b z 、b h and b γ Update the network and record the state of the entire network when the total loss is minimized, and save it as the optimal model; then use the optimal model as the pre-trained game decision model and save it in the dynamic game decision module.

[0054] The beneficial effects of the present invention are: 1. The GCN graph learning module is innovatively introduced to perform graph convolution analysis on the spatial distribution of drone sensors, so that the state detection model can deeply explore the spatial characteristics of the sensor signals and improve the decision-making reliability in drone applications; 2. Combining long time series and spatial dimensions, the dynamic game state data of the drone is comprehensively learned, and the dynamically changing spatiotemporal relationships are accurately extracted to improve the real-time and accuracy of decision-making; 3. According to the operating characteristics of the drone, the decision model is adaptively adjusted to improve the response capability to abnormal situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a structural diagram of a ground-air collaborative UAV dynamic game decision-making management and planning system;

[0056] Figure 2 A schematic diagram of the locations of the corresponding status sensors installed in the drone;

[0057] Figure 3 is the spatial position distribution map of the drone array;

[0058] In the figure, 1-state sensor; 2-UAV array; 3-UAV array state sensor group in the air; 4-Sensor database; 5-Data preprocessing module; 6-Spatiotemporal graph module; 7-Dynamic game decision module; 8-Planning executor; 9-Decision management center on the ground; 10-Execution module. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to illustrate the present invention, rather than to represent all embodiments. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments of the present invention without creative work are within the scope of protection of the present invention.

[0060] In order to overcome the challenges of mining the spatial coupling relationship between individual drones and capturing the individual temporal characteristics in the dynamic game decision-making management planning of drone arrays, the purpose of the present invention is to construct a ground-to-air collaborative decision-making system, collect status data from the air, build a database in the ground decision-making management center, and combine GCN to mine the spatial coupling relationship between individuals in the drone array and GRU to capture the temporal dynamic characteristics of individuals, intelligently manage the dynamic game decision-making of drones, and propose a ground-to-air collaborative drone dynamic game decision-making management planning system to achieve efficient task execution and system security.

[0061] It constructs a graph model to characterize the spatial topological relationship of drones in the array, uses GCN to learn the spatial dependency of the graph, and GRU to learn the temporal dependency, and uses this convolution kernel cyclic splicing to predict future states through historical flight data and current flight data and perform dynamic game decision-making tasks.

[0062] Example 1

[0063] like Figure 1 As shown, the present invention provides a ground-to-air coordinated UAV dynamic game decision-making management and planning system, comprising: an aerial UAV array state sensor group 3, a ground-based decision-making management center 9, and a planning executor 8. The ground-based decision-making management center 9 includes a sensor database 4, a data preprocessing module 5, a spatiotemporal graph module 6, and a dynamic game decision-making module 7.

[0064] The drone array status sensor group 3 located in the air is used to detect the real-time position and individual status of the drone array 2, and transmit the current flight data obtained by the new monitoring to the sensor database 4 in the decision-making management center 9 located on the ground.

[0065] The drone array 2 is composed of N drones, each of which is equipped with a corresponding state sensor 1, such as Figure 2 As shown. Figure 3 As shown, N drones are distributed in a certain spatial position and head to a designated mission location. The current flight data obtained by the new monitoring includes information such as the position, speed, and energy level of the drones.

[0066] The sensor database 4 is used to receive the current flight data obtained by new monitoring and store historical flight data, specifically:

[0067] The sensor database 4 receives the current flight data obtained by new monitoring, and the current flight data is in, is the flight data sent back by the status sensor on the j-th UAV in the UAV array 2 at the m-th moment, j = 1, 2, ..., j, ..., N, and m represents the current moment.

[0068] The sensor database 4 stores historical flight data, which are

[0069]

[0070] in, is the flight data set sent back by the state sensor on the jth UAV in the UAV array 2 at time 0 to m-1, is the flight data sent back by the j-th UAV at the i-th moment.

[0071] The current flight data and historical flight data are then transferred to the data preprocessing module 5 .

[0072] The data preprocessing module 5 is used to perform preprocessing operations on the current flight data and the historical flight data respectively. The preprocessing operations include missing value filling and data cleaning to obtain the preprocessed current flight data and historical flight data. Specifically,

[0073] The data preprocessing module 5 converts historical flight data Perform preprocessing operations, including missing value filling and data cleaning, to obtain preprocessed historical flight data

[0074]

[0075] in, is the historical flight data set after preprocessing operation of the j-th UAV; The historical flight data of the j-th UAV after the preprocessing operation at the i-th moment.

