Intelligent interaction one-stop digital scene display method and system based on multi-source data
By employing a one-stop digital scene display method based on intelligent interaction of multi-source data, and utilizing isolated forest anomaly detection and SVM models for fine-grained data classification and semantic similarity calculation, this approach solves the problem of information silos caused by fragmented enterprise data, achieves data integration and efficient utilization, and enhances the scientific rigor and timeliness of decision-making.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-07
- Publication Date
- 2026-03-24
- Estimated Expiration
- Not applicable · inactive patent
Smart Images
Figure CN121722959A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data integration technology. More specifically, this invention relates to a one-stop digital scene display method and system based on multi-source data for intelligent interaction. Background Technology
[0002] In today's digital age, the construction of digital scenarios is a crucial driver of innovation and development across all industries. Immersive experience environments created through virtual-real fusion technology not only enable natural interaction between users and digital spaces but also allow users to break through the limitations of traditional physical interfaces. They can deeply engage with digital content through multi-dimensional interaction methods such as voice. This immersive experience that transcends time and space is redefining the standard paradigm of service delivery, freeing users from the constraints of traditional interaction methods.
[0003] Currently, enterprises possess a large volume of business data, which typically belongs to different business systems. This data fragmentation brings numerous serious problems, making it difficult to fully explore and maximize the value of the data. On the one hand, data fragmentation prevents information from flowing and integrating across systems, making it difficult for different departments or business segments to obtain the necessary information comprehensively and accurately, thus affecting the scientific and timely nature of decision-making. Furthermore, because data is scattered across different systems, data analysis and mining are difficult to conduct comprehensively and deeply. Much valuable information remains hidden in isolated data oceans, unable to be effectively extracted and utilized, thus limiting business innovation and efficiency improvement. Summary of the Invention
[0004] To address the aforementioned technical problem of difficulty in extracting data from power systems, this invention provides solutions in several aspects.
[0005] In the first aspect, a one-stop digital scene display method based on multi-source data intelligent interaction: The system acquires historical data from multiple data sources, with each source containing data points from multiple time points. Anomaly detection algorithms are used to identify anomalous data points in the historical data of each data source, and all data points other than anomalous points are considered normal data points. Evaluation metrics for each data source are calculated, where the absolute value of the difference between a normal data point in the historical data and a preset standard value is proportional to the data source's evaluation metric. The evaluation metrics and normal data points from the historical data are visualized to generate data analysis charts. In response to received user input text, the system obtains the semantic text similarity between the user's input text and preset standard terminology names from each data source. Data analysis charts for data sources with a relevance greater than a preset threshold to the input text are output.
[0006] Preferably, the anomaly detection algorithm is an isolated forest anomaly detection algorithm.
[0007] Preferred methods for intelligent interactive one-stop digital scene display based on multi-source data also include: The abnormal data points are classified into categories using SVM, which are categorized as extreme high values, extreme low values, fluctuating anomalies, and null values. The results of visualizing the number of abnormal data points in each category from the historical data of the data source are then added to the data analysis chart of the data source.
[0008] Preferably, fine-grained classification of the outlier data points using SVM includes: obtaining the feature vector of the outlier data points, where the first... i The first historical data from the data source s Feature vectors of anomalous data points X i,s satisfy: For the first i The first historical data from the data source s One abnormal data point, x i,s,1 This represents the value of the data point. x i,s,2 This indicates that the data point is the first one of the preset size. i Standard values a i The absolute value of the difference between them x i,s,3 This indicates whether the data point is null. x i,s,4 This indicates that the data point is related to the first... i The first historical data from the data source s -1 difference between outlier data points, where the data point is null. x i,s,3 The value of is 1; the first i The first historical data from the data source s Feature vectors of anomalous data points X i,s Input is given to a preset SVM model consisting of extreme high values, extreme low values, volatility anomaly SVM models, and null values; the confidence level of the output of the extreme high value SVM model, extreme low value SVM model, volatility anomaly SVM model, and null value SVM model is used to determine the first... i The first historical data from the data source s The categories of anomalous data points.
[0009] Preferably, determine the first i The first historical data from the data source s The categories of outlier data points include: the first confidence score of obtaining extremely high SVM model output values.α 1. Second confidence level of SVM model output at extremely low values α 2. Third confidence level of the output of the fluctuation anomaly SVM model α 3. The fourth confidence level of the output of the null SVM model α 4; Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The outlier data points are classified as extremely high values; in response to the first confidence level... α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The outlier data points are classified as extremely low values; in response to the first confidence level... α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The category of the anomalous data points is fluctuation anomaly; in response to the first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The category of the abnormal data points is null.
[0010] Preferably, calculate the first i The formula for the evaluation metrics of each data source is: .
