Prediction method and device based on vehicle type, equipment, medium and program product

Abstract

The embodiment of the invention provides a prediction method and device based on a vehicle type, equipment, a medium and a program product, and relates to the technical field of intelligent driving. The method comprises the following steps: receiving a request input by a user; the request is used for requesting a prediction result, the request comprises target vehicle type information, target information related to the target vehicle type information is inquired in a database, the prediction result is obtained based on the target information, and the prediction result is displayed by adopting a first interaction chart, the first interaction icon table at least comprises a competitive product configuration table and a parameter comparison and association graph; according to the method, unified integration, accurate query and visual efficient output of cross-data-source vehicle model information are realized, and the accuracy of a vehicle data analysis prediction result is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to intelligent driving technology, and in particular to a prediction method and device based on vehicle type, equipment, medium and program product. BACKGROUND

[0002] Under the background of the current automobile industry accelerating the transformation towards electrification and intelligence, the accurate grasp of the competition pattern by vehicle enterprises increasingly relies on multi-dimensional and high-timeliness data analysis. The traditional competition analysis method relying on manual experience or static report has been difficult to cope with the rapidly changing market environment.

[0003] The existing vehicle data analysis platform mainly relies on a database and focuses on surface indicators such as historical sales, price range and market share. Some systems introduce a simple competitor benchmarking function.

[0004] However, the existing vehicle data analysis method is still mainly rule-driven, and the system lacks dynamic prediction and high-concurrency processing capability based on machine learning, resulting in a lag in response to market changes, which ultimately causes the problem of low accuracy of the existing vehicle data analysis prediction results. SUMMARY

[0005] The embodiments of the present application provide a prediction method and device based on vehicle type, equipment, medium and program product, which are used to solve the problem of low accuracy of the existing vehicle data analysis prediction results.

[0006] In a first aspect, the embodiments of the present application provide a prediction method based on vehicle type, which comprises:

[0007] receiving a request input by a user; the request is used to request a prediction result, and the request includes target vehicle type information;

[0008] querying target information related to the target vehicle type information in a database; the database is obtained based on multi-source heterogeneous data and a unified vehicle type standard coding system, the multi-source heterogeneous data refers to data from different data sources and inconsistent in format and structure, and the unified vehicle type standard coding system refers to a coding rule for mapping source information in different data sources to a unique identifier, the source information at least including vehicle type name, configuration parameter, supply chain information, vehicle field related rule information, vehicle enterprise value related information and market information;

[0009] obtaining a prediction result based on the target information;

[0010] displaying the prediction result by using a first interactive chart, and the first interactive chart at least includes a competitor configuration table, a parameter comparison and an association graph.

[0011] In a possible implementation, the obtaining of the prediction result based on the target information comprises:

[0012] Extract feature vectors from the target information; the feature vectors include at least the following five elements: supplier bargaining power, buyer bargaining power, threat of substitutes, threat of potential entrants, and intensity of competition from competitors;

[0013] The feature vector is input into the prediction model to obtain the prediction result. The prediction model includes an input layer, a backbone network, and an output layer. The input layer is used to receive the feature vector input, the backbone network is used as a multilayer perceptron, and the output layer is used to output the prediction result.

[0014] In one possible implementation, the prediction model supports dynamically adjusting the weight coefficients of each element based on a time series analysis model.

[0015] In one possible implementation, the prediction model also utilizes a graph neural network to simulate supply chain disturbances, in which a heterogeneous graph of suppliers, automakers, and component quality inspection is constructed.

[0016] In one possible implementation, the method further includes:

[0017] Receive user actions in the first interactive chart, including: delete, update and / or add operations;

[0018] The prediction result is updated based on the modification items corresponding to the operation in the first interactive chart;

[0019] The updated forecast results are displayed using a second interactive chart;

[0020] The method further includes updating the database based on the modifications.

[0021] In one possible implementation, updating the prediction result based on the modification item corresponding to the operation in the first interaction graph includes:

[0022] Delete information related to the modification item from the target information, or adjust the weight of information related to the modification item in the target information to obtain updated target information;

[0023] The prediction result is updated based on the updated target information.

[0024] In one possible implementation, when the database is obtained, a semantic mapping algorithm is specifically used to process the multi-source heterogeneous data to generate the unified vehicle model standard coding system; the semantic mapping algorithm is used to perform semantic analysis on entities in different data sources and establish mapping relationships through natural language processing algorithms.

[0025] Secondly, embodiments of this application provide a vehicle model-based prediction device, the device comprising:

[0026] A receiving module is used to receive a request input by a user; the request is for requesting a prediction result, and the request includes target vehicle information;

[0027] The query module is used to query target information related to the target vehicle model information in the database. The database is obtained based on multi-source heterogeneous data and a unified vehicle model standard coding system. The multi-source heterogeneous data refers to data from different data sources with inconsistent formats and structures. The unified vehicle model standard coding system refers to the coding rules that map source information from different data sources to unique identifiers. The source information includes at least: vehicle model name, configuration parameters, supply chain information, vehicle-related rule information, vehicle company value-related information, and market information.

[0028] The determination module is used to obtain the prediction result based on the target information;

[0029] The display module is used to display the prediction results using a first interactive chart, wherein the first interactive chart includes at least: a competitor configuration table, a parameter comparison, and a correlation graph.

[0030] In one possible implementation, the vehicle-type-based prediction device further includes: an extraction module;

[0031] The extraction module is used to extract feature vectors from the target information; the feature vectors include at least the following five elements: supplier bargaining power, buyer bargaining power, threat of substitutes, threat of potential entrants, and intensity of competition from competitors;

[0032] The determining module is used to input the feature vector into the prediction model to obtain the prediction result. The prediction model includes an input layer, a backbone network, and an output layer. The input layer is used to receive the feature vector input, the backbone network is used as a multilayer perceptron, and the output layer is used to output the prediction result.