[0076] The missing value filling is specifically as follows: when the flight data returned by the j-th UAV at the i-th moment Any variable in If there is a missing value, the variables before and after k moments are used to fill the missing value through mathematical calculation to obtain the variable after the missing value is filled. The formula is as follows:

[0077]

[0078] The data cleaning is specifically to delete the flight data of each drone at every moment. Complete data cleaning by removing the erroneous data in the data.

[0079] The data pre-processing module 5 converts the current flight data into Perform preprocessing operations to obtain the preprocessed current flight data in, is the current flight data of the j-th UAV after the preprocessing operation at the m-th moment.

[0080] The pre-processed historical flight data is then transferred to the space-time graph module 6 , and the pre-processed current flight data is transferred to the dynamic game decision module 7 .

[0081] The spatiotemporal graph module 6 is used to perform mutual information evaluation on the pre-processed historical flight data to screen variables, obtain the adjacency matrix using the maximum nearest neighbor method, train the spatiotemporal graph model using the GCN graph learning module and the GRU time series learning module, and save the optimal model as the pre-trained game decision model and save it in the dynamic game decision module 7. Specifically, it includes the following sub-steps:

[0082] (a.1) The key component of the spatiotemporal graph model is the spatiotemporal graph neural network, which is divided into three key elements: time domain, space domain, and graph structure. Graph structure: The construction of the graph includes two steps: screening drones that are highly relevant to the flight mission (based on mutual information evaluation), and determining the edge relationships between these drones and generating the adjacency matrix required by GCN (based on the maximum nearest neighbor method). Spatial domain: The graph convolutional neural network (GCN) is a neural network specifically designed for processing graph-structured data. It learns the mutual correlation between different drone flight data from a spatial perspective. It captures the spatial dependencies between drones by propagating information between each drone and its neighbors in the graph. Temporal domain: The gated recurrent unit (GRU) is a recurrent neural network used for time series analysis. It introduces a gating mechanism to control the flow of information from the perspective of the time domain to capture the dynamic changes in the time of each drone's flight data.

[0083] Using Mutual Information (MIC) to evaluate pre-processed historical flight data By analyzing the correlation between the pre-processed historical flight data sets corresponding to any two UAVs in the game, the pre-processed historical flight data set with the greatest information contribution can be selected to ensure that the decision model receives the richest input information. This method can identify the S pre-processed historical flight data sets with the strongest collaborative relationships in the current game state and use this most relevant data for modeling, eliminating data redundancy and reducing unnecessary information processing, thereby improving the accuracy of game decisions and the intelligence level of the system. At the same time, this targeted sensor selection also enhances the system's flexibility and adaptability in complex and changing environments. This selected sensor data is used as input variables in the spatiotemporal graph model, providing more effective support for subsequent decision-making and management.

[0084] First, the preprocessed historical flight data The mutual information value is obtained by evaluating the historical flight data set after preprocessing operations for all arbitrary two UAVs A and B. The calculation formula is as follows:

[0085]

[0086] Where A=1,2,...,A,…,N,B=1,2,…,B,…,N,A≠B;αis the historical flight data set after preprocessing operation The number of grids divided in the direction, β is the historical flight data set after preprocessing operation The number of grids divided in the direction; for and The joint probability of for The marginal probability of for The marginal probability of ; λ is the variable control parameter.

[0087] Arrange the mutual information values ​​corresponding to all two arbitrary drones from large to small, select the historical flight data set after preprocessing corresponding to the first S drones with the largest mutual information values ​​as the key historical flight data set, and obtain the key historical flight data

[0088]

[0089] in, represents the g-th key historical flight data set, g = 1, 2,…, g,…, S.

[0090] (a.2) Use the KNN method to select key historical flight data The core of constructing edge relationships is to calculate the maximum similarity between the selected key historical flight data sets and record the maximum similarity as the edge weight relationship between the key historical flight data sets, thereby characterizing the key historical flight data. The interdependencies and synergies between each key historical flight data set in the

[0091] Use the maximum nearest neighbor method to analyze key historical flight data Construct edge relationships and obtain edge weight matrix

[0092]

[0093] Among them, edge f,g Represents a collection of key historical flight data and key historical flight data sets The edge weights between them are f=1,2,…,f,…,S; when f≠g, Where ξ represents the bandwidth of the Gaussian kernel function; when f = g, edge f,g =0.