[0011] in, σ i For the first i The standard deviation of all normal data points in the historical data of each data source, and, n i For the first i The number of all normal data points in the historical data of each data source. Aj,i For the first i The historical data from the data source is the first j The value of a normal data point K The adjustment coefficient is the preset size. j It is a positive integer.
[0012] Preferably, the normal data points are linearly interpolated before visualizing the normal data points of the historical data from the data source.
[0013] Preferably, the user's input text is obtained and the first i The semantic text similarity between the standard term names preset by the data source includes: preprocessing the input text and converting it into a first TF-IDF vector; and then converting the first TF-IDF vector into a second TF-IDF vector. i The data source's preset standard terminology names are converted into a second TF-IDF vector; the cosine similarity between the first TF-IDF vector and the second TF-IDF vector is calculated, and the cosine similarity is determined to be the ratio between the user's input text and the first TF-IDF vector. i Semantic text similarity between standard terminology names preset by the data source.
[0014] Preferably, preprocessing the input text includes: removing stop words from the input text and then performing word segmentation on the input text.
[0015] In the second aspect, an intelligent interactive one-stop digital scene display system based on multi-source data includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the intelligent interactive one-stop digital scene display method based on multi-source data as described in any of the above-described inventions.
[0016] The beneficial effects of this invention are as follows: This invention integrates historical data from multiple data sources, retrieves the most relevant historical data based on the user's input text, and visualizes the data, breaking down data silos and improving data flow and value utilization. Attached Figure Description
[0017] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein: Figure 1 This is a flowchart illustrating the steps of an intelligent interactive one-stop digital scene display method based on multi-source data according to an embodiment of the present invention. Figure 2This is a schematic diagram illustrating the structural block of an intelligent interactive one-stop digital scene display system based on multi-source data according to an embodiment of the present invention. Detailed Implementation
[0018] 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.
[0019] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0020] Figure 1 This is a flowchart illustrating the steps of an intelligent interactive one-stop digital scene display method based on multi-source data according to an embodiment of the present invention.
[0021] like Figure 1 As shown, the intelligent interactive one-stop digital scene display method based on multi-source data includes steps S1 to S6.
[0022] Step S1: Obtain historical data from multiple data sources.
[0023] Historical data from a single data source comprises data points from multiple points in time. It should be noted that all data points have numerical values.
[0024] In one embodiment, historical data from multiple data sources includes: historical measurement data from the SCADA system, historical PMU synchronization phasor data, historical micro-meteorological monitoring data of transmission lines, and historical data from smart meter freezes.
[0025] Step S2: Identify abnormal data points in the historical data of each data source using an anomaly detection algorithm, and determine the data points in the historical data of the data source other than the abnormal data points as normal data points.
[0026] In one embodiment, the anomaly detection algorithm is the Isolation Forest anomaly detection algorithm. The Isolation Forest anomaly detection algorithm is a tree-based algorithm that utilizes the "isolation" characteristic of data to identify anomalous data points. The Isolation Forest anomaly detection algorithm constructs trees by randomly selecting features and randomly partitioning feature values. Each tree is cut within the data space until data points are isolated. During the isolation process, anomalous data points are easier to isolate than normal data points because anomalous data points are usually more dispersed in the feature space, making it easier to find a split point to isolate them from other data points.
[0027] Furthermore, the present invention also includes: the intelligent interactive one-stop digital scene display method based on multi-source data further includes: performing fine-grained classification of the abnormal data points using SVM to obtain the categories of abnormal data points, wherein the abnormal data points are classified as extreme high value, extreme low value, fluctuation anomaly, and null value; and adding the visualization results of the number of abnormal data points of each category in the historical data of the data source to the data analysis chart of the data source.
[0028] In one embodiment, fine-grained classification of the outlier data points using SVM includes: obtaining the feature vector of the outlier data points, where the first... i The first historical data from the data source s Feature vectors of anomalous data points X i,s satisfy: For the first i The first historical data from the data source s One abnormal data point, x i,s,1 This represents the value of the data point. x i,s,2 This indicates that the data point is the first one of the preset size. i Standard values a i The absolute value of the difference between them x i,s,3 This indicates whether the data point is null. x i,s,4 This indicates that the data point is related to the first... i The first historical data from the data source s -1 difference between outlier data points, where the data point is null. x i,s,3 The value of is 1; the first i The first historical data from the data source s Feature vectors of anomalous data points X i,s Input is given to a preset SVM model consisting of extreme high values, extreme low values, volatility anomaly SVM models, and null values; the confidence level of the output of the extreme high value SVM model, extreme low value SVM model, volatility anomaly SVM model, and null value SVM model is used to determine the first... i The first historical data from the data source s The categories of anomalous data points.
[0029] Among them, the first i The first historical data from the data source s The categories of outlier data points include: the first confidence score of obtaining extremely high SVM model output values. α1. Second confidence level of SVM model output at extremely low values α 2. Third confidence level of the output of the fluctuation anomaly SVM model α 3. The fourth confidence level of the output of the null SVM model α 4.