[0033] In one possible implementation, the determining module is used to enable the prediction model to dynamically adjust the weight coefficients of each element based on a time series analysis model.

[0034] In one possible implementation, the vehicle-type-based prediction device further includes: a simulation module;

[0035] The simulation module is used to simulate supply chain disturbances in the prediction model using a graph neural network, in which a heterogeneous graph of suppliers, car manufacturers and parts quality inspection is constructed.

[0036] In one possible implementation, the vehicle-type-based prediction device further includes: an update module;

[0037] The receiving module is used to receive user operations in the first interactive chart, including: deletion operation, update operation and / or add operation;

[0038] The update module is used to update the prediction result based on the modification item corresponding to the operation in the first interactive chart;

[0039] The update module is also used to display the updated prediction results using a second interactive chart;

[0040] The update module is also used to update the database based on the modifications.

[0041] In one possible implementation, the vehicle-type-based prediction device further includes: a processing module;

[0042] The processing module is further configured to delete information related to the modification item in the target information, or to adjust the weight of information related to the modification item in the target information to obtain updated target information;

[0043] The update module is also used to update the prediction result based on the updated target information.

[0044] In one possible implementation, the vehicle-type-based prediction device further includes: a generation module;

[0045] The generation module is used to process the multi-source heterogeneous data using a semantic mapping algorithm when the database is obtained, and to generate the unified vehicle standard coding system; the semantic mapping algorithm is used to perform semantic analysis on entities in different data sources and establish mapping relationships through natural language processing algorithms.

[0046] Thirdly, embodiments of this application provide a vehicle, including: a memory and a processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0047] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0048] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0049] The vehicle model-based prediction method, device, equipment, medium, and program products provided in this application, on the one hand, construct a database based on multi-source heterogeneous data and a unified vehicle model standard coding system. This maps source information such as vehicle names, configuration parameters, supply chain information, vehicle domain rule information, vehicle enterprise value information, and market information from different data sources with inconsistent formats and structures into unique identifiers. This breaks down information barriers between different data sources, achieving unified integration and standardized management of vehicle model information across data sources. It avoids information conflicts and query biases caused by differences in data formats and lays a solid data foundation for accurate queries, effectively improving the efficiency and accuracy of target information queries. On the other hand, after acquiring target information and generating prediction results, a first interactive chart containing competitor configuration tables, parameter comparisons, and correlation graphs is used for visualization output. This presents the complex vehicle data analysis and prediction results in an intuitive and easy-to-understand form, which not only facilitates users to quickly understand the prediction conclusions and clearly compare the core differences and correlations between different vehicle models, but also significantly improves the efficiency and practicality of data output, ultimately ensuring the accuracy of vehicle data analysis and prediction results. Attached Figure Description

[0050] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0051] Figure 1 A flowchart illustrating the vehicle model-based prediction process provided in this application. Figure 1 ;

[0052] Figure 2 A flowchart illustrating the vehicle model-based prediction process provided in this application. Figure 2 ;

[0053] Figure 3 A schematic diagram of the vehicle model-based prediction device provided in this application;

[0054] Figure 4 A schematic diagram of the vehicle model-based prediction device provided in this application.

[0055] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0056] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0057] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0058] Furthermore, the technical solution involved in this application, which involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.) and the use of artificial intelligence technology for automated decision-making, and makes decisions that have a significant impact on personal rights based on the results of automated decision-making, provides users with corresponding operation entry points for users to choose to agree to or reject the results of automated decision-making; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0059] In this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0060] In the embodiments of this application, the use of terms such as "first" and "second" is to distinguish between identical or similar items that have essentially the same function and purpose. For example, "first vehicle" and "second vehicle" are merely used to distinguish different vehicles and do not limit their order of execution. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that "first" and "second" do not necessarily imply that they are different.

[0061] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following associated objects have an "or" relationship.

[0062] As the automotive industry undergoes a profound transformation towards electrification and intelligentization, the pace and dimensions of market competition continue to escalate. Automakers increasingly rely on multi-dimensional, timely data analysis to assess and strategize their competitive landscape. Traditional competitive analysis methods, which depend on manual experience summaries and static report reviews, are no longer adequate for the rapidly changing market environment due to their slow response time and narrow coverage.

[0063] Most mainstream vehicle data analysis platforms currently operate with databases as their core, focusing their analysis on superficial quantitative indicators such as historical sales, product price ranges, and market share.

[0064] Current vehicle data analysis methods are still predominantly rule-driven, lacking dynamic prediction modules based on machine learning algorithms in their system architecture, and also have shortcomings in high-concurrency data processing capabilities. This directly leads to a lag in the platform's response to changes in market trends, making it difficult to capture potential competitive opportunities and risks in advance, and resulting in low accuracy of the final prediction results, failing to provide strong support for automakers' decision-making.

[0065] To address the aforementioned issues, this application provides a vehicle model-based prediction method. This method breaks down the format and structural barriers of multi-source heterogeneous data by establishing a unified vehicle model standard coding system. It maps source information such as vehicle names, configuration parameters, supply chain information, vehicle industry rules, vehicle manufacturer value, and market information scattered across different data sources into unique identifiers, constructing a standardized vehicle model database. Based on this, a prediction request containing target vehicle model information is first received. Then, relevant target information is accurately retrieved from the database. Analysis and calculations are performed on this integrated and complete data to generate prediction results. Finally, the prediction results are visualized through a first interactive chart integrating competitor configuration tables, parameter comparisons, and correlation graphs. Ultimately, this achieves unified integration, accurate querying, and efficient visualization of vehicle model information across data sources, thereby improving the accuracy of vehicle data analysis and prediction results.

