Wargame deduction result data analysis method

Abstract

The invention discloses a war game deduction result data analysis method. The method comprises the following steps: acquiring multi-source deduction data for war game deduction; preprocessing the multi-source deduction data to obtain a preprocessed standardized data set; constructing a multi-dimensional feature matrix based on the standardized data set, and performing data association analysis on the multi-dimensional feature matrix to obtain an element association relationship matrix; based on the element incidence relation matrix, key feature factors influencing the deduction result are identified through a data mining algorithm, and the influence weight of each key feature factor is calculated; and comprehensively evaluating the deduction result data according to the influence weight to obtain influence factor analysis data of the deduction result. The technical problem of low availability of war game deduction result data analysis is solved.

Description

Technical Field

[0001] This invention relates to the field of wargaming, and more specifically, to a method for analyzing wargaming results data. Background Technology

[0002] Wargaming is a crucial tool in military simulation research. By constructing models of elements such as combat units, geographical environments, and command systems, it simulates the execution process and outcomes of different combat plans, thereby evaluating combat effectiveness and the rationality of decision-making. With the development of intelligent warfare and information-based operations, wargaming systems can generate massive amounts of dynamic data, including combat action data, equipment status data, geographical environment data, and command decision-making data. This data is complex in dimensions, highly correlated, and suffers from problems such as multi-source heterogeneity and temporal misalignment.

[0003] Existing methods for analyzing wargaming results typically rely on empirical assessments or simple statistical comparisons, making it difficult to extract deep relationships between elements from multidimensional data and lacking the ability to quantitatively analyze the factors influencing the simulation results. Especially in complex battlefield environments, where nonlinear coupling effects exist between different elements, traditional methods struggle to accurately identify key influencing factors and their mechanisms of action, resulting in insufficient scientific rigor and interpretability of the analysis.

[0004] Furthermore, existing technologies generally employ static parameter configurations in data preprocessing, feature fusion, and result evaluation, failing to adaptively adjust the model structure according to dynamic changes in the simulation scenario. Consequently, the analysis results exhibit poor robustness to environmental and task variations. As wargaming evolves towards intelligence and real-time capabilities, there is an urgent need for an analytical method capable of integrating multi-source heterogeneous data, automatically identifying key influencing factors, and providing interpretable result evaluation.

[0005] There is currently no effective solution to the above problems. Summary of the Invention

[0006] This invention provides a method for analyzing wargaming simulation results, which at least solves the technical problem of inaccurate wargaming simulation results.

[0007] According to one aspect of the present invention, a method for analyzing wargaming results data is provided, comprising: acquiring multi-source wargaming data for wargaming, wherein the multi-source wargaming data includes at least one of the following: combat action data, logistical support action data, equipment status data, geographical environment data, and command and decision data; preprocessing the multi-source wargaming data to obtain a preprocessed standardized dataset, wherein the preprocessing includes at least one of the following: data cleaning, time synchronization, and feature standardization; constructing a multi-dimensional feature matrix based on the standardized dataset, and performing data correlation analysis on the multi-dimensional feature matrix to obtain an element correlation matrix, wherein the multi-dimensional feature matrix is ​​used to characterize the relationships between multi-dimensional elements in the wargaming process; based on the element correlation matrix, using a data mining algorithm to identify key feature factors affecting the wargaming results, and calculating the influence weight of each key feature factor; and comprehensively evaluating the wargaming results data according to the influence weights to obtain analysis data on the influencing factors of the wargaming results.

[0008] Optionally, according to another aspect of the present invention, a wargaming result data analysis device is also provided, comprising: an acquisition module configured to acquire multi-source wargaming data for wargaming, wherein the multi-source wargaming data includes at least one of the following: combat action data, equipment status data, geographical environment data, and command decision data; a preprocessing module configured to preprocess the multi-source wargaming data to obtain a preprocessed standardized dataset, wherein the preprocessing includes at least one of the following: data cleaning, time synchronization, and feature standardization processing; a construction module configured to construct a multi-dimensional feature matrix based on the standardized dataset, and perform data association analysis on the multi-dimensional feature matrix to obtain an element association matrix, wherein the multi-dimensional feature matrix is ​​used to characterize the relationships between multi-dimensional elements in the wargaming process; a weight calculation module configured to identify key feature factors affecting the wargaming results using a data mining algorithm based on the element association matrix, and calculate the influence weight of each key feature factor; and a wargaming module configured to comprehensively evaluate the wargaming result data according to the influence weights to obtain analysis data on the influencing factors of the wargaming results.

