Data adjustment method and device for vehicle, processor and electronic equipment

Abstract

This application discloses a method, apparatus, processor, and electronic device for adjusting vehicle data. The method includes: acquiring multi-source initial operating data of the vehicle, wherein the multi-source initial operating data represents the degree of influence of operating data from different sources on the vehicle's cost; invoking a target calibration model to calibrate the multi-source initial operating data to obtain multi-source target operating data, wherein the target calibration model is trained using multi-source initial operating data samples from historical time periods, and the multi-source initial operating data samples represent the degree of influence of operating data from different sources on the cost of the vehicle sample; performing error analysis on the multi-source initial operating data and the multi-source target operating data to obtain analysis results; and adjusting the multi-source target operating data in response to the analysis result indicating that the error does not meet an error threshold. This application solves the technical problem of low accuracy in vehicle data adjustment.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a method, apparatus, processor, and electronic device for adjusting vehicle data. Background Technology

[0002] Currently, the management of vehicle cost data often relies on single macroeconomic indicators for static cost prediction and adjustment. However, cost adjustment factors are mostly based on empirical estimates or simple averages of historical data, resulting in a lack of clear classification and quantitative correlation between various influencing factors. This makes it difficult to comprehensively reflect the combined impact of market environment, competitive landscape, and internal operational efficiency on costs, leading to low accuracy in cost data adjustments. Therefore, the technical problem of low accuracy in adjusting vehicle data remains.

[0003] There is currently no effective solution to the aforementioned technical problems. Summary of the Invention

[0004] This application provides a method, apparatus, processor, and electronic device for adjusting vehicle data, in order to at least solve the technical problem of low accuracy in adjusting vehicle data.

[0005] According to one aspect of the embodiments of this application, a method for adjusting vehicle data is provided. The method may include: acquiring multi-source initial operating data of the vehicle, wherein the multi-source initial operating data is used to represent the degree of influence of operating data from different sources on the cost of the vehicle; invoking a target calibration model to calibrate the multi-source initial operating data to obtain multi-source target operating data, wherein the target calibration model is trained using multi-source initial operating data samples from historical time periods, and the multi-source initial operating data samples are used to represent the degree of influence of operating data from different sources on the cost of the vehicle samples; performing error analysis on the multi-source initial operating data and the multi-source target operating data to obtain an analysis result, wherein the analysis result is used to indicate whether the error between the multi-source initial operating data and the multi-source target operating data meets an error threshold; and adjusting the multi-source target operating data in response to the analysis result indicating that the error does not meet the error threshold, wherein the accuracy of the adjusted multi-source target operating data is higher than that of the original multi-source target operating data.

[0006] Optionally, the multi-source target operation data includes at least one of the following: target economic data, target competition data, and target efficiency data. The target economic data represents the degree of influence of external economic dimensions on vehicle costs, the target competition data represents the degree of influence of different competition states on vehicle costs, and the target efficiency data represents the degree of influence of operational efficiency on vehicle costs. The target calibration model is invoked to calibrate the multi-source initial operation data to obtain multi-source target operation data. This includes: invoking a first target calibration model to calibrate the multi-source initial operation data to obtain target economic data. The first target calibration model is obtained by analyzing the multi-source initial operation data sample using time series regression.

[0007] Optionally, a target calibration model is invoked to calibrate the multi-source initial operating data to obtain multi-source target operating data. This includes: invoking a second target calibration model to calibrate the multi-source initial operating data to obtain target competition data. The second target calibration model includes at least one of the following: a demand price elasticity model, a competition intensity assessment model, and a value attribute assessment model. The demand price elasticity model is obtained by fitting the demand quantity of vehicles and the price information of vehicles. The competition intensity assessment model is obtained by fitting the price information and the average price information of vehicles of the same type. The value attribute assessment model is obtained by fitting the difference between the price information and the average price information.

[0008] Optionally, a target calibration model is invoked to calibrate the multi-source initial operating data to obtain multi-source target operating data. This includes: invoking a third target calibration model to calibrate the multi-source initial operating data to obtain target efficiency data. The third target calibration model includes at least one of the following: an economies of scale model, a learning curve model, a supply chain efficiency assessment model, or a time decay model. The economies of scale model is obtained by fitting a power function between vehicle output and vehicle cost. The learning curve model is obtained by fitting vehicle cost and total cumulative output. The supply chain efficiency assessment model is obtained by fitting the relationship between vehicle cost and vehicle delivery risk. The time decay model is obtained by fitting the exponential relationship between time changes over different time periods and vehicle cost.

[0009] Optionally, the first target calibration model is invoked to calibrate the multi-source initial operating data to obtain target economic data, including: invoking the first target calibration model to calibrate the multi-source initial operating data to obtain first target economic data, second target economic data, third target economic data, and fourth target economic data. The first target economic data is used to represent the inflation rate, the second target economic data is used to represent exchange rate fluctuations, the third target economic data is used to represent interest rate changes, and the fourth target economic data is used to represent the target GDP growth rate. The inflation rate is used to measure the upward pressure on vehicle costs, the exchange rate fluctuations are used to reflect the elasticity of vehicle costs, the interest rate changes are used to assess the impact on vehicle costs, and the target GDP growth rate is used to affect the scale effect corresponding to vehicle costs.

[0010] Optionally, a second target calibration model is invoked to calibrate the multi-source initial operating data to obtain target competition data. This includes: invoking a demand price elasticity model to calibrate the multi-source initial operating data to obtain first target competition data, where the first target competition data represents the price elasticity coefficient of vehicle cost; invoking a competition intensity assessment model to calibrate the multi-source initial operating data to obtain second target competition data, where the second target competition data represents the competition intensity coefficient of vehicle cost; and invoking a value assessment attribute model to calibrate the multi-source initial operating data to obtain third target competition data, where the third target competition data represents the premium coefficient of vehicle cost. The method further includes: performing technical evaluation on the multi-source initial operating data to obtain evaluation results; and based on the evaluation results, calibrating the multi-source initial operating data to obtain fourth target competition data, where the fourth target competition data represents the technical generation gap coefficient of vehicle cost.

[0011] Optionally, a third target calibration model is invoked to calibrate the multi-source initial operating data to obtain target efficiency data. This includes: invoking a scale economy model to calibrate the multi-source initial operating data to obtain first target efficiency data, where the first target efficiency data represents the scale effect coefficient of vehicle cost; invoking a learning curve model to calibrate the multi-source initial operating data to obtain second target efficiency data, where the second target efficiency data represents the learning curve coefficient of vehicle cost; invoking a supply chain efficiency assessment model to calibrate the multi-source initial operating data to obtain third target efficiency data, where the third target efficiency data represents the supply chain efficiency coefficient of vehicle cost; and invoking a time decay model to calibrate the multi-source initial operating data to obtain fourth target efficiency data, where the fourth target efficiency data represents the time decay coefficient of vehicle cost.

[0012] Optionally, the error includes a first error and a second error, where the first error represents the mean absolute error and the second error represents the root mean square error. The error threshold includes a first error threshold and a second error threshold. The method further includes: in response to the first error being less than the first error threshold and the second error being less than the second error threshold, determining the analysis result as an error satisfying the error threshold.

[0013] Optionally, in response to the analysis result that the error does not meet the error threshold, the multi-source target operation data is adjusted, including: in response to a first error being greater than or equal to a first error threshold, or a second error being greater than or equal to a second error threshold, determining that the error does not meet the error threshold, and adjusting the multi-source target operation data within the target time period.

[0014] According to another aspect of the embodiments of this application, a vehicle data adjustment device is also provided. The device may include: an acquisition unit for acquiring multi-source initial operating data of the vehicle, wherein the multi-source initial operating data represents the degree of influence of operating data from different sources on the cost of the vehicle; a calibration unit for calling a target calibration model to calibrate the multi-source initial operating data to obtain multi-source target operating data, wherein the target calibration model is trained using multi-source initial operating data samples from historical time periods, and the multi-source initial operating data samples represent the degree of influence of operating data from different sources on the cost of the vehicle sample; a verification unit for performing error analysis on the multi-source initial operating data and the multi-source target operating data to obtain an analysis result, wherein the analysis result indicates whether the error between the multi-source initial operating data and the multi-source target operating data meets an error threshold; and an adjustment unit for adjusting the multi-source target operating data in response to the analysis result indicating that the error does not meet the error threshold, wherein the accuracy of the adjusted multi-source target operating data is higher than that of the multi-source target operating data.

[0015] According to another aspect of the embodiments of this application, a processor is also provided. The processor is used to run a program, wherein the program executes the methods of the embodiments of this application during runtime.

[0016] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the method of the embodiments of this application when it runs.

[0017] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided. The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to perform the method of the embodiments of this application.

[0018] According to another aspect of the embodiments of this application, a vehicle is also provided. The vehicle includes a memory and a processor. The memory stores an executable program; the processor is used to run the program, which, when running, implements the methods described in the embodiments of this application.