[0094] Key historical flight data The similarity is obtained with each key historical flight data set The K other key historical flight data sets with the greatest similarity are in the edge weight matrix Only keep key historical flight data sets The edge weights of the corresponding K other key historical flight data sets are modified to 0, and the graph structure is obtained.

[0095] Then the graph structure Converted into adjacency matrix structure Finally get the key historical flight data Image in, Shows key historical flight data The time step relationship in .

[0096] (a.3) Using Graph Convolutional Network (GCN) to perform convolution calculations on the graph structure can effectively obtain key historical flight data The deep coupling between the S key historical flight data sets during collaboration provides rich spatiotemporal information for decision-making and management. In the dynamic environment of ground-air collaboration, drones need to quickly adapt to changing conditions, resulting in complex and variable relative positional relationships. The graph convolutional network in the GCN graph learning module can specifically extract feature information of different drone nodes, such as speed and energy. By aggregating information from neighboring nodes, it enhances the individual drone's perception of the overall environment. This is crucial for collaborative decision-making and helps drone swarms develop a global perspective. Furthermore, the GCN model is highly scalable and can adapt to drone swarms of varying sizes.

[0097] Then the figure Input into the GCN graph learning module, use the graph convolution network to perform convolution calculation, and get the output of the GCN graph learning module

[0098] Among them, σ(·) is the activation function; U is given by Calculated by Laplace:

[0099]

[0100] Where D is the graph The node degree matrix of

[0101] g θ Is the convolution kernel function:

[0102] Among them, θ k is the polynomial parameter, T k are the Chebyshev polynomial parameters, For key historical flight data The calculated diagonal matrix.

[0103] (a.4) Using a gated recurrent unit (GRU) to capture the temporal dependencies of individual drones can effectively enhance the performance of dynamic game decision management systems. GRU excels at modeling the dynamic changes in time series data and can accurately capture the evolution of drone states over time, which is crucial for decision management in dynamic environments. In addition, the GRU's gating mechanism can selectively retain and forget historical information, improving the efficiency of modeling long-term dependencies, helping individual drones maintain a memory of previous states and make more coherent decisions. Combining GRU with GCN, which excels at spatial modeling, can comprehensively capture the dynamic characteristics of drone groups in the spatiotemporal dimensions, providing richer information for decision management and further enhancing the continuity and robustness of the system.

[0104] The output It consists of sub-outputs from time 0 to time t-1: in, Represents the output at the i-th moment;

[0105] and key historical flight data Divide by time 0 to time t-1, divided into in, represents the key historical flight data set at the i-th moment;

[0106] It will then output and key historical flight data Input to the GRU timing learning module to capture temporal dependencies based on the gated recurrent unit;

[0107]

[0108] Where Θ is the Hadamard symbol; Represents the output of the update gate at the i-th moment; represents the output of the reset gate at the i-th moment; is the hidden state at the i-1th moment; is the hidden state at the i-th moment; is the candidate hidden state at the i-th moment; W xr 、W xz 、W xh 、W hr 、W hz and W hh are learnable parameters; b r 、b z and b h are bias parameters respectively;

[0109] After the above calculation, the time domain convolution output at the m-1th moment is obtained Among them, is a learnable parameter, b γ is the bias parameter;

[0110] (a.5) Then according to the time domain convolution output at the m-1th moment The key historical flight data set at the m-1th moment Calculate the mean square error to get the total loss; based on the total loss, the learnable parameter W xr 、W xz 、W xh 、W hr 、W hz 、W hh and W γ And the bias parameter b r 、b z 、b h and b γ The network is updated and the state of the entire network when the total loss is minimized is recorded and saved as the optimal model. The optimal model is then used as the pre-trained game decision model and saved in the dynamic game decision module 7.

[0111] The dynamic game decision module 7 is used to use the pre-trained game decision model to predict the pre-processed current flight data, obtain the predicted flight data at the next moment, and generate a state forecast based on it; and manually confirm and adjust the state forecast to generate corresponding decision instructions and send them to the planning executor 8.