[0066] This application can be applied to automobiles and mobile terminals, and is suitable for solving the problem that existing systems cannot support users in querying their own historical behavior data, nor can they automatically execute corresponding operations based on users' past behavior, thus failing to meet users' scenario-based service needs.

[0067] The implementing entity of this application can be the prediction system of the aforementioned vehicle, or it can be the prediction system of a mobile terminal. This application does not impose any special restrictions in this regard.

[0068] The technical solutions of this application will be described in detail below with reference to specific embodiments. The specific embodiments described below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0069] Figure 1 A flowchart illustrating the vehicle model-based prediction process provided in this application. Figure 1 .like Figure 1 As shown, the method includes:

[0070] S101: Receive a request input by the user; the request is used to request the prediction result, and the request includes the target vehicle model information.

[0071] Specifically, the system deploys a user interaction interface that supports access from multiple channels, including web pages, mobile apps, and in-vehicle terminals. Users can enter or select information such as the name, model code, or VIN code fragment of the target vehicle through any of these channels. The interface performs format validation and validity verification on the input content, filters out invalid data such as null values ​​and garbled characters, and encapsulates the compliant request instructions into standardized data messages, which are then transmitted to the system's query module.

[0072] For example, a market analyst from a car company can input the target model information "XX brand 2025 pure electric SUV A-type vehicle" through the web interface of the intelligent vehicle prediction system and submit a request to predict the market sales of this model in the next 6 months. After the system interface verifies that the model information format is correct, it generates a request message containing the model identifier and the sales forecast request and sends it to the query module.

[0073] S102: Query target information related to the target vehicle model in the database; the database is based on multi-source heterogeneous data and a unified vehicle model standard coding system. Multi-source heterogeneous data refers to data from different data sources with inconsistent formats and structures. The unified vehicle model standard coding system refers to the coding rules that map source information from different data sources to unique identifiers. The source information includes at least: vehicle model name, configuration parameters, supply chain information, vehicle-related rule information, vehicle company value-related information, and market information.

[0074] Among them, the database is a structured storage carrier that integrates raw data from multiple sources and in multiple formats, and constructs the data after normalization processing through unified coding rules.

[0075] Multi-source heterogeneous data encompasses non-same-source data from various sources, including automakers' official websites, third-party automotive platforms, supply chain management systems, and traffic regulations databases.

[0076] A unified vehicle model standard coding system is the core rule for achieving cross-source data association and accurate matching.

[0077] Specifically, the query module first converts the target vehicle information in the request message into a unique identifier corresponding to the unified vehicle standard coding system. Then, it uses an index retrieval algorithm to perform precise matching in the database, retrieving all source information associated with the identifier, including configuration data such as the vehicle's power parameters and driving range, supply chain data such as battery suppliers and parts procurement cycles, automotive industry rule data such as new energy vehicle subsidy policies, value data such as automaker R&D investment and brand premium, and market data such as the vehicle's historical sales and user reputation. Finally, the retrieved multi-dimensional data is integrated into a target information set and passed to the predictive analysis module.

[0078] For example, the system query module converts "XX brand 2025 pure electric SUV A type vehicle" into a unique code "QC-2025-001" under the unified coding system. It then retrieves the associated data corresponding to this code from the database, including configuration parameters such as motor power and battery capacity, supply chain information such as the delivery cycle of battery supplier Y company, rule information such as the exemption policy for new energy vehicles in the city, valuable information such as the R&D investment ratio of the car company in the past three years, and market information such as sales data and user complaint hotspots in the first month of the model's launch. This information is then integrated to form a target information set for the model.

[0079] S103: Obtain the prediction result based on the target information.

[0080] Specifically, the predictive analysis module loads an algorithm model that matches the user's request type. If the user's request is for sales forecasting, a time series forecasting model is loaded; if it is for market share forecasting, a regression analysis model is loaded. Configuration parameters, supply chain stability, policy support, historical market data, etc., from the target information set are used as input features for the model. Key influencing factors are extracted through feature engineering, and the model is used for training and computation. At the same time, it is corrected by combining real-time market dynamic data. Finally, a prediction result containing the predicted value, confidence interval, and influencing factor analysis is generated and transmitted to the results display module.

[0081] For example, in response to a sales forecast request from a market analyst, the system loads a time series forecasting model and takes target information such as the battery range advantage of the 2025 XX brand pure electric SUV A-type vehicle, the stable supply cycle of Company Y, the subsidy policy of the city, and the sales base in the first month of its launch as input features. Through model calculation, the system obtains a sales forecast of 12,000-15,000 vehicles per month for the next 6 months, with a confidence interval of 95%. The system also concludes that a stable battery supply chain is the core positive factor and generates a forecast report.

[0082] S104: The prediction results are displayed using a first interactive chart. The first interactive chart table includes at least: a competitor configuration table, parameter comparison, and correlation graph.

[0083] The first interactive chart is a visual interface that integrates multiple data display components.

[0084] The competitor configuration table is used to present the configuration differences between the target model and similar competing products.

[0085] Parameter comparison is used to visually demonstrate the advantages and disadvantages of core performance indicators.

[0086] Association maps are used to present the relationship between prediction results and various influencing factors.

[0087] Specifically, the results display module, based on the type and data dimensions of the prediction results, calls a pre-set visualization component library to generate interactive charts including a competitor configuration table, parameter comparison, and correlation graph. The competitor configuration table uses a tabular format to compare the core configuration parameters of the target model with 3-5 similar competitors. The parameter comparison uses a bar chart or radar chart to show the numerical differences of key performance indicators. The correlation graph uses a knowledge graph format to show the correlation strength between the sales forecast value and factors such as supply chain stability, policy support, and configuration advantages. At the same time, interactive functions are configured for the charts, allowing users to click to view details, filter comparison dimensions, and export data reports, which are displayed in the user interaction interface.