[0019] In this embodiment, multi-source initial operating data, representing the impact of operating data from different sources on vehicle costs, is acquired. Then, a target calibration model is invoked to calibrate the multi-source initial operating data, resulting in multi-source target operating data. Error analysis is then performed between the multi-source initial operating data and the multi-source target operating data to determine if the error meets an error threshold. If the error does not meet the threshold, the multi-source target operating data is adjusted. This four-step closed-loop mechanism—collecting multi-source initial operating data, invoking the target calibration model, performing error analysis, and adaptive adjustment—achieves the quantification, dynamic calibration, and continuous optimization of various internal and external factors affecting vehicle costs. This solves the technical problem of low accuracy in vehicle data adjustment and improves the accuracy of vehicle data adjustment. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0021] Figure 1 This is a flowchart of a vehicle data adjustment method according to an embodiment of this application;

[0022] Figure 2 This is a schematic diagram of a three-dimensional economic coefficient system architecture according to an embodiment of this application;

[0023] Figure 3 This is a flowchart of an inflation rate coefficient calibration method according to an embodiment of this application;

[0024] Figure 4 This is a flowchart of a method for calibrating the price elasticity coefficient according to an embodiment of this application;

[0025] Figure 5 This is a flowchart of a learning curve coefficient calibration method according to an embodiment of this application;

[0026] Figure 6 This is a flowchart of a method for dynamically adjusting coefficients according to an embodiment of this application;

[0027] Figure 7 This is a flowchart of a coefficient validity verification method according to an embodiment of this application;

[0028] Figure 8This is a schematic diagram of a coefficient calibration result comparison according to an embodiment of this application;

[0029] Figure 9 This is a schematic diagram showing a comparison of cost prediction accuracy results according to an embodiment of this application;

[0030] Figure 10 This is a schematic diagram of a vehicle data adjustment device according to an embodiment of this application. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] According to an embodiment of this application, an embodiment of a vehicle data adjustment method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0034] Figure 1 This is a flowchart of a vehicle data adjustment method according to an embodiment of this application, such as... Figure 1 As shown, the method may include the following steps.

[0035] Step S102: Obtain multi-source initial operating data of the vehicle.

[0036] In the technical solution provided by step S102 of this application, multi-source initial operating data can be used to represent the degree of impact of operating data from different sources on the cost of the vehicle.

[0037] In this embodiment, the multi-source initial operating data can be a set of raw operating indicators that can characterize the driving factors of changes in the vehicle's life cycle cost, collected from multiple heterogeneous data sources such as the enterprise's internal operating system, external macroeconomic databases, market intelligence platforms, and supply chain management platforms. The multi-source initial operating data is not a single-dimensional static value, but rather dynamic, multi-dimensional, and high-granular raw information covering three core aspects: the macroeconomic environment, the market competition situation, and the enterprise's internal operating efficiency.

[0038] Optionally, the aforementioned multi-source initial operating data may include the initial α dimension—macroeconomic coefficients, the initial β dimension—market competition coefficients, and the initial γ dimension—operational efficiency coefficients. The initial α dimension's macroeconomic coefficients may include data obtained from institutions such as the National Bureau of Statistics, the People's Bank of China, and the State Administration of Foreign Exchange, including the Consumer Price Index (CPI), the RMB / USD / EUR exchange rate volatility, benchmark interest rate adjustment records, and quarterly Gross Domestic Product (GDP) growth rates. These macroeconomic coefficients reflect the systematic impact of the external economic environment on vehicle costs.

[0039] Optionally, the initial β-dimensional market competition coefficient can be derived from vehicle industry associations, third-party market research institutions, sales terminal monitoring systems, and competitor pricing monitoring platforms. The market competition coefficient can include the demand change rate for a specific vehicle model, changes in the average selling price of competitors, market concentration index, brand premium index, and technology gap score, etc., to describe the squeezing or supporting effect of market competition on a company's pricing space and cost-bearing capacity.

[0040] Optionally, the initial γ dimension of operational efficiency coefficients can mainly come from the enterprise's internal and production management systems, and can cover cumulative output, unit labor hours consumption, inventory turnover rate, supplier delivery on-time rate, supply chain collaboration response cycle, and the decreasing trend of unit manufacturing cost with batches, etc., to quantify the effect of the enterprise's own efficiency improvement on cost in terms of production scale, process accumulation, supply chain collaboration and technology maturity.

[0041] In this embodiment of the application, by systematically integrating the original data of the three dimensions of initial α dimension - macroeconomic coefficient, initial β dimension - market competition coefficient, and initial γ dimension - operational efficiency coefficient, comprehensive data support can be provided for subsequent calling of the target calibration model to calibrate the multi-source initial operating data.

[0042] Step S104: Call the target calibration model to calibrate the multi-source initial running data to obtain multi-source target running data.

[0043] In the technical solution provided in step S104 of this application, the target calibration model can be trained using multi-source initial operating data samples from historical time periods. The multi-source initial operating data samples can be used to represent the degree of influence of operating data from different sources on the cost of vehicle samples.

[0044] In this embodiment, after obtaining the multi-source initial operating data of the vehicle, the corresponding target calibration model can be called to calibrate the multi-source initial operating data to obtain multi-source target operating data.

[0045] Optionally, the target calibration model is a dedicated modeling toolset specifically trained and validated based on historical data samples accumulated over a long period. The multi-source initial operating data samples upon which the target calibration model training relies are historical sequences of vehicle cost changes and corresponding original operating indicators in dimensions such as macroeconomics, market competition, and operational efficiency, collected and labeled over several complete financial cycles (e.g., the past 5–8 years). These multi-source initial operating data samples not only contain the numerical values ​​themselves but also undergo collaborative labeling by humans and the system, clearly recording the actual causal relationship between cost changes and each input factor in each period. For example, when the CPI rises by 3.2% in a certain year, the unit cost of a certain electronic and electrical system rises by 2.7% accordingly. After a competitor lowers its price by 8% in a certain quarter, the sales volume of this model declines by 15%. After the cumulative production of a certain model reaches 200,000 vehicles, the unit manufacturing cost decreases by 11.3% compared to the first vehicle. These fully labeled "input-output" mapping relationships constitute the sample set for training the target calibration model, enabling the target calibration model to learn and internalize the real transmission mechanism of costs from different data sources.

[0046] Optionally, the initial operational data from multiple sources can be calibrated using differentiated mathematical modeling methods based on different dimensions, rather than a uniform formula. For example, in the initial α dimension (macroeconomics), time series regression analysis can be used to establish multiple linear or nonlinear functional relationships between cost changes and CPI, exchange rates, interest rates, and GDP. The least squares method can be used to fit the coefficients, transforming the original macroeconomic indicators into standardized cost transmission coefficients. In the initial β dimension (market competition), price elasticity is calculated using the elasticity formula of demand change rate and price change rate, while competition intensity is assessed using a weighted comprehensive index model, integrating multiple dimensions such as the number of competitors, price war intensity, and market concentration into a single competition index. In the initial γ dimension (operational efficiency), the scale effect is modeled using a power function regression model, the learning curve is fitted using logarithmic linear regression, and supply chain efficiency is assessed using principal component analysis or entropy weighting to construct a comprehensive scoring model. This yields multi-source target operational data, such as calibration coefficients with clear economic explanatory power, such as "α1=0.85", "β1=–1.8", and "γ2=0.12", representing the proportion of cost change caused by each unit change in each external or internal factor.

[0047] In this embodiment of the application, the limitations of subjective judgment mode that relies on human experience scoring or simple averaging are overcome by automated calibration driven by the target calibration model, thereby improving the efficiency and consistency of multi-source target operation data generation.

[0048] Step S106: Perform error analysis on the multi-source initial running data and the multi-source target running data to obtain the analysis results.

[0049] In the technical solution of step S106 of this application, the analysis result can be used to indicate whether the error between the multi-source initial running data and the multi-source target running data meets the error threshold.

[0050] In this embodiment, through a systematic, statistical, and quantifiable error assessment mechanism, it can be determined whether the multi-source target operation data output by the target calibration model truly, stably, and reliably reflects the actual cost-driven relationship carried by the multi-source initial operation data.

[0051] Optionally, the analysis method can be based on the principle of historical data backtesting, employing a reverse verification strategy. For example, known multi-source initial operating data within a historical time period (e.g., the past three years) can be selected as input, substituted into the current version of the target calibration model, and the target calibration model can be run to calculate the corresponding multi-source target operating data. Subsequently, mean absolute error (MAE) and root mean square error (RMSE) analyses are performed on the multi-source initial operating data and the multi-source target operating data to obtain the analysis results.

[0052] In this embodiment of the application, the error between the initial multi-source running data and the multi-source target running data is determined based on whether the error threshold is met, thereby providing a basis for determining whether to adjust the multi-source target running data.

[0053] Step S108: In response to the analysis result that the error does not meet the error threshold, the multi-source target running data is adjusted.

[0054] In the technical solution of step S108 of this application, the accuracy of the adjusted multi-source target operation data is higher than that of the multi-source target operation data.

[0055] In this embodiment, if the analysis result shows that the error does not meet the error threshold, for example, MAE is greater than or equal to 5% and RMSE is greater than or equal to 8%, then the calibration result is determined to be substandard and the multi-source target operation data can be adjusted.