[0009] In this embodiment of the invention, multi-source simulation data for wargaming is acquired, wherein the multi-source simulation data includes at least one of the following: combat action data, equipment status data, geographical environment data, and command and decision data; the multi-source simulation data is preprocessed to obtain a preprocessed standardized dataset, wherein the preprocessing includes at least one of the following: data cleaning, time synchronization, and feature standardization; based on the standardized dataset, a multi-dimensional feature matrix is ​​constructed, and data correlation analysis is performed on the multi-dimensional feature matrix to obtain an element correlation matrix, wherein the multi-dimensional feature matrix is ​​used to characterize the relationships between multi-dimensional elements in the simulation process; based on the element correlation matrix, a data mining algorithm is used to identify key feature factors affecting the simulation results, and the influence weight of each key feature factor is calculated; the simulation result data is comprehensively evaluated according to the influence weights to obtain the influencing factor analysis data of the simulation results. The above solution solves the technical problem of inaccurate wargaming results. Attached Figure Description

[0010] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0011] Figure 1 This is a flowchart of an optional wargaming result data analysis method according to an embodiment of the present invention;

[0012] Figure 2 This is a flowchart of another optional wargaming result data analysis method according to an embodiment of the present invention;

[0013] Figure 3 This is a schematic diagram of the structure of an optional wargaming result data analysis device according to an embodiment of the present invention;

[0014] Figure 4 A schematic diagram of the structure of a computer device suitable for implementing embodiments of the present disclosure is shown. Detailed Implementation

[0015] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0016] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0017] According to an embodiment of the present invention, a method embodiment for analyzing wargaming simulation results data is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0018] Figure 1 This is a method for analyzing wargaming simulation results according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0019] Step S102: Obtain multi-source simulation data for wargaming, wherein the multi-source simulation data includes at least one of the following: combat action data, equipment status data, geographical environment data, and command and decision data.

[0020] Step S104: Preprocess the multi-source inference data to obtain a preprocessed standardized dataset, wherein the preprocessing includes at least one of the following: data cleaning, time synchronization and feature standardization.

[0021] Step S106: Based on the standardized dataset, construct a multidimensional feature matrix and perform data association analysis on the multidimensional feature matrix to obtain an element association matrix, wherein the multidimensional feature matrix is ​​used to characterize the relationships between multidimensional elements in the inference process.

[0022] First, a multidimensional feature matrix is ​​constructed based on the standardized dataset. For example, feature extraction is performed on the combat operation data in the standardized dataset to obtain combat feature vectors; state feature modeling is performed on the equipment status data in the standardized dataset to obtain equipment feature vectors; spatial feature encoding is performed on the geographical environment data in the standardized dataset to obtain environmental feature vectors; semantic parsing and strategy feature extraction are performed on the command and decision data in the standardized dataset to obtain decision feature vectors; the combat feature vectors, equipment feature vectors, environmental feature vectors, and decision feature vectors are then fused in a unified feature space to obtain the multidimensional feature matrix.

[0023] Then, data association analysis is performed on the multidimensional feature matrix to obtain the element association matrix. For example, based on the multidimensional feature matrix, the correlation coefficient and mutual information value between each feature dimension are calculated to obtain the initial feature association matrix; the initial feature association matrix is ​​weighted and normalized to obtain the weighted association matrix; a feature association graph model is constructed based on the weighted association matrix, and the connection strength between nodes is calculated to obtain the node connection strength matrix; the node connection strength matrix is ​​matrix-mapped and formatted to obtain the element association matrix.

