VR simulation data intelligent classification method

Abstract

The present application relates to the technical field of data processing, and particularly relates to a VR simulation data intelligent classification method; the scheme of the present application can acquire a plurality of keywords in a historical VR simulation scene, acquire corresponding features of the keywords, calculate the similarity between the keywords, obtain the membership between the corresponding features and the keywords, obtain the classification weight of each keyword by using the membership between the corresponding features and the keywords, train an XGboost model by using the classification weight to obtain the importance of a decision tree, input current simulation data into the trained XGboost model, and realize classification of the current simulation data. The present application determines the final classification result of simulation data in the current VR simulation scene according to the similarity between the simulation scene keywords in the current VR simulation scene and the keywords obtained by the XGboost model, effectively improving the classification rate and accuracy of VR simulation data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a VR simulation data intelligent classification method. BACKGROUND

[0002] With the application of VR simulation technology in various industries, its types and forms are increasingly diversified. VR simulation technology constructs highly realistic virtual environments, simulates various scenes and tasks in the real world, and provides powerful tools for scientific research, education, entertainment and other fields. The construction of VR simulation environment usually involves multiple parallel tasks and complex interaction relationships, which require efficient and accurate data management to support. Data classification and hierarchical management, as a recognized effective data management method, plays an important role in VR simulation environment. Through data intelligent classification, different tasks, scenes and objects can be accurately identified, so that the VR simulation system can handle multi-task parallelism and complex situations, thereby improving simulation efficiency and accuracy.

[0003] In VR simulation tasks, there are usually multiple simulation scenarios, and each scenario has different data requirements. For example, urban traffic flow analysis involves simulation scenarios such as peak traffic flow analysis, road construction impact on traffic flow analysis, and traffic flow analysis under new traffic control measures. These scenarios have close relationships and mutual influences, and their data requirements are also unique. However, existing data classification methods often do not fully consider the correlation between VR simulation scenarios and the membership between VR simulation data and simulation scenarios. This oversight may lead to inaccurate classification of VR simulation data by the classification model, thereby affecting the reliability and effectiveness of the simulation results.

[0004] Therefore, there is an urgent need for a method that can effectively solve the problem of inaccurate classification of VR simulation data by the classification model due to the lack of consideration of the correlation between VR simulation scenarios and the membership between VR simulation data and simulation scenarios in existing technology. SUMMARY

[0005] The present application aims to provide a VR simulation data intelligent classification method to solve the problem of inaccurate classification of VR simulation data by the classification model due to the lack of consideration of the correlation between VR simulation scenarios and the membership between VR simulation data and simulation scenarios in existing technology. To this end, the present application provides the following technical solution.

[0006] The present application provides a VR simulation data intelligent classification method, comprising:

[0007] Multiple keywords from a historical VR simulation scene are obtained, and corresponding features are acquired for each keyword to calculate the similarity between keywords. The membership degree between the corresponding features and the keywords is obtained. The similarity between keywords is obtained by calculating the similarity of the feature histograms corresponding to the keywords. The similarity of the feature histograms is obtained by statistically analyzing the various features contained in the keywords. The membership degree is used to characterize the degree of association between keywords and corresponding features.

[0008] By utilizing the membership degree between the corresponding features and keywords, the classification weight of each keyword is obtained, and the importance of the decision tree is obtained through the classification weight to train the XGboost model;

[0009] The current simulation data is input into the trained XGboost model to classify the current simulation data.

[0010] In the above scheme, the present invention obtains multiple keywords from historical VR simulation scenes, acquires corresponding features for the keywords, calculates the similarity between the keywords, obtains the membership degree between the corresponding features and the keywords, uses the membership degree between the corresponding features and the keywords to obtain the classification weight of each keyword, and uses the classification weight to obtain the decision tree importance to train the XGboost model, thereby realizing the classification of the current simulation data and effectively improving the classification speed and accuracy of the current simulation data.