[0056] Optionally, the adjustment method can be based on the results of error analysis, and targeted optimization can be performed on the input structure, algorithm parameters, or weight allocation of the target calibration model. For example, if the prediction error of a certain macroeconomic coefficient (e.g., α1 inflation rate) is too high, it is possible to trace back whether the CPI data used in the training phase of this coefficient missed specific categories (e.g., the price index of special raw materials for vehicle parts), or whether the lag effect was not considered, and thus introduce delayed variables or subdivide CPI subcategories for refitting. If the market competition coefficient β1 price elasticity shows a systematic overestimation, it may be because the target calibration model did not remove abnormal fluctuation data such as promotional seasons. In this case, seasonal filtering or outlier suppression mechanisms will be introduced to recalculate the elasticity coefficient. If the learning curve fitting deviation of the operating efficiency coefficient γ2 is large, it may be due to unreasonable production batch division or failure to identify process mutations. In this case, the learning curve stages will be re-divided, or process improvement nodes will be introduced as segmented breakpoints for segmented fitting.

[0057] Optionally, if multiple coefficients fail simultaneously, a dimensional weight reallocation mechanism can be initiated. This involves reassessing the contribution weight of each sub-dimension to cost using principal component analysis or entropy weighting, eliminating redundant or low-correlation variables, and enhancing the simplicity and generalization ability of the target calibration model. When necessary, expert experience can be introduced as a constraint to apply a reasonable range of correction boundaries to the target calibration model output, preventing extreme values.

[0058] In this embodiment, if the MAE is less than 5% and the RMSE is less than 8%, the multi-source target operation data is determined to be valid, and no adjustment is made to the multi-source target operation data. However, if the MAE is greater than or equal to 5% and the RMSE is greater than or equal to 8%, the multi-source target operation data is determined to be invalid. In this case, the initial α dimension coefficient can be updated quarterly, the initial β dimension coefficient can be updated semi-annually, and the initial γ dimension coefficient can be updated annually to adjust the multi-source target operation data, thus solving the problem of low accuracy in adjusting vehicle data.

[0059] In steps S102 to S108 of this application, multi-source initial operating data representing the impact of operating data from different sources on vehicle costs is acquired. Then, a target calibration model is invoked to calibrate the multi-source initial operating data, obtaining multi-source target operating data. Error analysis is then performed between the multi-source initial operating data and the multi-source target operating data to determine if the error meets an error threshold. If the error does not meet the threshold, the multi-source target operating data is adjusted. This four-step closed-loop mechanism—collecting multi-source initial operating data, invoking the target calibration model, performing error analysis, and adaptive adjustment—achieves the quantification, dynamic calibration, and continuous optimization of various internal and external factors affecting vehicle costs. This solves the technical problem of low accuracy in vehicle data adjustment and improves the accuracy of vehicle data adjustment.

[0060] The method described in this embodiment will be further described below.

[0061] As an optional implementation method, the multi-source target operating data includes at least one of the following: target economic data, target competition data, and target efficiency data. The target economic data represents the degree of influence of external economic dimensions on vehicle costs, the target competition data represents the degree of influence of different competition states on vehicle costs, and the target efficiency data represents the degree of influence of operating efficiency on vehicle costs. The multi-source target operating data is obtained by calling a target calibration model to calibrate the multi-source initial operating data. This includes calling a first target calibration model to calibrate the multi-source initial operating data to obtain target economic data. The first target calibration model is obtained by analyzing the multi-source initial operating data samples using time series regression.

[0062] In this embodiment, the target economic data can be α-dimensional – macroeconomic coefficients, which may include four coefficients: α1 inflation rate, α2 exchange rate volatility, α3 interest rate change, and α4 GDP growth rate. The target competition data can be β-dimensional – market competition coefficients, which may include four coefficients: β1 price elasticity, β2 competition intensity, β3 brand premium, and β4 technological gap. The target efficiency data can be γ-dimensional – operational efficiency coefficients, which may include four coefficients: γ1 scale effect, γ2 learning curve, γ3 supply chain efficiency, and γ4 time decay.

[0063] Optionally, target economic data reflects external macroeconomic variables, such as inflation rate, exchange rate fluctuations, interest rate changes, and GDP growth rate, and the systemic transmission strength of these variables to vehicle costs. Target competitive data describes changes in the market landscape, such as price elasticity, number of competitors, brand premium, and technological gap, and their squeezing or supporting effects on a company's pricing space and cost tolerance. Target efficiency data reflects improvements in a company's internal capabilities, such as economies of scale, learning curves, supply chain turnover, and time decay, and their endogenous improvement effects on unit manufacturing costs. These three types of data—target economic data, target competitive data, and target efficiency data—are not isolated but together constitute a complete driving network supporting cost forecasting. The generation process relies on a refined calibration mechanism that uses multiple dimensions and models.

[0064] Optionally, the first target calibration model is invoked to calibrate the multi-source initial operating data to obtain the target economic data. This is an independent modeling process for the macroeconomic coefficients in the α dimension. This first target calibration model is a time-series regression model trained on historical data samples accumulated over a long period. It establishes a quantitative mapping relationship between the changing trend of vehicle unit costs and macroeconomic variables (such as CPI, exchange rate, interest rate, and GDP) by fitting the synchronous change path of these variables over several past periods. For example, the first target calibration model analyzes questions such as "When the CPI rises by 1 percentage point, what is the average percentage increase in the cost of electronic and electrical systems?" and "When the RMB depreciates by 5% against the US dollar, is the increase in the cost of imported chips linear?" It uses methods such as multiple linear regression, vector error correction models, or autoregression with covariates to automatically identify the lag effects, interactions, and nonlinear contributions of each macroeconomic coefficient, ultimately outputting a set of standardized economic transmission coefficients, i.e., the target economic data. The target economic data no longer represents the original CPI index or exchange rate value, but is transformed into "the proportion of cost change caused by each unit of macroeconomic fluctuation". For example, α1=0.85 means that for every 1% increase in inflation, the cost of this model increases by an average of 0.85%.

[0065] In the embodiments of this application, firstly, the dynamic transmission mechanism between the changing trend of vehicle unit cost and macroeconomic variables is captured by the first target calibration model. This not only reflects the current impact but also identifies lag effects (for example, cost pressures may only appear three months after exchange rate fluctuations), thus improving the forward-looking nature of the forecast.

[0066] As an optional implementation method, a target calibration model is invoked to calibrate the multi-source initial operating data to obtain multi-source target operating data. This includes: invoking a second target calibration model to calibrate the multi-source initial operating data to obtain target competition data. The second target calibration model includes at least one of the following: a demand price elasticity model, a competition intensity assessment model, and a value attribute assessment model. The demand price elasticity model is obtained by fitting the demand quantity of vehicles and the price information of vehicles. The competition intensity assessment model is obtained by fitting the price information and the average price information of vehicles of the same type. The value attribute assessment model is obtained by fitting the difference between the price information and the average price information.

[0067] In this embodiment, the second target calibration model is invoked to calibrate the multi-source initial operating data to obtain target competition data. This model is based on the β dimension and the market competition coefficient. By introducing the demand price elasticity model, the competition intensity assessment model, and the value attribute assessment model, the originally vague and scattered market behavior data is transformed into target competition data with clear economic meaning, quantifiability, comparability, and predictability.

[0068] Optionally, the price elasticity of demand model establishes the elastic relationship between the rate of change in quantity demanded and the rate of change in price by analyzing historical sales data of the vehicle itself. For example, when the sales volume of a certain model changes significantly due to a price adjustment within a specific period, the price elasticity of demand model uses the formula β1=(ΔQ / Q) / (ΔP / P) to calculate the price elasticity coefficient, reflecting the sensitivity of consumers to price changes of that model. For instance, if the price of a smart cockpit model decreases by 5% and sales increase by 9%, the price elasticity of demand model will calculate β1≈1.8, indicating that the product has high price elasticity, and sufficient room for price reduction must be reserved for cost control; otherwise, it will lead to a decline in sales.

[0069] Optionally, ΔQ can be used to represent changes in quantity demanded; Q can be used to represent benchmark quantity demanded; ΔP can be used to represent changes in price; and P can be used to represent benchmark price.

[0070] Optionally, the competition intensity assessment model starts from the market structure level, constructing a comprehensive index by integrating multiple competition indicators. This model collects structured data such as the number of similar models within a region, market share concentration, price war frequency, and promotional intensity, assigning appropriate weights to different indicators. For example, the number of competitors accounts for 30%, market concentration for 40%, and price competition index for 30%, generating a comprehensive competition intensity coefficient β2 through weighted summation. This competition intensity assessment model effectively addresses the fundamental difference in cost pressure between a single competitor's price reduction and the overall market price reduction.

[0071] Optionally, the value attribute assessment model establishes a quantitative assessment mechanism for price premium from the perspective of brand and technological differentiation. This model uses the difference between the price of the vehicle model and the average price of competing products of the same type as a benchmark, and calculates the brand premium coefficient using the formula β3 = (Vehicle Price - Industry Average Price) / Industry Average Price. If a vehicle is priced 15% higher than the industry average due to its advanced intelligent driving system, then β3 = 0.15, indicating a certain technological premium capability. This premium can partially offset the cost pressure caused by rising raw material costs or exchange rate fluctuations. Simultaneously, this value attribute assessment model can also assess the cost-saving effect of technological advancement. For example, when a vehicle uses a self-developed battery architecture, resulting in a unit cost 10% lower than the industry average, β4 can be negative (e.g., -0.10), representing a cost advantage brought by technological leadership, rather than simply a price advantage.