[0024] Step S108: Based on the element association matrix, use data mining algorithms to identify key feature factors that affect the inference results, and calculate the influence weight of each key feature factor.

[0025] First, based on the element association matrix, data mining algorithms are used to identify key feature factors affecting the inference results, and the influence weight of each key feature factor is calculated. For example, based on the element association matrix, a clustering analysis algorithm is used to cluster similar features to obtain feature clustering results; the importance of each feature in the feature clustering results is evaluated using a feature importance calculation model to obtain a feature importance ranking; a set of key feature factors is selected according to the feature importance ranking, and a gradient boosting-based weighted calculation model is applied to the set of key feature factors to calculate the influence weight of each key feature factor.

[0026] Then, the importance of each feature in the feature clustering results is evaluated using a feature importance calculation model to obtain a feature importance ranking. For example, the feature clustering results are input into a feature evaluation model based on random forest, and the importance index of each feature is calculated by the node split gain and the change in model error; the importance index is normalized and ranked to obtain the feature importance ranking result.

[0027] Step S110: Based on the influence weights, comprehensively evaluate the simulation result data to obtain the influencing factor analysis data of the simulation result.

[0028] For example, the inference result data and the influence weight are weighted and fused to obtain a weighted feature matrix, and the weighted feature matrix is ​​statistically analyzed and normalized to obtain a comprehensive influence score; an influence factor distribution map of the inference result is generated based on the comprehensive influence score, and the influence factor analysis data of the inference result is output according to the influence factor distribution map.

[0029] Figure 2 This is a method for analyzing wargaming simulation results according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes the following steps:

[0030] Step S202: Obtain multi-source simulation data.

[0031] The distributed data acquisition module synchronously acquires different types of data streams from multiple subsystems within the wargaming platform. This data includes operational data, logistical support data, equipment status data, geographical environment data, and command and decision-making data.

[0032] Operational data and logistical support data reflect the mission sequence, troop deployment, fire strikes, maneuver and support operations, and logistical support operations during the simulation process. This data is generated in real time by the event logging module of the simulation system, and each data packet contains a timestamp, event type, execution unit identifier, and a set of operational parameters.

[0033] To ensure data accuracy, the system introduces a timing synchronization mechanism during the acquisition phase, employing an improved time alignment function:

[0034]

[0035] in, This is the original data collection time. For reference system clock, This is the synchronization weighting coefficient (range 0–1). This mechanism can automatically correct time offsets between different systems, ensuring the time sequence consistency of various types of data.

[0036] Equipment status data originates from the weapon platform simulation module, recording the performance parameters of each piece of equipment, such as fuel reserves, ammunition status, detection capabilities, and communication latency. The system employs a dynamic sampling frequency control algorithm, automatically adjusting the sampling interval based on equipment type and mission phase, thereby reducing redundant data while maintaining the integrity of key features.

[0037] The geographic environment data is provided by the environmental modeling module and includes terrain elevation, landform type, climate conditions, and battlefield obstacle information. This data is stored in a raster format, and the system achieves fast read and write speeds through a block-based caching mechanism. To improve the accuracy of terrain features, this invention introduces a geographic resolution adaptive strategy during the data acquisition phase, automatically adjusting the sampling accuracy based on the terrain change rate to ensure higher data resolution in complex terrain areas.

[0038] Command decision-making data originates from the command agent module and includes decision-making instructions, plan templates, and response strategies from commanders at different levels. The system converts these decision logs into structured feature vectors using a natural language processing module. This module then utilizes an instruction template matching algorithm to identify keywords, combat targets, and instruction categories, ultimately forming a quantifiable command strategy data stream.

[0039] Step S204, data preprocessing.

[0040] This step includes three sub-processes: data cleaning, time synchronization, and feature standardization. Its goal is to generate a standardized dataset with controllable quality.

[0041] First, the system performs a data cleaning process.

[0042] In response to anomalous events in combat operation data, the system uses a density-based anomaly detection algorithm (DBSCAN) to detect low-frequency or isolated event points.