[0112] The predicted flight data at the next moment can be updated to obtain the individual status information of each UAV at the next moment, which is the status forecast; including: position: adjusting the expected position; speed: updating the speed according to the current mission requirements and environmental information; energy level: reflecting the battery level and endurance of the UAV to ensure effective management of resources during mission execution; other key parameters: such as sensor status, mission priority, etc., accurately reflect the dynamic environment.

[0113] The planning executor 8 is used to use the internal execution module 10 to receive decision instructions and send corresponding operation instructions to achieve the synergy effect and adaptive adjustment planning of the drone cluster.

[0114] Decision instructions include: Synergy: comprehensively consider the synergy effect of drones to optimize task allocation and paths; Task allocation: dynamically adjust tasks according to drone status and cluster needs to improve efficiency; Dynamic adjustment: allows real-time adjustment to flexibly respond to environmental changes and emergencies.

[0115] Assign locations and tasks to the best coordinated drones. Place uncoordinated drones in low-power mode and on standby for emergencies. Adjust drone speeds to appropriate levels based on assigned tasks and locations.

[0116] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A ground-air coordinated UAV dynamic game decision-making management planning system, characterized by: The system includes: a drone array state sensor group located in the air, a decision management center located on the ground, and a planning executor; the decision management center located on the ground includes a sensor database, a data preprocessing module, a spatiotemporal graph module, and a dynamic game decision module; The drone array status sensor group located in the air is used to detect the real-time position and individual status of the drone array, and transmit the newly monitored current flight data to the sensor database in the decision-making management center located on the ground; The drone array consists of The system consists of drones, each of which is equipped with a corresponding status sensor; The sensor database is used to receive the current flight data obtained by new monitoring and store historical flight data, and then transfer the current flight data and historical flight data to the data preprocessing module; The sensor database receives the current flight data obtained through new monitoring and stores historical flight data, specifically: The sensor database receives the current flight data obtained by new monitoring, and the current flight data is ,in, The first in the drone array The status sensor on the drone is Flight data sent back at each moment, , Indicates the current moment; The sensor database stores historical flight data, which is : ； in, The first in the drone array The status sensor on the drone is The flight data set returned at each moment, For the The drone in Flight data returned at each moment; The data preprocessing module is used to perform preprocessing operations on the current flight data and the historical flight data respectively, wherein the preprocessing operations include missing value filling and data cleaning, to obtain preprocessed current flight data and historical flight data, and then transfer the preprocessed historical flight data to the spatiotemporal graph module, and transfer the preprocessed current flight data to the dynamic game decision module; The spatiotemporal graph module is used to perform mutual information evaluation on the pre-processed historical flight data to screen variables, obtain the adjacency matrix using the maximum nearest neighbor method, train the spatiotemporal graph model using the GCN graph learning module and the GRU time series learning module, and save the optimal model as the pre-trained game decision model and save it in the dynamic game decision module; The dynamic game decision module is used to use the pre-trained game decision model to predict the pre-processed current flight data, obtain the predicted flight data for the next moment, and generate a state forecast based on it; and manually confirm and adjust the state forecast to generate corresponding decision instructions and send them to the planning executor; The planning executor is used to use the internal execution module to receive decision instructions and send corresponding operation instructions to achieve the synergy effect and adaptive adjustment planning of the drone cluster.

2. The ground-air coordinated UAV dynamic game decision-making management planning system according to claim 1 is characterized in that: The data preprocessing module performs preprocessing operations on the current flight data and the historical flight data respectively. The preprocessing operations include missing value filling and data cleaning to obtain the preprocessed current flight data and historical flight data. Specifically, The data preprocessing module converts historical flight data Perform preprocessing operations, including missing value filling and data cleaning, to obtain preprocessed historical flight data : ； in, For the A collection of historical flight data after pre-processing operations on UAVs; For the The drone in Historical flight data after preprocessing operations at each moment; The missing value filling is specifically as follows: The drone in Flight data returned at each moment Any variable in If there is a missing value, the variables before and after k moments are used to fill the missing value through mathematical calculation to obtain the variable after the missing value is filled. , the formula is as follows: ； The data cleaning is specifically to delete the flight data of each drone at every moment. Complete data cleaning based on the erroneous data in the data; The data preprocessing module converts the current flight data into Perform preprocessing operations to obtain the preprocessed current flight data : ,in, For the On the first drone The current flight data after the preprocessing operation at this moment.