[0011] Optionally, the membership degree is calculated as follows:

[0012] ;

[0013] in, Let f be the membership degree of the f-th feature to the i-th keyword in the k-th simulation scenario; H represents the similarity between the feature histograms of the i-th keyword and the j-th keyword, which contain f features, in the k-th simulation scenario; k Let be the total number of keyword combinations in the k-th simulation scenario. It is a function of average value. In the k-th simulation scenario, the i-th keyword is the first of the other features after removing the f-th feature. The cosine similarity between the first feature and the features under the j-th keyword.

[0014] The above scheme calculates the membership degree to obtain the degree of association between keywords and corresponding features.

[0015] Optionally, the similarity between the feature histograms specifically includes:

[0016] ;

[0017] in, For the first The similarity of the feature histograms corresponding to the i-th keyword and the j-th keyword in a simulation scenario; and The first The feature histograms corresponding to the i-th and j-th keywords in the simulated scenario text; c is the value of the i-th keyword. The c-th feature in the feature histogram with the largest feature count among the i-th and j-th keywords in a simulation scenario; and The first The usage percentage of the c-th feature in the feature histograms corresponding to the i-th and j-th keywords in a simulation scenario.

[0018] The above scheme calculates the similarity between keywords by measuring the similarity of their corresponding feature histograms. Therefore, the similarity between keywords can intuitively reflect their similarities and differences, and by using the similarity of feature histograms, the degree of association between corresponding features and keywords can be determined more accurately.

[0019] Optionally, the calculation method for the classification weight of each keyword is as follows:

[0020] ;

[0021] in, The classification weight of the i-th keyword in the historical simulation data corresponding to the f-th feature in the historical VR simulation scene; Let f be the membership degree of the f-th feature to the i-th keyword in the k-th simulation scene; n is the total number of keywords in the k-th simulation scene; and K is the total number of historical VR simulation scenes.

[0022] The above scheme improves the accuracy of XGboost model decision-making by calculating the classification weights of each keyword.

[0023] Optionally, the method for obtaining the importance of the decision tree is as follows:

[0024] ;

[0025] in, Let v be the decision importance of the v-th decision tree in the XGboost model; The classification accuracy of the historical simulation data assigned to the v-th decision tree.

[0026] The above approach optimizes the accuracy of the XGboost model in classifying the current simulation data by calculating the importance of the decision tree.

[0027] Optionally, the classification accuracy of the historical simulation data can be obtained as follows:

[0028] ;

[0029] in, The classification accuracy of the historical simulation data assigned to the v-th decision tree; The total number of keywords for the f-th feature; The total number of types of historical simulation data assigned to the v-th decision tree; The classification weight of the i-th keyword in the historical simulation data corresponding to the f-th feature in the historical VR simulation scene; The classification accuracy of the historical simulation data corresponding to the f-th feature on the v-th decision tree as the i-th keyword.

[0030] Optionally, the keywords are obtained by segmenting the text of historical VR simulation scenes using the Jieba tool.

[0031] Optionally, the corresponding feature is a feature obtained by standardizing and normalizing the data of different dimensions in the simulation scenario.

[0032] Optionally, the XGboost model is a model built using VR simulation data.

[0033] Optionally, after the corresponding features are preprocessed, spline interpolation is used to complete the missing parts of the corresponding features.

[0034] The beneficial effects of this invention are as follows:

[0035] The present invention can acquire multiple keywords from historical VR simulation scenes, obtain corresponding features for these keywords, calculate the similarity between keywords, obtain the membership degree between the corresponding features and keywords, and use the membership degree between the corresponding features and keywords to obtain the classification weight of each keyword. The XGBoost model is then trained using the importance of the decision tree obtained from the classification weights. The current simulation data is input into the trained XGBoost model to classify the current simulation data. This invention determines the final classification result of simulation data in the current VR simulation scene by comparing the similarity between the keywords of the proposed simulation scene in the current VR simulation scene and the keywords obtained from the XGBoost model, effectively improving the classification speed and accuracy of VR simulation data. Attached Figure Description

[0036] Figure 1 The flowchart illustrating the steps of a VR simulation data intelligent classification method in this embodiment is shown in the illustration. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0039] This invention addresses a data classification and management scheme for VR simulation environments. The intelligent classification scheme for VR simulation data improves the accuracy of VR simulation data classification by calculating the similarity between keywords and obtaining the membership degree between corresponding features and keywords.