[0088] For example, the system results display module generates the first interactive chart for the sales forecast results of the 2025 XX brand pure electric SUV A model. The competitor configuration table shows the configuration parameters of this model compared with competitors B and C, such as battery capacity, motor power, and intelligent driving level. The parameter comparison uses a radar chart to intuitively present the advantages of this model in terms of range and intelligent configuration. The correlation graph displays the sales forecast value in the form of nodes, and the thickness of the lines reflects the correlation strength with factors such as Y company's supply cycle and new energy subsidy policies. Market analysts can click on the "supply chain" node in the graph to view the specific impact weight of this factor on the sales forecast.

[0089] The vehicle model-based prediction method provided in this application receives a user-input request. The request is used to request prediction results and includes target vehicle model information. The method queries a database for target information related to the target vehicle model. The database is obtained based on multi-source heterogeneous data and a unified vehicle model standard coding system. Multi-source heterogeneous data refers to data from different data sources with inconsistent formats and structures. The unified vehicle model standard coding system refers to coding rules that map source information from different data sources to unique identifiers. The source information includes at least: vehicle model name, configuration parameters, supply chain information, vehicle-related rule information, vehicle company value-related information, and market information. Based on the target information, a prediction result is obtained and displayed using a first interactive chart. The first interactive chart includes at least: a competitor configuration table, parameter comparison, and a correlation graph. This method achieves unified integration, accurate querying, and efficient visualization output of vehicle model information across data sources, improving the accuracy of vehicle data analysis and prediction results.

[0090] Figure 2 A flowchart illustrating the vehicle model-based prediction process provided in this application. Figure 2 .like Figure 1 As shown. This embodiment is... Figure 3 Based on the embodiments, the implementation process of the vehicle model-based prediction method is described in detail. The vehicle model-based prediction method provided in this embodiment includes:

[0091] S201: Receive a request from the user; the request is used to request the prediction result, and the request includes the target vehicle model information.

[0092] Step S201 is similar to step S101, and will not be described again here.

[0093] S202: Query target information related to the target vehicle model in the database; the database is based on multi-source heterogeneous data and a unified vehicle model standard coding system. Multi-source heterogeneous data refers to data from different data sources with inconsistent formats and structures. The unified vehicle model standard coding system refers to the coding rules that map source information from different data sources to unique identifiers. The source information includes at least: vehicle model name, configuration parameters, supply chain information, vehicle-related rule information, vehicle company value-related information, and market information.

[0094] Step S202 is similar to step S102, and will not be described again here.

[0095] Optionally, when obtaining the database, a semantic mapping algorithm is specifically used to process the multi-source heterogeneous data to generate a unified vehicle model standard coding system; the semantic mapping algorithm is used to perform semantic analysis on entities in different data sources and establish mapping relationships through natural language processing algorithms.

[0096] Specifically, firstly, multi-source heterogeneous data from various channels such as automaker websites, dealer systems, and industry policy documents are collected. The data undergoes initial cleaning to filter out invalid and duplicate data. Then, a semantic mapping algorithm is invoked, and entity recognition technology from natural language processing algorithms is used to extract core entities from each data source (such as "2026 XX brand Model A" and "Model A's pure electric range of 600km"). Next, semantic similarity calculation and ontology matching techniques are used to analyze the semantic relationships between entities in different data sources (e.g., identifying "2026 XX A model" on the automaker's website and "XX brand 2026 Model A model" from the dealer system as the same model entity). Finally, based on the established semantic mapping relationship, a unique standard code is assigned to each unique model entity, integrating them to form a unified model standard coding system to support the construction of the database.

[0097] For example, when building a vehicle model prediction database, multi-source heterogeneous data was collected. Among them, the official website of the car manufacturer marked "2026 XXA model pure electric range 600km", the dealer system marked "XX brand 2026 A model range 600km", and the industry data platform marked "2026XX-A pure electric range 600km". Through semantic mapping algorithm, the entities such as "2026 XXA model" and "XX brand 2026 A model" were first extracted with the help of natural language processing algorithm. Then, it was analyzed that these entities have consistent semantics and all point to the same vehicle model. Subsequently, a unique standard code "XX-A-2026-001" was assigned to the vehicle model. At the same time, the configuration parameters, supply information and other entities of the vehicle model in each data source were mapped with the code. Finally, a unified vehicle model standard coding system containing the code and the corresponding mapping relationship was generated and integrated into the database construction.

[0098] S203: Extract feature vectors from target information; the feature vectors shall include at least the following five elements: supplier bargaining power, buyer bargaining power, threat of substitutes, threat of potential entrants, and intensity of competition from competitors.

[0099] Among them, supplier bargaining power refers to the ability of parts suppliers to influence the pricing / supply terms of automakers.

[0100] Buyer bargaining power refers to a consumer's ability to negotiate the price / configuration of a vehicle model.

[0101] The threat of substitutes refers to the impact of other modes of transportation on the market demand for that vehicle model.

[0102] The threat of potential entrants refers to the competitive possibility of new car manufacturers / new models entering the market.

[0103] Competition intensity refers to the degree of competition among brands of vehicles in the same class.

[0104] Specifically, quantitative rules are set for various types of target information data, and quantitative scores are performed on five dimensions respectively. The quantitative results are then combined in a fixed order to form a feature vector.