[0030] Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The abnormal data points are classified as extremely high values.
[0031] Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The abnormal data points are classified as extremely low values.
[0032] Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The abnormal data points are classified as fluctuation anomalies.
[0033] Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The category of the abnormal data points is null.
[0034] It should be noted that the extreme high-value SVM model is used to detect data points that are significantly higher than the normal range among outlier data points. The decision function of the extreme high-value SVM model is: .
[0035] in, The first weight vector is preset. The first bias is set to a preset size. Based on the decision model, the extreme high-value SVM model in this invention calculates the first confidence level. α The formula for 1 is: .
[0036] It should be noted that the first confidence level α 1 is to make the first i The first historical data from the data source s Feature vectors of anomalous data points X i,s The output result obtained after inputting into the SVM model with extremely high values.
[0037] The extreme low value SVM model is used to detect and identify data points that are significantly below the normal range in outlier data points. The decision function of the extreme low value SVM model is: .
[0038] in, The second weight vector is preset. The second bias is set to a preset size. Based on the decision model, the extreme low-value SVM model in this invention calculates the second confidence level. α The formula for 2 is: .
[0039] It should be noted that the second confidence level α 2 is to make the first i The first historical data from the data source s Feature vectors of anomalous data points X i,s The output result obtained after inputting into an SVM model with extremely low values.
[0040] The fluctuation anomaly SVM model is used to detect data points with drastic changes in value among outlier data points. The decision function of the fluctuation anomaly SVM model is: .
[0041] in, The third weight vector is preset. The third bias is set to a preset size. Based on the decision model, the fluctuation anomaly SVM model in this invention calculates the third confidence level. α The formula for 3 is: .
[0042] It should be noted that the third confidence level α 3 is to make the first i The first historical data from the data sources Feature vectors of anomalous data points X i,s The output results obtained after inputting into the fluctuation anomaly SVM model.
[0043] Calculating the fourth confidence level in a null SVM model α The formula for 4 is: .
[0044] Step S3: Calculate the evaluation metrics for each data source.
[0045] Among them, the absolute value of the difference between the normal data points in the historical data of the data source and the preset standard value is proportional to the evaluation index of the data source.
[0046] In one embodiment, the calculation of the first i The formula for the evaluation metrics of each data source is: .
[0047] in, σ i For the first i The standard deviation of all normal data points in the historical data of each data source, and, n i For the first i The number of all normal data points in the historical data of each data source. A j,i For the first i The historical data from the data source is the first j The value of a normal data point K The adjustment coefficient is the preset size. j It is a positive integer. Calculate the... i Standard deviation of all normal data points in historical data from one data source σ i The formula is: .
[0048] Step S4: Visualize the evaluation metrics of the data source and normal data points in historical data to obtain data analysis charts of the data source.
[0049] In one embodiment, the normal data points are linearly interpolated before the normal data points of the historical data from the data source are visualized.
[0050] It should be noted that after linear interpolation of the normal data points, the time interval between two adjacent data points is a constant.
[0051] Step S5: In response to receiving the user's input text, obtain the semantic text similarity between the user's input text and the preset standard terminology names of each data source.
[0052] In one embodiment, the user's input text is obtained and the... i The semantic text similarity between the standard term names preset by the data source includes: preprocessing the input text and converting it into a first TF-IDF vector; and then converting the first TF-IDF vector into a second TF-IDF vector. i The data source's preset standard terminology names are converted into a second TF-IDF vector; the cosine similarity between the first TF-IDF vector and the second TF-IDF vector is calculated, and the cosine similarity is determined to be the ratio between the user's input text and the first TF-IDF vector. i Semantic text similarity between standard terminology names preset by the data source.
[0053] It's important to note that TF-IDF (Term Frequency-Inverse Document Frequency) is a weighted technique commonly used in information retrieval and text mining. TF-IDF is used to evaluate the importance of a word to one document in a collection of documents. The basic idea is that if a word appears frequently in one document but infrequently in others, then that word is considered to have good class distinction.
[0054] The preprocessing of the input text includes: removing stop words from the input text and then performing word segmentation on the input text.
[0055] Step S6: Output data analysis charts for data sources whose relevance to the input text is greater than a preset threshold.
[0056] In one embodiment, the data analysis graph is transmitted as image data to a relevant display unit, enabling the display unit to display the data analysis graph.
[0057] Figure 2 This is a schematic diagram illustrating the structural block of an intelligent interactive one-stop digital scene display system based on multi-source data according to an embodiment of the present invention.