[0043] For missing values ​​in equipment status data, the system uses a bidirectional temporal interpolation algorithm to repair them.

[0044]

[0045] in, Indicates the interpolated value. and These are the effective values ​​for the preceding and following time steps, respectively. The weighting parameter is denoted as . Unlike traditional linear interpolation, this method introduces smoothing weights in the time dimension, which can adaptively adjust the interpolation trend.

[0046] During the time synchronization phase, the system precisely aligns the data streams of different modules based on a unified master clock. A multi-channel time drift correction algorithm is employed, using minimum mean square error to optimize the solution for the time offset.

[0047]

[0048] in, This is the original timestamp. For reference time, This is the optimal offset.

[0049] After the calculation is completed, the system batch corrects the time field to ensure that all data has a unified time series reference.

[0050] During the feature standardization phase, the system employs an adaptive normalization function for different feature dimensions:

[0051]

[0052] in, These are the standardized eigenvalues. For the first The mean of each feature, Standard deviation This is the proportionality coefficient (usually taken as 0.01–0.05).

[0053] This improved formula, by introducing an average value adjustment term, can maintain numerical stability in features with significant distribution shifts and avoid the amplification effect of eigenvalue extrema.

[0054] After completing the above processing, the system integrates all processing results into a unified standardized dataset. This dataset structurally stores four types of information: combat operations, equipment status, geographical environment, and command decisions. Each record is accompanied by a standardized feature vector, a unified timestamp, and a data source identifier, providing complete input for subsequent multi-dimensional feature modeling.

[0055] Step S206: Construct a multidimensional feature matrix.

[0056] In this step, the system constructs a multidimensional feature matrix based on the standardized dataset to characterize the relationships between multidimensional elements in the deduction process.

[0057] First, the system performs feature extraction on various types of data.

[0058] For operational data, the system calculates indicators such as operational frequency, mission duration, maneuver speed, and firepower intensity based on the sequence of operations and combat phases. These indicators constitute the operational feature vector. .

[0059] To enhance the ability to represent dynamic battlefields, this invention introduces an action complexity index:

[0060]

[0061] in, For the first Number of operational nodes in a phase For the distance of action, For resource consumption, This represents the total number of stages. This indicator dynamically reflects the non-linear relationship between the pace of action and resource consumption.

[0062] For equipment status data, the system captures the trend of equipment performance changes through time series modeling and extracts equipment lifespan ratio, health index and mission adaptability index.

[0063] This invention proposes an improved equipment health model:

[0064]

[0065] in, For a moment Equipment health This represents the performance degradation. This is the maximum performance value. The attenuation coefficient is... This represents random interference noise.

[0066] This model can effectively describe the exponential decay of equipment performance with mission pressure.

[0067] For geographic environment data, the system employs a terrain grid coding method to convert two-dimensional geographic data into a three-dimensional feature description. Each geographic unit is defined by a height gradient. Obstacle density and surface accessibility Composition. The environmental feature vector is represented as... .

[0068] In addition, the system learns spatial features from geographic rasters through a convolutional feature extraction network, further enhancing its ability to represent complex terrain.

[0069] For command and decision-making data, the system uses semantic parsing and strategy feature extraction models to identify elements such as command intent, decision-making level, and response delay.

[0070] Decision feature vectors are formed through embedding vector representation. .

[0071] The embedding process employs an improved Transformer model, using a multi-head attention mechanism to capture the relationships between different instructions.

[0072] During the feature fusion stage, the system maps the above four types of feature vectors to a unified feature space and synthesizes them using a nonlinear fusion function:

[0073]

[0074] in, For activation function, to The feature weight matrix, For equipment feature vectors, This is a bias term.

[0075] Through nonlinear fusion functions, the system can capture high-order interaction relationships between different elements.

[0076] Finally, the system stacks the fused outputs according to the time series to form a multidimensional feature matrix. Each row corresponds to a comprehensive state vector for a deduction time step.

[0077] This matrix not only provides input for subsequent correlation analysis, but can also serve as input features for machine learning models to predict or evaluate the effectiveness of different combat strategies.