[0072] In this embodiment of the application, different models such as the demand price elasticity model, the competition intensity assessment model, and the value attribute assessment model are used to provide a basis for subsequent targeted calibration processing of multi-source initial operating data to obtain different data.

[0073] As an optional implementation method, a target calibration model is invoked to calibrate the multi-source initial operating data to obtain multi-source target operating data. This includes: invoking a third target calibration model to calibrate the multi-source initial operating data to obtain target efficiency data. The third target calibration model includes at least one of the following: an economies of scale model, a learning curve model, a supply chain efficiency assessment model, or a time decay model. The economies of scale model is obtained by fitting a power function between vehicle output and vehicle cost. The learning curve model is obtained by fitting vehicle cost and total cumulative output. The supply chain efficiency assessment model is obtained by fitting the relationship between vehicle cost and vehicle delivery risk. The time decay model is obtained by fitting the exponential relationship between time changes over different time periods and vehicle cost.

[0074] In this embodiment, a third target calibration model is invoked to calibrate the multi-source initial operating data, specifically for modeling the operational efficiency coefficient in the γ dimension. Through four types of models—scale economy model, learning curve model, supply chain efficiency assessment model, and time decay model—key operational data in the manufacturing, supply chain, and production evolution process are deeply fitted, transforming previously difficult-to-quantify internal efficiency improvements into standardized target efficiency data that is economically significant, predictable, and traceable.

[0075] Optionally, the aforementioned economies of scale model establishes a power function relationship between cost and output by analyzing changes in unit cost at different output levels, mathematically expressed as C_unit = C0 × Q^(-γ1). The economies of scale model fits an exponential decay rate γ1 to historical production data, accurately reflecting the marginal cost reduction efficiency brought about by scale expansion. For example, when the annual output of a platform-based chassis system increases from 100,000 units to 500,000 units, the unit cost decreases by 18%. The economies of scale model will automatically fit γ1 = 0.15, indicating that for every 10% increase in output, the unit cost can decrease by approximately 1.5%.

[0076] Optionally, the aforementioned C_ unit can be used to represent unit cost, C0 can be used to represent benchmark cost, and Q can be used to represent output. γ1 is obtained by performing power function regression on cost data under different output levels.

[0077] Alternatively, the learning curve model described above takes an experience-accumulation perspective, capturing the continuous cost reduction resulting from process proficiency, operational optimization, and process standardization during production. Based on the formula C_n=C1×n^(-γ2), where C_n is the cumulative average cost of the nth product, C1 is the first-piece cost, and n is the cumulative output, the learning curve model can automatically identify the strength of the learning effect by analyzing the evolution trajectory of unit costs across different batches of products. For example, if the cost decreases by 12% when the cumulative output doubles, then γ2=0.12.

[0078] Optionally, C_n can be used to represent the cost of the nth product, C1 can be used to represent the cost of the first product, and n can be used to represent the cumulative output. γ2 is obtained by fitting a learning curve to the production data.

[0079] Optionally, the aforementioned supply chain efficiency assessment model breaks through the limitations of traditional cost analysis that only focuses on procurement prices, and models the operational quality of the supply chain as a cost driver factor. The supply chain efficiency assessment model does not directly use supplier quotations, but instead constructs a quantitative relationship between operational indicators such as inventory turnover rate, on-time delivery rate, order response speed, and supplier collaboration responsiveness and changes in unit cost. For example, when inventory turnover rate increases by 20%, on-time delivery rate increases by 15%, and supplier collaboration score increases by 10%, a weighted fitting yields γ3=0.08, meaning that an 8% increase in supply chain efficiency can lead to a 0.8% reduction in unit cost.

[0080] Optionally, the aforementioned time decay model is used to describe the long-term impact of technology maturity, product iteration cycle, and process stability on cost. The time decay model adopts an exponential decay form, capturing the process by which initial high costs are gradually smoothed out by process optimization, design simplification, and reduced testing and verification as the product lifecycle progresses. For example, a certain intelligent driving module may have a high unit cost in its initial launch phase, but with technology iteration, accumulation of test data, and improvement in manufacturing yield, the cost slowly decreases exponentially over 18 months. The time decay model can fit a value of γ4 = 0.05, indicating that the cost naturally decreases by 5% after each standardized time period.

[0081] In the embodiments of this application, more effective target efficiency data can be obtained by calibrating multi-source initial operating data through economies of scale models, learning curve models, supply chain efficiency assessment models, or time decay models.

[0082] As an optional implementation method, a first target calibration model is invoked to calibrate the multi-source initial operating data to obtain target economic data. This includes: invoking the first target calibration model to calibrate the multi-source initial operating data to obtain first target economic data, second target economic data, third target economic data, and fourth target economic data. The first target economic data is used to represent the inflation rate, the second target economic data is used to represent exchange rate fluctuations, the third target economic data is used to represent interest rate changes, and the fourth target economic data is used to represent the target GDP growth rate. The inflation rate is used to measure the upward pressure on vehicle costs, the exchange rate fluctuations are used to reflect the elasticity of vehicle costs, the interest rate changes are used to assess the impact on vehicle costs, and the target GDP growth rate is used to affect the scale effect corresponding to vehicle costs.

[0083] In this embodiment, the first target calibration model is invoked to calibrate the multi-source initial operating data to obtain target economic data. This involves breaking down the complex impact of the macroeconomic environment on vehicle costs into four sub-dimensional coefficients with clear economic meanings, independent modeling capabilities, and dynamic updates. Specifically, these are the first target economic data (e.g., inflation rate), the second target economic data (e.g., exchange rate fluctuations), the third target economic data (e.g., interest rate changes), and the fourth target economic data (e.g., GDP growth rate).

[0084] Optionally, the primary target economic data is obtained by fitting the company's historical unit cost data to the Consumer Price Index (CPI) or the Producer Price Index (PPI) over a long-term trend. This reflects the general upward pressure on costs such as raw materials, components, logistics, and labor due to rising overall price levels. For example, if the primary target calibration model fits α1=0.85, it means that when the CPI rises by 1%, the total cost of this vehicle model increases by an average of 0.85%. This primary target economic data is not a fixed value but dynamically adjusts according to different cost structures (e.g., a high proportion of electronic components or a high proportion of metal components), enabling companies to accurately identify which cost items are most sensitive to inflation, thereby locking in procurement windows or adjusting inventory strategies in advance.

[0085] Optionally, the second objective economic data involves modeling vehicle models or subsystems that rely on imported components (e.g., chips, high-end sensors, powertrains). This is achieved by establishing a regression relationship between import costs and changes in the RMB exchange rate against major currencies (e.g., USD, EUR). If the first objective calibration model fits and outputs α2 = 1.2, it means that when the exchange rate depreciates by 1%, the procurement cost of imported components increases by 1.2%.

[0086] Optionally, the third target economic data reflects the impact of changes in financing costs on the overall capital expenditure and working capital costs of enterprises. This can be achieved by conducting a correlation analysis between changes in the enterprise's historical financial expenses, loan size, and changes in the central bank's benchmark interest rate or corporate bond yield, resulting in α3 = 0.3. This means that for every 1 percentage point increase in interest rates, the unit cost due to the increase in financing costs increases by approximately 0.3%.

[0087] Optionally, the fourth objective, economic data, describes the transmission path of macroeconomic conditions to market demand, indirectly affecting the scale effect of costs. The first objective calibration model, by analyzing the linkage between enterprise sales growth and regional or national GDP growth rates, derives α4 = 0.5. This means that for every 1% increase in GDP, the enterprise's sales scale expands accordingly, and unit fixed costs are diluted by approximately 0.5%.

[0088] As an optional implementation, a second target calibration model is invoked to calibrate the multi-source initial operating data to obtain target competition data. This includes: invoking a demand price elasticity model to calibrate the multi-source initial operating data to obtain first target competition data, wherein the first target competition data represents the price elasticity coefficient of vehicle cost; invoking a competition intensity assessment model to calibrate the multi-source initial operating data to obtain second target competition data, wherein the second target competition data represents the competition intensity coefficient of vehicle cost; and invoking a value assessment attribute model to calibrate the multi-source initial operating data to obtain third target competition data, wherein the third target competition data represents the premium coefficient of vehicle cost. The method further includes: performing technical evaluation on the multi-source initial operating data to obtain evaluation results; and based on the evaluation results, calibrating the multi-source initial operating data to obtain fourth target competition data, wherein the fourth target competition data represents the technical generation gap coefficient of vehicle cost.

[0089] In this embodiment, the second target calibration model is invoked to calibrate the multi-source initial running data to obtain target competition data. Instead of treating market competition as a general external disturbance variable, it is deconstructed into four independent, logically complementary, and data-traceable sub-dimensions, which are independently modeled and calibrated by the second target calibration model, and finally form a complete competition coefficient system consisting of the first target competition data to the fourth target competition data.

[0090] Optionally, by using a demand price elasticity model and performing a nonlinear fit on historical sales volume and price change data for vehicles, the first target competitive data, i.e., the price elasticity coefficient, can be obtained. The demand price elasticity model is based on the elasticity principle in economics, calculating the relative change rate of demand caused by a unit price change, i.e., β1=(ΔQ / Q) / (ΔP / P). For example, if the price of a certain smart cockpit model is reduced by 10% during a promotional period, and sales increase by 18%, the output will be β1=1.8, indicating that consumers are highly sensitive to the price of this product.