[0105] For example, based on the target information of the 2026 XX brand model A, the system quantifies the following: there are only 2 core battery suppliers with no alternative resources, resulting in a supplier bargaining power score of 8 out of 10; there are 15 competing models in the same segment, giving consumers a wide range of choices, resulting in a buyer bargaining power score of 7; the penetration rate of new energy motorcycles in the same city is 40%, indicating a threat from substitutes of 6 out of 10; the entry threshold for the new energy vehicle industry has decreased, with 3 emerging car companies planning to launch models in the same segment, resulting in a threat from potential entrants of 5 out of 10; and the average monthly promotions for models in the same segment are 8 times, resulting in a competitive intensity score of 9 out of 10. These findings ultimately form a feature vector. .

[0106] S204: Input the feature vector into the prediction model to obtain the prediction result. The prediction model includes an input layer, a backbone network, and an output layer. The input layer is used to receive the feature vector input, the backbone network is used as a multilayer perceptron, and the output layer is used to output the prediction result.

[0107] Specifically, firstly, the system transmits the extracted feature vectors to the input layer of the prediction model. The input layer normalizes the feature vectors to ensure that the data in each dimension are of the same magnitude, thereby improving computational accuracy. Subsequently, the normalized feature vectors are fed into the backbone network. The backbone network uses the neural connections of a multilayer perceptron and activation function operations to deeply mine and analyze the information in each dimension of the feature vectors. Finally, the computation results of the backbone network are transmitted to the output layer. The output layer performs format conversion and interpretation of the computation results, generates clear prediction results, and outputs them to the subsequent display stage.

[0108] For example, the feature vector extracted from the 2026 XX brand model A. The system first inputs the feature vector into the input layer of the prediction model. The input layer normalizes the vector data to eliminate the differences in the magnitude of the values ​​in each dimension. Then, the normalized vector is fed into the backbone network, which uses a multilayer perceptron to perform in-depth analysis on information such as the supplier's bargaining power and the intensity of competition among competitors. Finally, the output layer outputs the prediction result: "The annual market share of XX brand model A in 2026 is 8.5%".

[0109] Optionally, the prediction model supports dynamically adjusting the weight coefficients of each element based on the time series analysis model.

[0110] Among them, time series analysis models refer to analytical models that, based on the sequence data of factors changing over time, mine the inherent trends, periodicity, and correlation characteristics of the data, and then quantify the degree of influence of factors on the prediction target.

[0111] Specifically, the historical time series data of each element are first input into the time series analysis model. Through algorithms such as trend decomposition, correlation test, and fluctuation feature extraction, the correlation strength between each element and the prediction result is identified in different time periods. Then, based on the dynamic change law of the correlation strength, the weight coefficient adjustment instructions corresponding to each element are output in real time. After receiving the instructions, the prediction model updates the weight configuration, thereby improving the prediction accuracy.

[0112] For example, in sales forecasting scenarios, "promotional efforts," "competitor prices," and "seasonal factors" are used as core elements. The time series analysis model analyzes time series data from the past 12 months and finds that during the fourth quarter of each year (peak season), the weight of "promotional efforts" on sales increases from 0.3 in normal periods to 0.6, the weight of "seasonal factors" decreases from 0.2 to 0.1, while the weight of "competitor prices" remains stable at 0.3. The forecasting model then dynamically adjusts the weights of each element based on this analysis result, reducing the sales forecasting error during the peak season by 15% compared to the fixed weight model.

[0113] Optionally, the predictive model also utilizes graph neural networks to simulate supply chain disruptions, with the graph neural network constructing a heterogeneous graph of suppliers, automakers, and parts quality inspection.

[0114] Among them, graph neural networks are a type of deep learning model specifically designed for processing graph-structured data, capable of uncovering hidden features of data through the relationships between nodes.

[0115] Heterogeneous graphs refer to graph structures that contain multiple types of nodes and edges.

[0116] Specifically, suppliers, automakers, and parts inspection agencies are first set as three core nodes, with "supply," "quality inspection," and "procurement" as the links between nodes. The node attributes and edge weights are initialized based on historical business data and disturbance event records of the supply chain. Then, through the message passing mechanism of the graph neural network, the feature information of each node is transmitted and aggregated along the edges, thereby capturing the changes in the correlation strength between nodes. This simulates the impact of different types of supply chain disturbance events, such as supplier delivery delays and parts failing quality inspection, on the overall supply chain.

[0117] For example, in the scenario of automotive parts supply forecasting, when the "supply delay" attribute of a certain supplier node is triggered, the graph neural network will transmit the disturbance information to the corresponding car manufacturer node through the "supply" edge in the heterogeneous graph. At the same time, it will link the "quality inspection pass rate" feature of the parts quality inspection node to calculate the degree of impact of the disturbance on the car manufacturer's production plan. The prediction model, combined with this simulation result, can adjust the allocation of production resources in advance, shortening the supply chain's response time to disturbances by 20%.

[0118] S205: The prediction results are displayed using a first interactive chart. The first interactive chart table shall include at least: a competitor configuration table, parameter comparison, and a correlation graph.

[0119] Specifically, the system first calls the preset visualization component library and determines the core display dimensions of the charts based on the obtained prediction results and the target vehicle information acquired in the early stage. Then, the key configuration information of the target vehicle and its core competitors in the same class is filled into the competitor configuration table module. In the parameter comparison module, the differences in the core performance parameters of the target vehicle and its competitors are displayed in intuitive chart forms such as bar charts and line charts. In the association graph module, the relationship and strength of the relationship between the target vehicle and its suppliers, car manufacturers, industry policies, competitors and other elements are constructed and displayed in the form of "node-connection". At the same time, interactive functions are added to the charts (such as clicking on nodes to view detailed information, filtering the display range of competitors, etc.). After the charts are generated, they are displayed on the user terminal (web page, client, etc.).