[0078] Step S208, data correlation analysis.

[0079] First, the system is based on a multidimensional feature matrix. Calculate the initial correlation degree between features. To balance linear and nonlinear relationships, this invention employs an improved method for calculating the composite correlation coefficient:

[0080]

[0081] in, Indicates the first Features and the first The comprehensive correlation value of the features; Pearson correlation coefficient, used to describe linear correlation; These are mutual information coefficients, used to measure nonlinear dependence; The weighting coefficients are used to balance the system (ranging from 0 to 1). This improved model maintains high stability even when there is a non-linear shift in the feature relationships.

[0082] Subsequently, the system will combine all feature pairs. Value composition feature association initial matrix .

[0083] To avoid the influence of different characteristic dimensions on the analysis results, the system... After weighted normalization, the weighted correlation matrix is ​​obtained. The calculation formula is as follows:

[0084]

[0085] in, and These are the adaptive weights for the corresponding features, determined by the inverse ratio of the feature variance, to improve stability.

[0086] Based on this, the system constructs a feature association graph model. Among them, the node set Represents the feature dimension, edge set This represents the weighted correlation between nodes. The weight value of each edge corresponds to... .

[0087] To measure the interaction strength between features, this application proposes an adaptive node connection strength model, the core formula of which is:

[0088]

[0089] in, For nodes With nodes The strength of the connection between them; It is the smoothing exponent (typically ranging from 1.1 to 1.5). To prevent stable terms with a denominator of zero; This represents the total number of nodes.

[0090] Unlike traditional normalized weight models, this formula adaptively amplifies the contribution of weakly correlated nodes through power-law adjustment, making the model more sensitive to potential key factors.

[0091] The system calculates the node connection strength matrix. Then, it is subjected to matrix mapping and formatting to generate the final element relationship matrix. .matrix elements Representation of features With features The global correlation between them is represented by a symmetric matrix structure, with all elements on the main diagonal being 1.

[0092] This matrix not only visualizes the network of relationships between elements, but can also be used for downstream data mining and clustering calculations.

[0093] Step S210: Identify key feature factors and calculate their influence weights.

[0094] First, the system analyzes the element relationship matrix. Perform feature clustering analysis.

[0095] This embodiment employs an improved density-based clustering algorithm (Improved-DBSCAN), which no longer uses Euclidean distance as the distance metric, but instead uses a similarity function constructed from association strength:

[0096]

[0097] in, This is the clustering smoothing coefficient. This improvement allows the feature clustering process to better reflect the actual interdependencies between elements.

[0098] After clustering, the system calculates the center vector of features within each cluster. The cluster centers are calculated using a weighted average method:

[0099]

[0100] in, Indicates the first A set of indices for feature clusters For feature correlation vector, These are the initial weights for feature importance. The correlation between cluster centers can be used as a preliminary indicator of feature importance.

[0101] Next, the system uses a feature importance calculation model based on random forests to evaluate the contribution of each feature to the prediction results. The random forest model consists of multiple decision trees, and the system calculates the split gain for each feature. and error change rate The feature importance index is defined as follows:

[0102]

[0103] in, To prevent small constants with a denominator of zero.

[0104] Compared with traditional models, this index takes into account both information gain and error sensitivity, and can more accurately reflect the actual contribution of features to prediction performance.

[0105] The system is based on each feature Sort the values ​​and filter out the top ones. The features are used as the set of key feature factors. Then, a gradient boosting weighted model (GBWM) is used to calculate the influence weight of each key feature factor. The core formula is:

[0106]

[0107] in, This is the smoothing adjustment coefficient.

[0108] The model uses Softmax normalization to map feature importance into a probability distribution, thereby obtaining the relative influence weight of each feature.

[0109] Unlike traditional linear weighting, this method can create a non-linear competition mechanism among key features, making the weights of salient features more prominent.

[0110] Step S212: Comprehensive evaluation of the simulation results.