[0091] Optionally, a competition intensity assessment model can be invoked to integrate structural data such as the number of competing products of the same type of vehicle within the region, market concentration, and frequency of price wars, to construct a weighted comprehensive assessment system. This system can output a second target competition data point, namely the competition intensity coefficient β2. This competition intensity assessment model does not rely on the behavior of a single competitor but rather judges the systemic pressure of competition from the perspective of market structure. For example, if there are more than 20 similar models in the market, the combined market share of the top three companies is less than 50%, and 80% of competitors have launched price promotions in the past six months, then β2 can be set at 0.82, indicating that the market is in a state of fragmented, highly competitive, and low-profit vicious competition.

[0092] Optionally, a value assessment attribute model can be invoked to standardize the difference between the price of this model and the average price of similar products in the industry, thereby obtaining the third target competitive data, namely, the premium coefficient β3. For example, if this model is priced 15% higher than the industry average due to its industry-leading voice interaction system and luxurious seat materials, then β3 = 0.15, meaning that the brand and configuration advantages can support a 15% cost premium.

[0093] Optionally, a technical evaluation mechanism can be introduced to systematically assess the key technologies used in the vehicle, such as the maturity of intelligent driving algorithms, battery energy density, electric drive system integration, and the proportion of self-developed electronic control units. Through expert scoring, patent comparison, and industry benchmarking, a technical advancement evaluation result can be formed. Based on this evaluation result, a fourth competitive data point, namely the technology gap coefficient β4, can be calculated to quantify the cost difference between this model and the industry average. For example, if this model's electric drive system cost is 12% lower than the industry average due to the use of a self-developed 800V high-voltage platform, then β4 = -0.12, indicating that technological leadership directly translates into a cost advantage.

[0094] As an optional implementation, a third target calibration model is invoked to calibrate the multi-source initial operating data to obtain target efficiency data. This includes: invoking a scale economy model to calibrate the multi-source initial operating data to obtain first target efficiency data, where the first target efficiency data represents the scale effect coefficient of vehicle cost; invoking a learning curve model to calibrate the multi-source initial operating data to obtain second target efficiency data, where the second target efficiency data represents the learning curve coefficient of vehicle cost; invoking a supply chain efficiency assessment model to calibrate the multi-source initial operating data to obtain third target efficiency data, where the third target efficiency data represents the supply chain efficiency coefficient of vehicle cost; and invoking a time decay model to calibrate the multi-source initial operating data to obtain fourth target efficiency data, where the fourth target efficiency data represents the time decay coefficient of vehicle cost.

[0095] In this embodiment, the four sets of data—first target efficiency data (e.g., scale effect coefficient), second target efficiency data (e.g., learning curve coefficient), third target efficiency data (e.g., supply chain efficiency coefficient), and fourth target efficiency data (e.g., time decay coefficient)—together constitute the γ dimension, which is the complete expression system of the operational efficiency coefficient.

[0096] Optionally, by invoking the economies of scale model and fitting a power function to the unit cost data at different output levels, the first target efficiency data, i.e., the scale effect coefficient γ1, can be obtained. The economies of scale model does not presuppose a linear decrease in cost with output; instead, it automatically identifies the non-linear relationship between increased output and decreased unit cost through historical data, mathematically expressed as C_unit = C0 × Q^(-γ1). For example, when the annual output of a certain platform-based battery pack increases from 50,000 units to 300,000 units, the unit cost decreases by 17%. This can be automatically fitted with γ1 = 0.15, indicating that for every 10% increase in output, the unit cost can be reduced by approximately 1.5%.

[0097] Optionally, a learning curve model can be invoked to obtain the second objective efficiency data, i.e., the learning curve coefficient γ2, by exponentially fitting the evolution path of cumulative production and unit cost. The formula for the learning curve model is C_n=C1×n^(-γ2), where n is the cumulative production. For example, if the cost of a new electronic control unit decreases by 12% as the cumulative production doubles in the first 100,000 units produced, γ2=0.12 can be output. This coefficient allows companies to predict the cost evolution path after the mass production of a new model in advance, and to rationally formulate pricing strategies, supplier negotiation pace, and profit targets, avoiding losses due to prematurely promising low prices or missing market opportunities due to conservative pricing.

[0098] Optionally, by employing a supply chain efficiency assessment model and correlating operational performance data such as inventory turnover rate, on-time delivery rate, order response time, and supplier collaboration score with changes in unit cost, a third target efficiency data can be derived, namely, the supply chain efficiency coefficient γ3. This supply chain efficiency assessment model treats the supply chain as a collaborative and dynamic system. For example, when a company implements a supplier managed inventory model, resulting in a 25% increase in inventory turnover rate, an 18% increase in on-time delivery rate, and a 40% decrease in line stoppages due to material shortages, a fitting result of γ3 = 0.08 can be obtained. This means that for every 1 percentage point increase in supply chain collaboration efficiency, unit cost decreases by 0.08%.

[0099] Optionally, a time decay model can be invoked to exponentially fit the cost changes of the same product at different stages of its lifecycle, deriving the fourth objective efficiency data, namely, the time decay coefficient γ4. The time decay model reflects the cost reduction that naturally occurs over time due to technological maturity, process stability, reduced testing and verification, and improved yield. For example, a certain intelligent driving domain controller has a high unit cost in its initial year, but after 18 months of mass production iterations, due to design freeze, process solidification, and increased test case reuse, the cost slowly decreases exponentially, resulting in a fitted γ4 = 0.05. This indicates that the cost naturally decays by 5% after each standardized cycle.

[0100] As an optional embodiment, the error includes a first error and a second error, the first error representing the mean absolute error and the second error representing the root mean square error. The error threshold includes a first error threshold and a second error threshold. The method further includes: in response to the first error being less than the first error threshold and the second error being less than the second error threshold, determining the analysis result as the error satisfying the error threshold.

[0101] In this embodiment, the first error is the mean absolute error (MAE), which measures the average deviation between the predicted cost and the actual historical cost, reflecting the central tendency of the overall deviation. The second error is the root mean square error (RMSE), which amplifies the impact of outliers by squaring them, more sensitively capturing extreme prediction deviations and reflecting the stability and anti-interference ability of the prediction results. The two complement each other, forming a dual verification standard for the quality of coefficient calibration from the two dimensions of average accuracy and volatility risk.

[0102] Optionally, in practice, the calibrated economic coefficients of α, β, and γ can be used to retrospectively predict historical cost data over several periods, calculating the first and second errors between the predicted and actual values. If and only if the first error is less than a preset first error threshold (e.g., 5%), it indicates that the overall prediction deviation is at an acceptable average level. Simultaneously, if the second error is also less than the second error threshold (e.g., 8%), it indicates that the prediction result has not shown significant extreme deviations or runaway fluctuations. In this case, the error of the calibration result is deemed to meet the error threshold, meaning the coefficients are valid.

[0103] In this embodiment of the application, by introducing a combined judgment mechanism of mean absolute error and root mean square error, a rigorous, scientific and executable coefficient validity verification system is constructed, which realizes the goal of intelligent cost management with more accurate prediction, faster response and more stable decision-making.

[0104] As an optional embodiment, in response to the analysis result that the error does not meet the error threshold, the multi-source target operation data is adjusted, including: in response to a first error being greater than or equal to a first error threshold, or a second error being greater than or equal to a second error threshold, determining that the error does not meet the error threshold, and adjusting the multi-source target operation data within the target time period.

[0105] In this embodiment, when the first error (MAE) obtained through backtesting of historical data is greater than or equal to the first error threshold (e.g., 5%), or the second error (RMSE) is greater than or equal to the second error threshold (e.g., 8%), it is determined that the currently used economic coefficients can no longer accurately reflect the true cost change trend, i.e., the model has failed or the environment has changed abruptly. At this time, a data adjustment procedure is actively triggered to perform targeted correction and recalibration on the multi-source target operating data.

[0106] Optionally, the multi-source target operating data can be adjusted within the target time period. For example, the α coefficient can be updated quarterly, the β coefficient can be updated semi-annually, and the γ coefficient can be updated annually.

[0107] Optionally, by following the above steps, instead of relying on manual periodic review or post-audit, a preset error threshold is used as an alarm to initiate intervention as soon as the model performance deteriorates, ensuring that the coefficients are always in an effective service state.

[0108] In this embodiment, multi-source initial operating data, representing the impact of operating data from different sources on vehicle costs, is acquired. Then, a target calibration model is invoked to calibrate the multi-source initial operating data, resulting in multi-source target operating data. Error analysis is then performed between the multi-source initial operating data and the multi-source target operating data to determine if the error meets an error threshold. If the error does not meet the threshold, the multi-source target operating data is adjusted. This four-step closed-loop mechanism—collecting multi-source initial operating data, invoking the target calibration model, performing error analysis, and adaptive adjustment—achieves the quantification, dynamic calibration, and continuous optimization of various internal and external factors affecting vehicle costs. This solves the technical problem of low accuracy in vehicle data adjustment and improves the accuracy of vehicle data adjustment.

[0109] The technical solutions of the embodiments of this application will be illustrated below with reference to preferred embodiments.

[0110] Currently, cost forecasting and dynamic adjustment are key aspects of vehicle cost management. However, the following problems exist in related technologies.