[0120] For example, regarding the predicted result that "the annual market share of XX brand model A in 2026 will be 8.5%", the system generates the first interactive chart. The competitor configuration table clearly lists the core configuration information of XX brand model A in 2026 compared with five competing models of the same level, such as C, D, and E, including battery capacity, fast charging time, body size, and price. The parameter comparison module displays the differences in pure electric range of the six models in the form of a bar chart, intuitively showing the advantage of model A's 600km range compared with some competitors. In the association graph module, "2026 XX brand model A" is the core node, and is connected to its core battery supplier, company B, XX car manufacturer, new energy vehicle purchase tax reduction policy documents, and various competitor model nodes. The thickness of the connection corresponds to the strength of the association. Users can click on the company B node to view detailed information such as its production capacity and supply stability.

[0121] S206: Receive user actions in the first interactive chart, including: delete, update and / or add operations.

[0122] The delete operation refers to the user removing a portion of data that already exists in the chart (such as deleting the configuration information of a competitor).

[0123] The update operation refers to the user modifying the existing data content in the chart (such as correcting the parameter value of a certain car model).

[0124] Adding an action refers to users supplementing new data information into the chart (such as adding comparative data of a competitor).

[0125] Specifically, the system continuously monitors user actions on the first interactive chart through a front-end interaction listening program. When actions such as mouse clicks, selections, keyboard inputs, and form submissions are detected, the operation recognition logic is triggered. First, the type of operation is determined (delete, update, or add). Then, the corresponding chart data module (such as the competitor configuration table or parameter comparison module) and specific data items are accurately located. At the same time, the detailed content of the operation is recorded (such as the name of the deleted competitor, the original value before the update and the new value after the update, and the complete information of the added data). Finally, the operation type, associated data module, and specific operation content are associated and encapsulated to form a standardized operation instruction and transmitted to the subsequent data processing stage to ensure that the prediction results can be accurately updated based on the instruction.

[0126] For example, in the first interactive chart displaying "2026 XX brand model A's annual market share is 8.5%", the user discovered in the competitor configuration table module that the popular model F in the same segment was missing from the list of 5 competitors (C, D, E, etc.). The user then performed an "add" operation, filling in core information such as the battery capacity (580km) and price (248,000 yuan) of model F through the chart's built-in "Add Competitor" form and submitting it. Simultaneously, the user noticed that model D's pure electric range was listed as 510km, which differed from the officially published 530km. Therefore, an "update" operation was performed, selecting model D's range data and modifying it to 530km. The system captured these two actions through an interactive monitoring program, identifying them as "Add Operation - Competitor Configuration Table - Model F Information" and "Update Operation - Competitor Configuration Table - Model D Range", respectively. After recording the corresponding operation details, the standardized instructions were passed to the subsequent update prediction result stage.

[0127] S207: Update the prediction results based on the modification items corresponding to the operation in the first interactive chart.

[0128] Specifically, first, the system parses the received user operation commands and corresponding modification items to clarify the basic information of the target vehicle model associated with the modification item (such as competitor configurations, parameter data, etc.); second, it synchronously adjusts the previously acquired target information according to the modification item type (add, update, delete). If it is an add operation, it supplements the corresponding new data; if it is an update operation, it corrects the original data; if it is a delete operation, it removes the relevant invalid data to form the updated target information; then, it uses the same feature extraction rules as the initial prediction to re-extract feature vectors from the updated target information; finally, it inputs the new feature vectors into the prediction model, and the model recalculates to obtain the updated prediction result, completing the full-link update from modification item to prediction result.

[0129] For example, regarding the user's modifications to "Add F model configuration" and "Update D model range parameters" in the first interactive chart, the system first parses out that these two modifications are associated with competitor information data for the target model. Next, it adjusts the original target information for the 2026 XX brand A model, supplementing it with configuration data such as the battery capacity and price of the F model, while revising the D model's range from 510km to 530km, resulting in updated target information. Based on this updated target information, it re-extracts feature vectors, adjusting the competitor competitiveness score from 9 to 9.6 due to the addition of new competitors, ultimately obtaining new feature vectors. The vector was then input into the prediction model for recalculation, resulting in the updated prediction: "The annual market share of XX brand model A in 2026 is 7.8%".

[0130] Optionally, based on the modifications corresponding to the operation in the first interactive graph, update the prediction results, including:

[0131] Delete information related to the modified items from the target information, or adjust the weight of information related to the modified items in the target information to obtain the updated target information;

[0132] The prediction results are updated based on the updated target information.

[0133] The first interactive chart refers to the interactive interface that supports users to perform visual operations such as adding, deleting, and adjusting weights, and is used to carry the information related to the prediction target. The modification items are the specific content corresponding to the deletion or weight adjustment operations performed by the user on the target information in the chart.

[0134] Specifically, the system first captures the user's actions on the target information in the first interactive chart, and identifies whether the corresponding modification is a deletion operation or a weight adjustment operation. If it is a deletion operation, the content matching the modification is directly removed from the target information. If it is a weight adjustment operation, the weight coefficient of the corresponding information is changed according to the proportion set by the user or the system's preset rules to generate updated target information. The updated target information is then input into the prediction model, and the model recalculates and outputs the updated prediction result.

[0135] For example, in the automotive parts supply forecasting scenario, the first interactive chart displays target information such as "supplier supply capacity," "parts quality inspection pass rate," and "automaker procurement demand." When a user lowers the weight of "supplier supply capacity" in the chart (i.e., the modification item is weight adjustment), the system will first adjust the weight coefficient of this information in the target information, and then input the updated target information into the forecasting model. The model will recalculate accordingly and output a more realistic supply chain risk forecast result. If the user deletes the information "supply capacity of a temporary supplier" in the chart (i.e., the modification item is information deletion), the system will directly remove this information, and the forecasting model will recalculate based on the removed target information to avoid the unstable factors of temporary suppliers interfering with the forecast results.