[0040] Therefore, based on the above, this invention provides an intelligent classification method for VR simulation data to improve the accuracy of VR simulation data classification.

[0041] Specifically, such as Figure 1 As shown, a VR simulation data intelligent classification method in this embodiment includes the following steps:

[0042] Step S1: Obtain multiple keywords from historical VR simulation scenes.

[0043] Specifically, VR simulation tasks typically involve multiple simulation scenarios, each with different data requirements. For example, a VR simulation scenario for urban traffic flow analysis might include scenarios such as peak-hour traffic flow analysis, the impact of road construction on traffic flow, and traffic flow under new traffic control measures. Therefore, it is necessary to extract keywords from the descriptive text of historical VR simulation scenarios and preprocess the features corresponding to these keywords. This preprocessing includes standardizing and normalizing the features corresponding to the extracted keywords. After preprocessing, if any corresponding feature values ​​are missing, spline interpolation is used to fill in the missing values ​​to facilitate feature extraction.

[0044] The corresponding features include geometric and physical features required in the simulation scene (features describing the shape, size, and position of objects in the virtual environment, as well as features simulating the physical behavior of objects in the virtual environment), perceptual features (including corresponding features such as texture, color, lighting, shadows, and sound effects in the environment), interactive features (features related to user operation commands and interactive intentions collected by input devices), and contextual and plot features (features describing the background, atmosphere, and storyline of the virtual environment, as well as defining information such as event sequences, character behaviors, and task objectives in the virtual environment). In other words, the corresponding features in this invention vary depending on the specific simulation scene being created. The keywords refer to key textual terms or phrases used when describing or constructing the simulation scene. These keywords play a crucial role in clarifying scene features, guiding scene construction, and subsequent scene applications.

[0045] For example, the simulation scenario text describes the city's traffic network in a geographic VR simulation scenario, focusing on analyzing peak-hour traffic flow. This scenario description includes: busy streets, main roads, intersections, and dynamic changes in traffic lights. Users can observe, through VR devices, the density of traffic flow on the roads, vehicle speed, and the impact of traffic signal changes on overall traffic flow during simulated peak hours; as well as environmental features, weather conditions (sunny, rainy, etc.), audio features such as traffic signal sounds and alarms, and image features such as urban spatial structure and building structure. Keywords are obtained by acquiring descriptions of corresponding features from the simulation text. The examples described in this invention are for illustrative purposes only and do not constitute any limitation on the invention.

[0046] In this embodiment, by acquiring the corresponding features and keywords required in historical VR simulation scenes, the core elements and key information of VR simulation scenes can be quickly understood, thereby avoiding blindly searching in a large amount of data. Based on historical corresponding features, new VR simulation scenes can be designed more accurately, ensuring that task scenes, parameter settings, etc. are closer to actual needs, and improving the practicality and effectiveness of tasks.

[0047] Step S2: Obtain corresponding features for the keywords to calculate the similarity between keywords, and obtain the membership degree between the corresponding features and the keywords.

[0048] Specifically, VR simulation scene text may contain multiple keywords. Therefore, it is necessary to calculate the similarity between keywords to obtain the membership degree between the corresponding features and the keywords. The similarity between keywords is obtained by calculating the similarity of the feature histograms corresponding to the keywords. The similarity of the feature histograms is obtained by statistically analyzing the various features contained in the keywords. The membership degree is used to characterize the degree of association between keywords and their corresponding features.

[0049] Step S21: Calculate the similarity between feature histograms. The specific calculation process is as follows:

[0050] ;

[0051] in, For the first The similarity of the feature histograms corresponding to the i-th keyword and the j-th keyword in a simulation scenario; and The first The feature histograms corresponding to the i-th and j-th keywords in the simulated scenario text; c is the value of the i-th keyword. The c-th feature in the feature histogram with the largest feature count among the i-th and j-th keywords in a simulation scenario; and The first The usage percentage of the c-th feature in the feature histograms corresponding to the i-th and j-th keywords in a simulation scenario.

[0052] Step S22: Calculate the membership degree between the corresponding features and keywords based on the similarity of the feature histograms.