[0111] The cost adjustment factor is too simplistic. Traditional cost management methods usually only consider single macroeconomic factors such as the inflation rate, which fails to fully reflect the multidimensional factors that affect costs, resulting in insufficient accuracy in cost forecasting.

[0112] The lack of a systematic coefficient system means that existing methods have not established a systematic economic coefficient system, and there is a lack of clear classification and correlation among various influencing factors, making it difficult to carry out structured management.

[0113] The coefficient calibration methods are unscientific. Traditional methods often rely on empirical estimation or simple historical data averaging, lacking scientific calibration methods and quantitative models, resulting in insufficient accuracy and reliability of the coefficients.

[0114] The dynamic adjustment capability is weak. The existing coefficients are usually static values ​​and cannot be dynamically adjusted according to changes in the market environment and the company's operating conditions, making it difficult to adapt to the rapidly changing market environment.

[0115] The lack of a verification mechanism means that traditional methods cannot verify the validity of coefficients and therefore cannot assess the accuracy and applicability of coefficient calibration.

[0116] To address the aforementioned issues, this application proposes a multi-dimensional economic coefficient calibration method for automotive cost management. It establishes a three-dimensional economic coefficient system (α macroeconomic coefficient, β market competition coefficient, and γ operating efficiency coefficient) to comprehensively cover various factors influencing costs. A coefficient calibration model based on scientific methods such as regression analysis, elasticity coefficients, and learning curves is proposed, improving the accuracy and reliability of the coefficients. The coefficients can be dynamically adjusted based on real-time data and market changes, enhancing the adaptability of cost management. A coefficient validity verification mechanism is established, validating the accuracy of the coefficients through backtesting with historical data and practical applications. Therefore, in practical applications, this method improves the accuracy of cost prediction and cost fluctuation early warning.

[0117] The embodiments of this application will be further described below.

[0118] Figure 2 This is a schematic diagram of a three-dimensional economic coefficient system architecture according to an embodiment of this application, such as... Figure 2 As shown, the three-dimensional economic coefficient system 200 includes: α: macroeconomics 201, β: market competition 202, and γ: operational efficiency 203.

[0119] In this embodiment, the three-dimensional economic coefficient system design includes three dimensions: α (macroeconomic), β (market competition), and γ (operational efficiency), totaling 12 coefficients. The α dimension – macroeconomic coefficients – include four coefficients: α1 (inflation rate), α2 (exchange rate fluctuation), α3 (interest rate change), and α4 (GDP growth rate), reflecting the impact of the macroeconomic environment on costs. The β dimension – market competition coefficients – include four coefficients: β1 (price elasticity), β2 (competitive intensity), β3 (brand premium), and β4 (technological gap), reflecting the impact of market competition on costs. The γ dimension – operational efficiency coefficients – include four coefficients: γ1 (scale effect), γ2 (learning curve), γ3 (supply chain efficiency), and γ4 (time decay), reflecting the impact of enterprise operational efficiency on costs.

[0120] Figure 3 This is a flowchart of an inflation rate coefficient calibration method according to an embodiment of this application, such as... Figure 3 As shown, the method may include the following steps.

[0121] Step S301: Obtain CPI data and corresponding cost data.

[0122] In this embodiment, CPI data and corresponding cost data can be obtained.

[0123] Step S302: Perform data preprocessing.

[0124] In this embodiment, CPI data and corresponding cost data can be preprocessed.

[0125] Step S303: Establish a regression model.

[0126] In this embodiment, time series regression analysis can be used to establish a regression model between cost and CPI.

[0127] C(t) = C0 × (1 + α1 × ΔCPI)

[0128] Where C(t) is the cost at time t, C0 is the benchmark cost, and ΔCPI is the CPI change rate. The calibration value of α1 is obtained through regression analysis of historical data.

[0129] Step S304: Perform regression analysis and estimate parameters using the least squares method.

[0130] In this embodiment, regression analysis can be performed, and parameter estimation can be performed using the least squares method.

[0131] Step S305: Extract the α1 inflation rate.

[0132] In this embodiment, regression analysis is performed, and after parameter estimation using the least squares method, the α1 inflation rate can be extracted.

[0133] Step S306: Check if the goodness of fit is greater than 0.85.

[0134] In this embodiment, it can be determined whether the goodness-of-fit test is greater than 0.85. 0.85 can be obtained through regression analysis of CPI and cost data from 2020 to 2024.

[0135] Step S307: Determine if the α1 inflation rate is valid.

[0136] In this embodiment, if the goodness-of-fit test is greater than 0.85, the α1 inflation rate can be determined to be effective.

[0137] Step S308: Adjust the data or model.

[0138] In this embodiment, if the goodness-of-fit test is less than or equal to 0.85, the data or model can be adjusted.

[0139] Step S309: The inflation rate α1 is calibrated and output as α1=0.85.

[0140] In this embodiment, if the α1 inflation rate is valid, then the α1 inflation rate calibration is complete, and the output is α1=0.85.

[0141] Similarly, the α2 exchange rate volatility coefficient is calibrated as follows. For imported components, a model relating cost and exchange rate is established:

[0142] C_imports = C_local currency × (1 + α2 × Δexchange rate) × import share

[0143] α2 was obtained through regression analysis using historical exchange rate data and import cost data.

[0144] The α3 interest rate change coefficient and the α4 GDP growth rate coefficient can be calibrated using similar regression analysis methods.

[0145] Figure 4 This is a flowchart of a method for calibrating the price elasticity coefficient according to an embodiment of this application, such as... Figure 4 As shown, the method may include the following steps.

[0146] Step S401: Data collection.

[0147] In this embodiment, data collection can be performed.

[0148] Step S402: Calculate the rate of price change.

[0149] In this embodiment, the price change rate ΔP can be calculated.

[0150] Step S403: Calculate the rate of change in demand.

[0151] In this embodiment, the rate of change in demand ΔQ can be calculated.

[0152] Step S404: Calculate the price elasticity coefficient.

[0153] In this embodiment, the price elasticity coefficient β1 can be calculated.

[0154] Step S405: Determine the elasticity type.

[0155] In this embodiment, a demand price elasticity model can be determined.

[0156] β1=(ΔQ / Q) / (ΔP / P)

[0157] Where ΔQ represents the change in quantity demanded, Q is the baseline quantity demanded, ΔP represents the change in price, and P is the baseline price. β1 is calculated using market research data and sales data.

[0158] Step S406: Check if the data reasonableness verification is passed.

[0159] In this embodiment, it can be determined whether the data reasonableness check passes. If it passes, step S407 is executed. Otherwise, step S401 is executed.

[0160] Step S407: Determine the validity of the coefficients.

[0161] In this embodiment, if the data reasonableness test passes, the coefficient can be determined to be valid.

[0162] In step S408, the price elasticity of β1 is calibrated, and the output is β1 = -0.18.

[0163] In this embodiment, if the determination coefficient is valid, the price elasticity of β1 is calibrated and the output is β1=-0.18.

[0164] Similarly, the calibration process for the β2 competition intensity coefficient is as follows.

[0165] The following competition intensity assessment model is established.

[0166] β2 = (Number of competitors × 0.3) + (Market concentration × 0.4) + (Price competition index × 0.3)

[0167] β2 is obtained by calculating various indicators using market analysis data and then weighting and summing them.

[0168] The calibration process for the β3 brand premium coefficient is as follows.

[0169] The brand value assessment method is: β3 = (brand product price - average price of similar products) / average price of similar products, to obtain β3.

[0170] The β4 technology gap coefficient can be calibrated through technology level assessment and market comparison analysis.

[0171] Figure 5 This is a flowchart of a learning curve coefficient calibration method according to an embodiment of this application, such as... Figure 5 As shown, the method may include the following steps.

[0172] Step S501: Data collection.

[0173] In this embodiment, data collection can be performed.

[0174] Step S502: Perform logarithmic transformation.

[0175] In this embodiment, logarithmic transformation can be performed.

[0176] Step S503: Establish a learning curve model.

[0177] In this embodiment, a learning curve model can be established:

[0178] C_n = C1 × n^(-γ2)

[0179] Where C_n is the cost of the nth product, C1 is the cost of the first product, and n is the cumulative output. γ2 is obtained by fitting a learning curve to the production data.

[0180] Step S504: Calculate the price elasticity coefficient.

[0181] In this embodiment, the price elasticity coefficient can be calculated.

[0182] Step S505: Perform linear regression and fit a straight line in a logarithmic coordinate system.

[0183] In this embodiment, linear regression can be performed to fit a straight line in a logarithmic coordinate system.

[0184] Step S506: Extract the γ2 coefficient value.

[0185] In this embodiment, the γ2 coefficient value can be extracted.

[0186] Step S507: Calculate the learning rate.

[0187] In this embodiment, the learning rate can be calculated.

[0188] Step S508: Check if the goodness of fit is greater than 0.90.

[0189] In this embodiment, it can be determined whether the goodness-of-fit test is greater than 0.90. If the goodness-of-fit test is greater than 0.90, then step S509 is executed. Otherwise, step S510 is executed.

[0190] Step S509: Determine the validity of the coefficients.

[0191] In this embodiment, if the goodness-of-fit test is greater than 0.90, the coefficients can be considered valid.

[0192] Step S510: Add data points or remove outliers.