[0136] S208: The updated prediction results are displayed using a second interactive chart.

[0137] Specifically, the system first calls the visualization component library, using the updated prediction results and adjusted target information as the data foundation, and constructs the chart with reference to the module layout of the first interactive chart (including the competitor configuration table, parameter comparison, and correlation graph). The system then populates each module with the synchronously updated data; for example, the competitor configuration table is supplemented with new competitor information and incorrect parameters are corrected; the parameter comparison module updates the chart data to match the adjusted parameter differences; and the correlation graph module synchronously updates the node correlation information. Simultaneously, a "comparison before and after modification" interactive function is added, allowing users to switch between viewing the first interactive chart (original data and initial prediction results) and the second interactive chart (updated data and updated prediction results) with a single click. After the chart is generated, it is displayed on the user's terminal.

[0138] For example, regarding the updated prediction that "the annual market share of XX brand model A in 2026 is 7.8%", the system generates a second interactive chart, whose module layout is consistent with the first interactive chart; the competitor configuration table adds information such as the battery capacity (580km) and price (248,000 yuan) of model F, while correcting the range of model D to 530km; the bar chart of the parameter comparison module is updated synchronously, intuitively showing the differences in range among the six models, including model F; the association graph module adds a node for model F and establishes a connection with the core node; users can quickly switch between viewing the difference between the initial predicted market share of 8.5% and the updated 7.8% through the "Compare Before and After Modification" function, as well as the corresponding data adjustments and changes.

[0139] S209: Update the database based on the modifications.

[0140] Specifically, the system first verifies the received user modifications to ensure that the data format conforms to database storage specifications and the content conforms to a unified vehicle model standard coding system, thus avoiding data conflicts or format errors. Then, based on the modification type (add, update, delete), the system generates corresponding database operation instructions: add operations correspond to INSERT statements (writing new data to the corresponding data table), update operations correspond to UPDATE statements (correcting existing erroneous data in the data table), and delete operations correspond to DELETE statements (removing invalid data from the data table). After executing the operation instructions to complete the data writing or modification, the system verifies the data again to confirm that the corresponding data in the database has been successfully updated, and records the data update log (including update time, update content, and operator-related identifiers) to ensure data traceability.

[0141] For example, regarding user modifications such as "adding configuration for model F" and "updating range parameters for model D," the system first verifies that the battery capacity (580km) and price (248,000 yuan) of model F meet the requirements, and that the corrected range (530km) of model D conforms to the unified coding system specifications. Next, it generates corresponding database operation instructions: for "adding configuration for model F," it executes an INSERT statement to write the unique code and configuration parameters of model F into the "Competitor Configuration Information Table" in the database; for "updating range parameters for model D," it executes an UPDATE statement to locate the corresponding record of model D in the "Model Core Parameter Table" and change the range field from 510km to 530km. After executing the instructions, it verifies that the data has been successfully updated and records the update log to ensure traceability. When other users subsequently query competitor information for the 2026 XX brand model A, they can obtain the latest data including the parameters of model F and the corrected parameters of model D.

[0142] The vehicle model-based prediction method provided in this application first receives a user's request to obtain prediction results. After extracting the target vehicle model information from the request, it queries relevant target information in a database constructed based on multi-source heterogeneous data and a unified vehicle model standard coding system. The multi-source heterogeneous data refers to data from different sources with varying formats and structures. The unified vehicle model standard coding system is a coding rule that maps source information such as vehicle model names, configuration parameters, supply chain information, vehicle sector rules, vehicle manufacturer value information, and market information from various data sources into unique identifiers. Subsequently, it extracts feature vectors from the target information, including five core elements: supplier bargaining power, buyer bargaining power, threat of substitutes, threat of potential entrants, and competitor intensity. These feature vectors are then input into the input layer and backbone network. The prediction model, consisting of a network and an output layer, receives feature vectors through the input layer, performs multi-layer perceptron operations through the backbone network, and outputs the prediction results through the output layer. These results are then displayed in a first interactive chart that includes a competitor configuration table, parameter comparison, and correlation graph. Subsequently, user operations such as deletion, update, and addition on the first interactive chart are received, and the prediction results are updated in real time based on the corresponding modifications. The updated results are displayed in a second interactive chart, and the database is iterated synchronously based on the modifications. This method leverages the five-force model feature vectors and the multi-layer perceptron prediction model to ensure the reliability of the prediction results. The combination of dual interactive charts supports user interaction to achieve dynamic updates of the prediction results and synchronous iteration of the database, effectively improving the accuracy, interactivity, and dynamic data adaptability of vehicle-related predictions.

[0143] Figure 3 A schematic diagram of the vehicle model-based prediction device provided in this application is shown below. Figure 4 As shown in the figure, this application provides a vehicle-type-based prediction device applied to a vehicle. The device includes:

[0144] The receiving module 301 is used to receive a request input by the user; the request is used to request the prediction result, and the request includes the target vehicle model information;

[0145] The query module 302 is used to query target information related to the target vehicle model in the database. The database is based on multi-source heterogeneous data and a unified vehicle model standard coding system. Multi-source heterogeneous data refers to data from different data sources with inconsistent formats and structures. The unified vehicle model standard coding system refers to the coding rules that map source information from different data sources to unique identifiers. The source information includes at least: vehicle model name, configuration parameters, supply chain information, vehicle-related rule information, vehicle company value-related information, and market information.

[0146] The determination module 303 is used to obtain the prediction result based on the target information;

[0147] Display module 304 is used to display the prediction results using a first interactive chart. The first interactive chart includes at least: a competitor configuration table, parameter comparison, and correlation graph.