[0058] This invention also provides an intelligent interactive one-stop digital scene display system based on multi-source data. For example... Figure 2 As shown, the system includes a processor and a memory, the memory storing computer program instructions. When the computer program instructions are executed by the processor, the intelligent interactive one-stop digital scene display method based on multi-source data according to the first aspect of the present invention is implemented.
[0059] The system also includes other components well known to those skilled in the art, such as communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.
[0060] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions that can be stored or otherwise maintained by such a computer-readable medium.
[0061] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise explicitly specified.
[0062] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.
Claims
1. A one-stop digital scene display method based on multi-source data with intelligent interaction, characterized in that, include: Obtain historical data from multiple data sources, where the historical data from one data source includes data points from multiple time points; Anomaly detection algorithms are used to identify anomalous data points in the historical data of each data source, and to identify normal data points in the historical data of the data source other than anomalous data points. Calculate the evaluation index for each data source, where the absolute value of the difference between the normal data points in the historical data of the data source and the preset standard value is proportional to the evaluation index of the data source. Visualize the evaluation metrics of the data source and normal data points in historical data to obtain data analysis charts of the data source; In response to receiving user input text, obtain the semantic text similarity between the user input text and the preset standard terminology names of each data source; Output data analysis charts from data sources whose relevance to the input text is greater than a preset threshold.
2. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 1, characterized in that, The anomaly detection algorithm is the isolated forest anomaly detection algorithm.
3. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 1, characterized in that, Also includes: The abnormal data points are classified in a fine-grained manner using SVM to obtain the categories of abnormal data points, which are classified as extreme high value, extreme low value, fluctuation anomaly, and null value. The results of visualizing the number of outlier data points of each category in the historical data of the data source are added to the data analysis chart of the data source.
4. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 3, wherein fine-grained classification of the abnormal data points using SVM includes: Obtain the feature vector of the outlier data point, where the first... i The first historical data from the data source s Feature vectors of anomalous data points X i,s satisfy: For the first i The first historical data from the data source s One abnormal data point, x i,s,1 This represents the value of the data point. x i,s,2 This indicates that the data point is the first one of the preset size. i Standard values a i The absolute value of the difference between them x i,s,3 This indicates whether the data point is null. x i,s,4 This indicates that the data point is related to the first... i The first historical data from the data source s -1 difference between outlier data points, where the data point is null. x i,s,3 The value is 1. i , s All are positive integers; The first i The first historical data from the data source s Feature vectors of anomalous data points X i,s Input into the preset SVM model for extreme high values, extreme low values, fluctuation anomaly, and null values; The confidence level of the outputs of the extreme high value SVM model, the extreme low value SVM model, the fluctuation anomaly SVM model, and the null value SVM model is used to determine the [number of models]. i The first historical data from the data source s The categories of anomalous data points.
5. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 4, determines the first... i The first historical data from the data source s The categories of the outlier data points include: First confidence score for obtaining the output of an SVM model with extremely high values α 1. Second confidence level of SVM model output at extremely low values α 2. Third confidence level of the output of the fluctuation anomaly SVM model α 3. Fourth confidence level of the output of the null SVM model α 4; Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The category of the abnormal data points is extremely high value; Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The category of the abnormal data points is extremely low value; Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The category of the abnormal data points is fluctuation anomaly; Response at first confidence level α 1. Second confidence level α 2. Third confidence level α 3 and fourth confidence levels α 4. Satisfies the formula: Determine the first i The first historical data from the data source s The category of the abnormal data points is null.
6. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 4, characterized in that, Calculate the first i The formula for the evaluation metrics of each data source is: ； in, σ i For the first i The standard deviation of all normal data points in the historical data of each data source, and, n i For the first i The number of all normal data points in the historical data of each data source. A j,i For the first i The historical data from the data source is the first j The value of a normal data point K The adjustment coefficient is the preset size. j It is a positive integer.
7. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 1, characterized in that, Before visualizing the normal data points of the historical data from the data source, the normal data points are linearly interpolated.
8. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 1, characterized in that, Obtain the user's input text and the first i The semantic text similarity between the standard term names preset by the data source includes: After preprocessing the input text, the input text is converted into a first TF-IDF vector; The first i The default standard terminology names of the data source are converted into a second TF-IDF vector; Calculate the cosine similarity between the first TF-IDF vector and the second TF-IDF vector, and determine the cosine similarity as the ratio between the user's input text and the second TF-IDF vector. i Semantic text similarity between standard terminology names preset by the data source.
9. The intelligent interactive one-stop digital scene display method based on multi-source data according to claim 8, characterized in that, Preprocessing the input text includes: removing stop words from the input text and then performing word segmentation on the input text.
10. A one-stop intelligent interactive digital scene display method based on multi-source data, comprising a processor and a memory, wherein the memory stores a computer program, characterized in that, The processor executes the computer program to implement the intelligent interactive one-stop digital scene display method based on multi-source data as described in any one of claims 1-9.