[0193] In this embodiment, if the goodness-of-fit test is less than or equal to 0.90, data points can be added or outliers can be removed.

[0194] Step S511: γ2 calibration completed, output γ2=0.12, learning rate=92%.

[0195] In this embodiment, if the coefficients are valid, the γ2 calibration is complete, and the output γ2=0.12, learning rate=92%.

[0196] Similarly, the calibration process for the γ1 scale effect coefficient is as follows.

[0197] The following economies of scale model is adopted: C_unit = C0 × Q^(-γ1), where C_unit is the unit cost and Q is the output. γ1 is obtained by performing power function regression on cost data at different output levels.

[0198] The calibration process for the γ3 supply chain efficiency coefficient is as follows.

[0199] Establish the following supply chain efficiency assessment model;

[0200] γ3 = (Inventory turnover improvement × 0.4) + (On-time delivery rate improvement × 0.3) + (Supplier collaboration × 0.3)

[0201] The γ4 time decay coefficient can be calibrated using an exponential decay model.

[0202] Figure 6 This is a flowchart of a method for dynamically adjusting coefficients according to an embodiment of this application, such as... Figure 6 As shown, the method may include the following steps.

[0203] Step S601: Set the update cycle.

[0204] In this embodiment, a dynamic adjustment mechanism for the coefficients can be established to update the coefficients periodically based on real-time data and market changes.

[0205] Extend the point by setting the coefficient update cycle. For example, the α coefficient is updated quarterly, the β coefficient is updated every six months, and the γ coefficient is updated annually.

[0206] Step S602: Has the update cycle been reached?

[0207] In this embodiment, it can be determined whether the update cycle has been reached. If the update cycle has been reached, step S604 is executed. Otherwise, step S603 is executed.

[0208] Step S603, wait.

[0209] In this embodiment, if the update cycle has not been reached, the process waits until the update cycle is reached.

[0210] Step S604: Automatically collect data.

[0211] In this embodiment, if the update cycle is reached, a data collection mechanism can be established to automatically collect macroeconomic data, market data, and enterprise operation data.

[0212] Step S605: Perform coefficient calibration.

[0213] In this embodiment, the coefficients can be recalculated, compared with the old ones, and the magnitude of change can be calculated to perform coefficient calibration.

[0214] Step S606: Verify the validity of the coefficients.

[0215] In this embodiment, the accuracy of the new coefficients can be verified by backtesting with historical data.

[0216] Optionally, historical data backtesting can use calibrated coefficients to predict historical costs and calculate prediction errors. Error analysis can calculate the mean absolute error (MAE) and root mean square error (RMSE).

[0217] Step S607: Are the coefficients valid, and is MAE < 5% and RMSE < 8%?

[0218] In this embodiment, it can be determined whether the coefficient is valid and whether MAE < 5% and RMSE < 8%. If yes, then step S609 is executed. Otherwise, step S608 is executed.

[0219] Step S608: Perform coefficient optimization and adjustment.

[0220] In this embodiment, coefficient optimization can be performed by analyzing errors, adjusting the model, and recalibrating. For example, coefficients that do not meet the validity criteria can be adjusted and optimized.

[0221] Step S609: Update the coefficient database.

[0222] In this embodiment, when MAE < 5% and RMSE < 8%, the determination coefficient is valid, and the coefficient database can be updated by saving the new value, recording the time, and retaining the history.

[0223] Step S610: Send an update notification.

[0224] In this embodiment, the system can notify departments, update the prediction model, generate reports, and send update notifications.

[0225] Figure 7 This is a flowchart of a coefficient validity verification method according to an embodiment of this application, such as... Figure 7 As shown, the method may include the following steps.

[0226] Step S701: Input the 12 calibrated coefficients.

[0227] In this embodiment, 12 calibrated coefficients can be input, namely, α macroeconomic coefficient (α1-α4), β market competition coefficient (β1-β4), and γ operating efficiency coefficient (γ1-γ4).

[0228] Step S702: Prepare historical data.

[0229] In this embodiment, historical data can be prepared in advance.

[0230] Step S703: Use coefficients to predict costs.

[0231] In this embodiment, coefficients can be used for cost forecasting. For example, the α coefficient calibration method uses time series regression analysis to establish a model of the relationship between costs and macroeconomic variables. The β coefficient calibration method uses demand price elasticity models, competition intensity assessment models, and other methods. The γ coefficient calibration method uses economies of scale models, learning curve models, supply chain efficiency assessment models, and other methods.

[0232] Step S704: Calculate the prediction error.

[0233] In this embodiment, the prediction error can be calculated.

[0234] Step S705, error distribution analysis.

[0235] In this embodiment, error distribution analysis can be performed.

[0236] Step S706: Is MAE < 5% and RMSE < 8%?

[0237] In this embodiment, it can be determined whether MAE < 5% and RMSE < 8%. If MAE < 5% and RMSE < 8%, then step S707 is executed. Otherwise, step S708 or step S709 is executed.

[0238] Step S707, coefficients are valid.

[0239] In this embodiment, the coefficients are valid if MAE < 5% and RMSE < 8%.

[0240] Step S708, the coefficient part is valid.

[0241] In this embodiment, if either MAE < 5% or RMSE < 8% is not met, then the coefficient portion is valid.

[0242] Step S709, coefficients are invalid.

[0243] In this embodiment, if neither MAE < 5% nor RMSE < 8% is satisfied, the coefficient is invalid.

[0244] Step S710: Output the verification report.

[0245] In this embodiment, the judgment result can be output as a verification report.

[0246] The following section will further illustrate this point using the application of economic coefficient calibration in electronic and electrical systems as an example. Figure 8 This is a schematic diagram illustrating a comparison of coefficient calibration results according to an embodiment of this application, as shown below. Figure 8 As shown.

[0247] The α coefficients are calibrated as follows: α1 Inflation rate coefficient = 0.85 (obtained through regression analysis of CPI and cost data from 2020-2024). α2 Exchange rate volatility coefficient = 1.2 (chip imports account for 60%, resulting in high exchange rate sensitivity). α3 Interest rate change coefficient = 0.3 (financing costs account for a relatively low proportion). α4 GDP growth rate coefficient = 0.5 (moderate correlation with market demand).

[0248] The β coefficients are as follows: β1 Price Elasticity Coefficient = -1.8 (High price sensitivity in the smart cockpit market). β2 Competition Intensity Coefficient = 0.75 (Intense market competition). β3 Brand Premium Coefficient = 0.15 (Moderate brand premium capability). β4 Technological Leap Coefficient = -0.2 (Technological leadership brings cost advantages). The γ coefficients are as follows: γ1 Economies of Scale Coefficient = 0.15 (Increasing annual output from 100,000 to 500,000 reduces unit cost by 18%). γ2 Learning Curve Coefficient = 0.12 (Cumulative output doubling reduces cost by 12%). γ3 Supply Chain Efficiency Coefficient = 0.08 (Supply chain optimization leads to an 8% cost reduction). γ4 Time Decay Coefficient = 0.05 (Increased technological maturity leads to a 5% cost reduction). In this embodiment, coefficient verification can be performed. The calibrated coefficients are used to predict the costs for Q1-Q3 of 2024. The prediction error MAE=3.2% and RMSE=5.8%, which meets the validity criteria. The application effect can be obtained. After applying this coefficient system, the cost prediction accuracy rate increased from the original 68% to 92%, and the cost fluctuation early warning accuracy rate reached 87%.

[0249] Specifically, Table 1 is a comparison table of coefficient calibration results according to an embodiment of this application.

[0250] Table 1 Comparison of Coefficient Calibration Results

[0251]

[0252] Table 2 is a comparison table of cost forecast accuracy according to an embodiment of this application. As shown in Table 2, it illustrates the accuracy and forecast error for different months. Specifically, Figure 9 This is a schematic diagram illustrating a comparison of cost prediction accuracy results according to an embodiment of this application, as shown below. Figure 9 The figure shows the accuracy and prediction error for different months.

[0253] Table 2 Comparison of Cost Forecast Accuracy

[0254]

[0255] The following explanation uses the application of full-vehicle economic coefficient calibration as an example. Twelve economic coefficients are calibrated for each of the seven product systems (body, chassis, powertrain, electronics and electrical systems, interior and exterior trim, intelligent connectivity, and new energy). A coefficient database is established, storing 84 coefficient values ​​(7 systems × 12 coefficients) and their calibration basis. A dynamic adjustment mechanism is implemented: the α coefficient is updated quarterly, the β coefficient is updated semi-annually, and the γ coefficient is updated annually. Coefficient validity is verified: historical data backtesting is performed on the 84 coefficients, and 81 coefficients meet the validity standard (MAE < 5%), achieving an effectiveness rate of 96.4%. Through the above methods, the accuracy of full-vehicle cost prediction is improved by 30%, cost control efficiency is improved by 45%, and annual cost savings reach 23 million yuan.

[0256] According to an embodiment of this application, a vehicle data adjustment device is also provided. It should be noted that this vehicle data adjustment device can be used to perform the vehicle data adjustment method described in the embodiments.

[0257] Figure 10 This is a schematic diagram of a vehicle data adjustment device according to an embodiment of this application, such as... Figure 10 As shown, the data adjustment device 1000 for the vehicle may include: an acquisition unit 1002, a calibration unit 1004, a verification unit 1006, and an adjustment unit 1008.