[0111] First, the system weights and fuses the simulation results data (including win / loss determination indicators, resource consumption indicators, task completion rate, etc.) with the influence weights to generate a weighted feature matrix. The calculation method is as follows:

[0112]

[0113] in, For time step Time of the first Observations of key features, The corresponding influence weights are shown. This matrix represents the combined contribution of each key factor at different time stages.

[0114] The system performs statistical analysis and normalization on the weighted feature matrix, and calculates the comprehensive influence score:

[0115]

[0116] in, For time steps, and These represent the maximum and minimum values ​​of the weighted matrix, respectively. This score directly reflects the overall effectiveness of the entire combat process.

[0117] Subsequently, the system generates a distribution map of influencing factors based on the comprehensive score. This distribution map uses a two-dimensional heatmap to show the degree of influence of each key characteristic factor at different stages of the simulation.

[0118] To further improve interpretability, this invention proposes a "multidimensional causal chain visualization algorithm" that transforms influence paths into a causal chain graph structure:

[0119] in, Representation of features Features The algorithm reveals the strength of causal influence. It can uncover the hierarchical relationships between different features, helping analysts identify the core mechanisms leading to differences in results.

[0120] Finally, the system synthesizes all analysis results and generates influencing factor analysis data for the projection results. This analysis data includes: a list of key characteristic factors, a weight distribution table, a phased comprehensive score curve, an influence path diagram, and an overall assessment conclusion.

[0121] This application also provides a wargaming simulation result data analysis device, such as... Figure 3As shown, the system includes: an acquisition module 32, configured to acquire multi-source simulation data for wargaming, wherein the multi-source simulation data includes at least one of the following: combat action data, equipment status data, geographical environment data, and command decision data; a preprocessing module 34, configured to preprocess the multi-source simulation data to obtain a preprocessed standardized dataset, wherein the preprocessing includes at least one of the following: data cleaning, time synchronization, and feature standardization; a construction module 36, configured to construct a multi-dimensional feature matrix based on the standardized dataset, and perform data association analysis on the multi-dimensional feature matrix to obtain an element association matrix, wherein the multi-dimensional feature matrix is ​​used to characterize the relationships between multi-dimensional elements in the simulation process; a weight calculation module 38, configured to identify key feature factors affecting the simulation results using a data mining algorithm based on the element association matrix, and calculate the influence weight of each key feature factor; and a simulation module 39, configured to comprehensively evaluate the simulation result data according to the influence weights to obtain analysis data on the influencing factors of the simulation results.

[0122] It should be noted that the wargame simulation result data analysis device provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the wargame simulation result data analysis device and the wargame simulation result data analysis method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0123] Figure 4 A schematic diagram of a computer device suitable for implementing embodiments of the present disclosure is shown. It should be noted that... Figure 4 The computer device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0124] like Figure 4 As shown, the computer device includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage section 1008 into a random access memory (RAM) 1003. The RAM 1003 also stores various programs and data required for system operation. The CPU 1001, ROM 1002, and RAM 1003 are interconnected via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0125] The following components are connected to I / O interface 1005: an input section 1006 including a keyboard, mouse, etc.; an output section 1007 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to I / O interface 1005 as needed. A removable medium 1011, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 1010 as needed so that computer programs read from it can be installed into storage section 1008 as needed.

[0126] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for analyzing wargaming simulation results, characterized in that, include: Acquire multi-source simulation data for wargaming, wherein the multi-source simulation data includes at least one of the following: combat action data, logistical support action data, equipment status data, geographical environment data, and command and decision data; The multi-source inference data is preprocessed to obtain a preprocessed standardized dataset, wherein the preprocessing includes at least one of the following: data cleaning, time synchronization and feature standardization. Based on the standardized dataset, a multidimensional feature matrix is ​​constructed, and data association analysis is performed on the multidimensional feature matrix to obtain an element association matrix. The multidimensional feature matrix is ​​used to characterize the relationships between multidimensional elements in the inference process. Based on the aforementioned element correlation matrix, data mining algorithms are used to identify key feature factors that affect the inference results, and the influence weight of each key feature factor is calculated. The inference results data are comprehensively evaluated based on the aforementioned influence weights to obtain the influencing factor analysis data of the inference results.