[0053] Specifically, once the similarity of the feature histograms is obtained, i.e., the similarity of keywords, the membership degree between the corresponding features and keywords is calculated:

[0054] ;

[0055] in, Let f be the membership degree of the f-th feature to the i-th keyword in the k-th simulation scenario; H represents the similarity between the feature histograms of the i-th keyword and the j-th keyword, which contain f features, in the k-th simulation scenario; k Let be the total number of keyword combinations in the k-th simulation scenario. It is a function of average value. In the k-th simulation scenario, the i-th keyword is the first of the other features after removing the f-th feature. The cosine similarity between the first feature and the features under the j-th keyword.

[0056] Let be the mean similarity between the feature histograms of the i-th keyword and the j-th keyword containing f features in the k-th simulation scenario. The larger this value, the smaller the required feature difference between the i-th keyword and the j-th keyword containing the f-th corresponding feature, the greater the correlation between the f-th corresponding feature under the i-th keyword, and the smaller the membership degree of the i-th keyword corresponding to the f-th corresponding feature. The smaller this value, the greater the required feature difference between the i-th keyword and the j-th keyword containing the f-th corresponding feature, the smaller the correlation between the f-th corresponding feature under the i-th keyword, and the greater the membership degree of the i-th keyword corresponding to the f-th corresponding feature.

[0057] Let be the cosine similarity between the category of the corresponding feature under the i-th keyword and the corresponding feature under the j-th keyword in the k-th simulation scenario after removing the f-th corresponding feature. Let be the mean cosine similarity between the corresponding feature categories of the i-th keyword (after removing the f-th corresponding feature) and the corresponding feature categories of the other keywords in k simulation scenarios. A larger value indicates greater similarity between the corresponding feature categories of the i-th keyword (after removing the f-th corresponding feature) and the corresponding feature categories of the other keywords. The greater the downward correction, the weaker the correlation between the f-th corresponding feature under the i-th keyword, and the greater the membership degree of the VR simulation data corresponding to the f-th corresponding feature to the i-th keyword. A smaller value indicates that the corresponding feature type under the i-th keyword after removing the f-th corresponding feature is less similar to the feature types under the other keywords. The smaller the downward correction.

[0058] In this embodiment, the present invention obtains corresponding features for keywords to calculate the similarity between keywords and obtains the membership degree between corresponding features and keywords. By calculating the similarity of feature histograms corresponding to keywords, the present invention can intuitively reflect the similarity and difference between data by statistically analyzing the distribution of features corresponding to keywords. When calculating the membership degree between features corresponding to keywords and keywords, the similarity of feature histograms is used as a basis to more accurately determine the degree of association between corresponding features and keywords.

[0059] Step S3: Utilize the membership degree between the corresponding features and keywords to obtain the classification weight of each keyword, and use the classification weight to obtain the importance of the decision tree to train the XGboost model.

[0060] Specifically, the formula for calculating the classification weight of each keyword using the membership degree is as follows:

[0061] ;

[0062] in, The classification weight of the i-th keyword in the historical simulation data corresponding to the f-th feature in the historical VR simulation scene; Let f be the membership degree of the f-th feature to the i-th keyword in the k-th simulation scene; n is the total number of keywords in the k-th simulation scene; and K is the total number of historical VR simulation scenes.

[0063] The importance of the decision tree is obtained through the classification weights, and the calculation formula is as follows:

[0064] ;

[0065] in, Let v be the decision importance of the v-th decision tree in the XGboost model; The classification accuracy of the historical simulation data assigned to the v-th decision tree.

[0066] The method for obtaining the classification accuracy of the historical simulation data is as follows:

[0067] ;

[0068] in, The classification accuracy of the historical simulation data assigned to the v-th decision tree; The total number of keywords for the f-th feature; The total number of types of historical simulation data assigned to the v-th decision tree; The classification weight of the i-th keyword in the historical simulation data corresponding to the f-th feature in the historical VR simulation scene; The classification accuracy of the historical simulation data corresponding to the f-th feature on the v-th decision tree as the i-th keyword.

[0069] In this embodiment, the classification weight of each keyword is obtained by utilizing the membership degree between the corresponding features and keywords, and the decision importance of the decision tree is obtained through the classification weight. Then, the importance of the decision tree is used to train the XGboost model to improve the accuracy of the XGboost model in classifying simulation data.