[0148] In one possible implementation, the vehicle model-based prediction device further includes: an extraction module 305;

[0149] The extraction module 305 is used to extract feature vectors from target information; the feature vectors include at least the following five elements: supplier bargaining power, buyer bargaining power, threat of substitutes, threat of potential entrants, and intensity of competition from competitors;

[0150] The determination module 303 is used to input the feature vector into the prediction model to obtain the prediction result. The prediction model includes an input layer, a backbone network, and an output layer. The input layer is used to receive the feature vector input, the backbone network is used as a multilayer perceptron, and the output layer is used to output the prediction result.

[0151] In one possible implementation, the determination module 303 is used to predict the model to support the dynamic adjustment of the weight coefficients of each element based on the time series analysis model.

[0152] In one possible implementation, the vehicle model-based prediction device further includes: a simulation module 306;

[0153] Simulation module 306 is used to simulate supply chain disturbances in the prediction model using a graph neural network, which contains a heterogeneous graph of suppliers, car manufacturers and parts quality inspection.

[0154] In one possible implementation, the vehicle model-based prediction device further includes: an update module 307;

[0155] The receiving module 301 is used to receive user operations in the first interactive chart, including: delete operation, update operation and / or add operation;

[0156] Update module 307 is used to update the prediction results based on the modification items corresponding to the operation in the first interactive chart;

[0157] The update module 307 is also used to display the updated prediction results using a second interactive chart;

[0158] Update module 307 is also used to update the database based on the modifications.

[0159] In one possible implementation, the vehicle model-based prediction device further includes a processing module 308;

[0160] The processing module 308 is also used to delete information related to the modification item in the target information, or to adjust the weight of information related to the modification item in the target information to obtain the updated target information;

[0161] The update module 307 is also used to update the prediction results based on the updated target information.

[0162] In one possible implementation, the vehicle model-based prediction device further includes: a generation module 309;

[0163] The generation module 309 is used to process multi-source heterogeneous data using a semantic mapping algorithm when the database is obtained, and to generate a unified vehicle standard coding system. The semantic mapping algorithm is used to perform semantic analysis on entities in different data sources and establish mapping relationships through natural language processing algorithms.

[0164] The vehicle model-based prediction device provided in this application can be used to execute the technical solution of the vehicle model-based prediction method in any of the above embodiments of this application. Its implementation principle and technical effect are similar, and will not be described again here.

[0165] Figure 4 This is a schematic diagram of the vehicle model-based prediction device provided in this application. ​ As shown, the vehicle model-based prediction device 400 provided in this embodiment includes at least one processor 401 and a memory 402. Optionally, the device 400 also includes a communication component 403. The processor 401, memory 402, and communication component 403 are connected via a bus 404.

[0166] In a specific implementation, at least one processor 401 executes computer execution instructions stored in memory 402, causing at least one processor 401 to perform the above-described method.

[0167] The specific implementation process of processor 401 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0168] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0169] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0170] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0171] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0172] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0173] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0174] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0175] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0176] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0177] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0178] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0179] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0180] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A prediction method based on vehicle model, characterized in that, The method includes: Receive a request input by the user; the request is for requesting a prediction result, and the request includes target vehicle information; The database is used to query target information related to the target vehicle model. The database is obtained based on multi-source heterogeneous data and a unified vehicle model standard coding system. The multi-source heterogeneous data refers to data from different data sources with inconsistent formats and structures. The unified vehicle model standard coding system refers to coding rules that map source information from different data sources to unique identifiers. The source information includes at least: vehicle model name, configuration parameters, supply chain information, vehicle-related rule information, vehicle company value-related information, and market information. The prediction result is obtained based on the target information; The prediction results are displayed using a first interactive chart, which includes at least: a competitor configuration table, a parameter comparison, and a correlation graph.

2. The method according to claim 1, characterized in that, The process of obtaining the prediction result based on the target information includes: Extract feature vectors from the target information; the feature vectors include at least the following five elements: supplier bargaining power, buyer bargaining power, threat of substitutes, threat of potential entrants, and intensity of competition from competitors; The feature vector is input into the prediction model to obtain the prediction result. The prediction model includes an input layer, a backbone network, and an output layer. The input layer is used to receive the feature vector input, the backbone network is used as a multilayer perceptron, and the output layer is used to output the prediction result.

3. The method according to claim 2, characterized in that, The prediction model supports dynamically adjusting the weight coefficients of each element based on the time series analysis model.

4. The method according to claim 2 or 3, characterized in that, The prediction model also utilizes a graph neural network to simulate supply chain disturbances, in which a heterogeneous graph of suppliers, automakers, and component quality inspection is constructed.

5. The method according to any one of claims 1-3, characterized in that, The method further includes: Receive user actions in the first interactive chart, including: delete, update and / or add operations; The prediction result is updated based on the modification items corresponding to the operation in the first interactive chart; The updated forecast results are displayed using a second interactive chart; The method further includes updating the database based on the modifications.

6. The method according to claim 5, wherein updating the prediction result based on the modification item corresponding to the operation in the first interactive chart includes: Delete information related to the modification item from the target information, or adjust the weight of information related to the modification item in the target information to obtain updated target information; The prediction result is updated based on the updated target information.

7. The method according to any one of claims 1-3, wherein when obtaining the database, a semantic mapping algorithm is specifically used to process the multi-source heterogeneous data to generate the unified vehicle model standard coding system; the semantic mapping algorithm is used to perform semantic analysis on entities in different data sources and establish mapping relationships through natural language processing algorithms.

8. An electronic device, characterized in that, include: Memory and processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-7.

Images