2. The method according to claim 1, characterized in that, Based on the standardized dataset, a multidimensional feature matrix is ​​constructed, including: Feature extraction is performed on the combat operation data and logistical support operation data in the standardized dataset to obtain operation feature vectors; The equipment status data in the standardized dataset are modeled for status features to obtain equipment feature vectors. Spatial feature encoding is performed on the geographic environment data in the standardized dataset to obtain an environmental feature vector; Semantic parsing and strategy feature extraction are performed on the command and decision data in the standardized dataset to obtain a decision feature vector; The action feature vector, the equipment feature vector, the environment feature vector, and the decision feature vector are fused in a unified feature space to obtain the multidimensional feature matrix.

3. The method according to claim 1, characterized in that, Data association analysis is performed on the multidimensional feature matrix to obtain the element association matrix, including: Based on the multidimensional feature matrix, the correlation coefficient and mutual information value between each feature dimension are calculated to obtain the initial feature association matrix; The initial feature association matrix is ​​weighted and normalized to obtain a weighted association matrix; Based on the weighted correlation matrix, a feature correlation graph model is constructed, and the connection strength between nodes is calculated to obtain the node connection strength matrix; The node connection strength matrix is ​​matrix-mapped and formatted to obtain the element association matrix.

4. The method according to claim 1, characterized in that, Based on the aforementioned element correlation matrix, data mining algorithms are used to identify key feature factors affecting the inference results, and the influence weight of each key feature factor is calculated, including: Based on the aforementioned element association matrix, a clustering analysis algorithm is used to cluster similar features to obtain feature clustering results. The importance of each feature in the feature clustering results is evaluated using a feature importance calculation model to obtain a ranking of feature importance. The key feature factors are selected based on the importance of the features, and a gradient boosting-based weighted calculation model is applied to the key feature factors to calculate the influence weight of each key feature factor.

5. The method according to claim 4, characterized in that, The importance of each feature in the feature clustering results is evaluated using a feature importance calculation model to obtain a feature importance ranking, including: The clustering results of the features are input into a feature evaluation model based on random forest, and the importance index of each feature is calculated by the node splitting gain and the change in model error. The importance indicators are normalized and sorted to obtain the feature importance ranking results.

6. The method according to claim 1, characterized in that, Based on the aforementioned influence weights, a comprehensive evaluation of the projection results data is performed to obtain the influencing factor analysis data for the projection results, including: The inference results data and the influence weights are weighted and fused to obtain a weighted feature matrix. The weighted feature matrix is ​​then statistically analyzed and normalized to obtain a comprehensive influence score. Based on the comprehensive impact score, an impact factor distribution map of the inference results is generated, and the impact factor analysis data of the inference results is output according to the impact factor distribution map.

7. A wargaming simulation result data analysis device, characterized in that, include: The acquisition module is configured to acquire multi-source simulation data for wargaming, wherein the multi-source simulation data includes at least one of the following: combat action data, logistical support action data, equipment status data, geographical environment data, and command and decision data. The preprocessing module is configured to preprocess the multi-source inference data to obtain a preprocessed standardized dataset, wherein the preprocessing includes at least one of the following: data cleaning, time synchronization, and feature standardization. The construction module is configured to construct a multidimensional feature matrix based on the standardized dataset, and perform data association analysis on the multidimensional feature matrix to obtain an element association matrix, wherein the multidimensional feature matrix is ​​used to characterize the relationships between multidimensional elements in the inference process; The weight calculation module is configured to identify key feature factors affecting the inference results based on the element association matrix and using data mining algorithms, and to calculate the influence weight of each key feature factor. The evaluation module is configured to comprehensively evaluate the simulation result data based on the influence weights to obtain the influencing factor analysis data of the simulation result.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform the method according to any one of claims 1 to 6.

9. A computer device, characterized in that, include: Memory and processor The memory stores computer programs; The processor is configured to execute a computer program stored in the memory, wherein when the computer program is executed, the processor performs the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images