[0070] Step S4: Train the XGboost model using the decision importance of the decision tree, and input the current simulation data into the trained XGboost model to classify the current simulation data.

[0071] In this embodiment, the XGboost model is trained using the decision importance of the decision tree, and the current simulation data is input into the trained XGboost model so that the trained XGboost model can classify the VR simulation data. The present invention can determine which features have the greatest impact on the classification results by extracting the decision importance from the decision tree, thereby improving the accuracy of classification and reducing unnecessary features, making the XGboost model more concise and thus improving the computational efficiency of the XGboost model.

[0072] The training process of the XGboost model is as follows: each type of VR simulation data and its corresponding keywords are input into the XGboost model so that the XGboost model can predict the classification results of the VR simulation data. The XGboost model is trained using a loss function, cross-entropy loss, and gradient magnitude. The XGboost model stops training when it reaches the set maximum number of training iterations or when the model's loss is less than a preset loss threshold.

Claims

1. A VR simulation data intelligent classification method, characterized in that, include: Multiple keywords from historical VR simulation scenes are obtained, and corresponding features are acquired for each keyword to calculate the similarity between keywords. The membership degree between the corresponding features and the keywords is then obtained. The similarity between keywords is calculated by evaluating the similarity of the feature histograms corresponding to the keywords. The similarity of the feature histograms is obtained by statistically analyzing multiple features contained in the keywords. The membership degree is used to characterize the degree of association between keywords and their corresponding features. The membership degree is: ； Let f be the membership degree of the f-th feature to the i-th keyword in the k-th simulation scenario; H represents the similarity between the feature histograms of the i-th keyword and the j-th keyword, which contain f features, in the k-th simulation scenario; k Let be the total number of keyword combinations in the k-th simulation scenario. It is a function of average value. In the k-th simulation scenario, the i-th keyword is the first of the other features after removing the f-th feature. The cosine similarity between the first feature and the feature under the j-th keyword is calculated; the classification weight of each keyword is obtained using the membership degree between the corresponding feature and the keyword, and the importance of the decision tree is obtained through the classification weight to train the XGboost model; the classification weight of each keyword is: ； Let be the classification weight of the i-th keyword in the historical simulation data corresponding to the f-th feature in the historical VR simulation scene; n is the total number of keywords in the k-th simulation scene; and K is the total number of historical VR simulation scenes. The importance of decision trees is: ； Let v be the decision importance of the v-th decision tree in the XGboost model; The classification accuracy of the historical simulation data assigned to the v-th decision tree; The classification accuracy of historical simulation data is: ； The total number of keywords for the f-th feature; The total number of types of historical simulation data assigned to the v-th decision tree; The classification accuracy of the historical simulation data corresponding to the f-th feature on the v-th decision tree as the i-th keyword is calculated; the current simulation data is input into the trained XGboost model to classify the current simulation data.

2. The intelligent classification method for VR simulation data according to claim 1, characterized in that, The similarity between the feature histograms specifically includes: ； in, For the first The similarity of the feature histograms corresponding to the i-th keyword and the j-th keyword in a simulation scenario; and The first The feature histograms corresponding to the i-th and j-th keywords in the simulated scenario text; c is the value of the i-th keyword. The c-th feature in the feature histogram with the largest feature count among the i-th and j-th keywords in a simulation scenario; and The first The usage percentage of the c-th feature in the feature histograms corresponding to the i-th and j-th keywords in a simulation scenario.

3. The intelligent classification method for VR simulation data according to claim 1, characterized in that, The keywords were obtained by segmenting the text of historical VR simulation scenes using the Jieba tool.

4. The intelligent classification method for VR simulation data according to claim 1, characterized in that, The corresponding features are the features obtained after standardizing and normalizing the data of different dimensions in the simulation scenario.

5. The intelligent classification method for VR simulation data according to claim 1, characterized in that, The XGboost model is a model built using VR simulation data.

6. The intelligent classification method for VR simulation data according to claim 4, characterized in that, After the corresponding features are preprocessed, spline interpolation is used to fill in the missing parts of the corresponding features.

Images