3. The ground-air coordinated UAV dynamic game decision-making management planning system according to claim 2 is characterized in that: The spatiotemporal graph module performs mutual information evaluation on the pre-processed historical flight data to screen variables, obtains the adjacency matrix using the maximum nearest neighbor method, and trains the spatiotemporal graph model using the GCN graph learning module and the GRU time series learning module. The optimal model is saved as the pre-trained game decision model and is saved in the dynamic game decision module. Specifically, the steps include: (a.1) The key component of the spatiotemporal graph model is the spatiotemporal graph neural network, which is divided into three key elements: the time domain, the spatial domain, and the graph structure; Graph structure: Graph construction involves two steps: screening drones that are highly relevant to flight missions and determining the edge relationships between these drones to generate the adjacency matrix required by GCN. Spatial domain: Graph convolutional neural networks are a type of neural network specifically designed to process graph-structured data. They learn the interrelationships between different drone flight data from a spatial perspective. They capture the spatial dependencies between drones by propagating information between each drone and its neighbors in the graph. Temporal domain: The gated recurrent unit (GRU) is a recurrent neural network used for time series analysis. It introduces a gating mechanism to control information flow from a temporal perspective to capture the dynamic changes in the flight data of each drone over time. First, the preprocessed historical flight data The mutual information value is obtained by evaluating the historical flight data set after preprocessing operations for all arbitrary two UAVs A and B. , the calculation formula is as follows: ； in, , , ; A collection of historical flight data after preprocessing The number of grids divided in the direction, A collection of historical flight data after preprocessing The number of grids divided in the direction; for and The joint probability of for The marginal probability of for The marginal probability of Control parameters for variables; Arrange the mutual information values ​​corresponding to all two arbitrary drones from large to small, select the historical flight data set after preprocessing corresponding to the first S drones with the largest mutual information values ​​as the key historical flight data set, and obtain the key historical flight data : ； in, Indicates the A collection of key historical flight data, ; (a.2) Use the maximum neighbor method to analyze key historical flight data Construct edge relationships and obtain edge weight matrix : ； in, Represents a collection of key historical flight data and key historical flight data sets The edge weights between ;when hour, ,in, represents the bandwidth of the Gaussian kernel function; when hour, ; Key historical flight data The similarity is obtained with each key historical flight data set The K other key historical flight data sets with the greatest similarity are in the edge weight matrix Only keep key historical flight data sets The edge weights of the corresponding K other key historical flight data sets are modified to 0, and the graph structure is obtained ; Then the graph structure Converted into adjacency matrix structure , get key historical flight data Image : ,in, Shows key historical flight data The time step relationship in (a.3) Then the figure Input into the GCN graph learning module, use the graph convolution network to perform convolution calculation, and get the output of the GCN graph learning module : ; in, is the activation function; From the picture Calculated by Laplace: ； in, It's a picture The node degree matrix of Is the convolution kernel function: ; in, are the polynomial parameters, are the Chebyshev polynomial parameters, For key historical flight data The calculated diagonal matrix; (a.4) Output From time 0 to time The sub-output consists of: ,in, Indicates the Output at each moment; and key historical flight data By time 0 to time Divide into ,in, Indicates the A collection of key historical flight data at each moment; It will then output and key historical flight data Input to the GRU timing learning module to capture temporal dependencies based on the gated recurrent unit; ； ； ； ； in, is the Hadamard symbol; Indicates the The output of the update gate at this moment; Indicates the The output of the reset gate at this moment; For the The hidden state at a moment; For the The hidden state at a moment; For the The candidate hidden state at the moment; 、 、 、 、 and are the learnable parameters respectively; 、 and are bias parameters respectively; After the above calculations, we get Time domain convolution output at each moment : , where is a learnable parameter, is the bias parameter; (a.5) Then according to Time domain convolution output at each moment With the Key historical flight data collection at each moment Calculate the mean square error to get the total loss; based on the total loss, the learnable parameters 、 、 、 、 、 and And the bias parameters 、 、 and Update the network and record the state of the entire network when the total loss is minimized, and save it as the optimal model; then use the optimal model as the pre-trained game decision model and save it in the dynamic game decision module.

Images