[0258] The acquisition unit 1002 is used to acquire multi-source initial operating data of the vehicle, wherein the multi-source initial operating data is used to represent the degree of impact of operating data from different sources on the cost of the vehicle.

[0259] The calibration unit 1004 is used to call the target calibration model to calibrate the multi-source initial operating data and obtain multi-source target operating data. The target calibration model is trained using multi-source initial operating data samples from historical time periods. The multi-source initial operating data samples are used to represent the degree of influence of operating data from different sources on the cost of vehicle samples.

[0260] The verification unit 1006 is used to perform error analysis on the multi-source initial operating data and the multi-source target operating data, and obtain the analysis results. The analysis results are used to indicate whether the error between the multi-source initial operating data and the multi-source target operating data meets the error threshold.

[0261] The adjustment unit 1008 is used to adjust the multi-source target operation data in response to the analysis result that the error does not meet the error threshold. The accuracy of the adjusted multi-source target operation data is higher than that of the original multi-source target operation data.

[0262] Optionally, the multi-source target operating data includes at least one of the following: target economic data, target competition data, and target efficiency data. The target economic data represents the degree of influence of external economic dimensions on vehicle costs, the target competition data represents the degree of influence of different competition states on vehicle costs, and the target efficiency data represents the degree of influence of operating efficiency on vehicle costs. The calibration unit 1004 includes a first calibration subunit, used to call a first target calibration model to calibrate the multi-source initial operating data to obtain the target economic data. The first target calibration model is obtained by analyzing the multi-source initial operating data samples using time series regression.

[0263] Optionally, the calibration unit 1004 includes: a second calibration subunit, used to call the second target calibration model to perform calibration processing on the multi-source initial operating data to obtain target competition data, wherein the second target calibration model includes at least one of the following: a demand price elasticity model, a competition intensity assessment model, and a value attribute assessment model, wherein the demand price elasticity model is obtained by fitting the demand quantity of vehicles and the price information of vehicles, the competition intensity assessment model is obtained by fitting the price information and the average price information of vehicles of the same type, and the value attribute assessment model is obtained by fitting the difference between the price information and the average price information.

[0264] Optionally, the calibration unit 1004 includes a third calibration subunit, used to call a third target calibration model to calibrate the multi-source initial operating data and obtain target efficiency data. The third target calibration model includes at least one of the following: an economies of scale model, a learning curve model, a supply chain efficiency assessment model, or a time decay model. The economies of scale model is obtained by fitting a power function between the vehicle's output and its cost. The learning curve model is obtained by fitting the vehicle's cost and its total cumulative output. The supply chain efficiency assessment model is obtained by fitting the relationship between the vehicle's cost and its delivery risk. The time decay model is obtained by fitting the exponential relationship between the time change over different time periods and the vehicle's cost.

[0265] Optionally, the first calibration subunit includes: a first processing subunit, used to call the first target calibration model to perform calibration processing on the multi-source initial operating data to obtain first target economic data, second target economic data, third target economic data and fourth target economic data, wherein the first target economic data is used to represent the inflation rate, the second target economic data is used to represent exchange rate fluctuations, the third target economic data is used to represent interest rate changes, the fourth target economic data is used to represent the target GDP growth rate, the inflation rate is used to measure the upward pressure on vehicle costs, the exchange rate fluctuations are used to reflect the elasticity of vehicle costs, the interest rate changes are used to assess the impact on vehicle costs, and the target GDP growth rate is used to affect the scale effect corresponding to vehicle costs.

[0266] Optionally, the second calibration subunit includes: a second processing subunit, used to call a demand price elasticity model to calibrate the multi-source initial operating data and obtain first target competition data, wherein the first target competition data is used to represent the price elasticity coefficient of vehicle cost; a third processing subunit, used to call a competition intensity assessment model to calibrate the multi-source initial operating data and obtain second target competition data, wherein the second target competition data is used to represent the competition intensity coefficient of vehicle cost; a fourth processing subunit, used to call a value assessment attribute model to calibrate the multi-source initial operating data and obtain third target competition data, wherein the third target competition data is used to represent the premium coefficient of vehicle cost; the vehicle data adjustment device 1000 further includes: an evaluation unit, used to perform technical evaluation on the multi-source initial operating data and obtain evaluation results; and a processing unit, used to calibrate the multi-source initial operating data based on the evaluation results and obtain fourth target competition data, wherein the fourth target competition data is used to represent the technical generation gap coefficient of vehicle cost.

[0267] Optionally, the third calibration subunit includes: a fifth processing subunit, used to call a scale economy model to calibrate the multi-source initial operating data and obtain first target efficiency data, wherein the first target efficiency data is used to represent the scale effect coefficient of vehicle cost; a sixth processing subunit, used to call a learning curve model to calibrate the multi-source initial operating data and obtain second target efficiency data, wherein the second target efficiency data is used to represent the learning curve coefficient of vehicle cost; a seventh processing subunit, used to call a supply chain efficiency assessment model to calibrate the multi-source initial operating data and obtain third target efficiency data, wherein the third target efficiency data is used to represent the supply chain efficiency coefficient of vehicle cost; and an eighth processing subunit, used to call a time decay model to calibrate the multi-source initial operating data and obtain fourth target efficiency data, wherein the fourth target efficiency data is used to represent the time decay coefficient of vehicle cost.

[0268] Optionally, the error includes a first error and a second error, the first error being used to represent the mean absolute error and the second error being used to represent the root mean square error. The error threshold includes a first error threshold and a second error threshold. The data adjustment device 1000 for the vehicle further includes: a determining unit, used to determine the analysis result as the error satisfying the error threshold in response to the first error being less than the first error threshold and the second error being less than the second error threshold.

[0269] Optionally, the adjustment unit 1008 includes: an adjustment subunit, configured to determine that the error does not meet the error threshold in response to a first error being greater than or equal to a first error threshold, or a second error being greater than or equal to a second error threshold, and to adjust the multi-source target operating data within the target time period.

[0270] In this embodiment, the acquisition unit 1002 acquires multi-source initial operating data of the vehicle, wherein the multi-source initial operating data is used to represent the degree of influence of operating data from different sources on the cost of the vehicle; the calibration unit 1004 calls the target calibration model to calibrate the multi-source initial operating data to obtain multi-source target operating data, wherein the target calibration model is trained using multi-source initial operating data samples from historical time periods, and the multi-source initial operating data samples are used to represent the degree of influence of operating data from different sources on the cost of the vehicle sample; the verification unit 1006 performs error analysis on the multi-source initial operating data and the multi-source target operating data to obtain analysis results, wherein the analysis results are used to indicate whether the error between the multi-source initial operating data and the multi-source target operating data meets the error threshold; the adjustment unit 1008 adjusts the multi-source target operating data in response to the analysis result indicating that the error does not meet the error threshold, wherein the accuracy of the adjusted multi-source target operating data is higher than that of the multi-source target operating data, thereby solving the technical problem of low accuracy in adjusting vehicle data and achieving the technical effect of improving the accuracy of vehicle data adjustment.

[0271] 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, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0272] According to another aspect of the embodiments of this application, a processor is also provided. The processor is used to run a program, wherein the program executes the methods of the embodiments of this application during runtime.

[0273] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the method of the embodiments of this application when it runs.

[0274] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided. The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to perform the method of the embodiments of this application.

[0275] According to another aspect of the embodiments of this application, a vehicle is also provided. The vehicle includes a memory and a processor. The memory stores an executable program; the processor is used to run the program, which, when running, implements the methods described in the embodiments of this application.

[0276] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0277] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0278] 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 units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0279] Furthermore, the functional units in the various embodiments of this application 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. The integrated unit can be implemented in hardware or as a software functional unit.

[0280] If the integrated unit 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 application, in essence, or the part that contributes to the prior art, or all or 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 described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0281] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for adjusting vehicle data, characterized in that, include: Acquire multi-source initial operating data of the vehicle, wherein the multi-source initial operating data is used to represent the degree of impact of operating data from different sources on the cost of the vehicle; The target calibration model is invoked to calibrate the multi-source initial operating data to obtain multi-source target operating data. The target calibration model is trained using multi-source initial operating data samples from historical time periods. The multi-source initial operating data samples are used to represent the degree of influence of operating data from different sources on the cost of vehicle samples. Error analysis is performed on the multi-source initial operating data and the multi-source target operating data to obtain analysis results, wherein the analysis results are used to indicate whether the error between the multi-source initial operating data and the multi-source target operating data meets the error threshold; In response to the analysis result that the error does not meet the error threshold, the multi-source target operation data is adjusted, wherein the accuracy of the adjusted multi-source target operation data is higher than that of the original multi-source target operation data.

2. The method according to claim 1, characterized in that, The multi-source target operation data includes at least one of the following: target economic data, target competition data, and target efficiency data, wherein the target economic data is used to represent the degree of influence of external economic dimensions on the cost of the vehicle, the target competition data is used to represent the degree of influence of different competition states on the cost of the vehicle, and the target efficiency data is used to represent the degree of influence of operational efficiency on the cost of the vehicle. A target calibration model is invoked to calibrate the multi-source initial operation data to obtain the multi-source target operation data, including: The first target calibration model is invoked to calibrate the multi-source initial operating data to obtain the target economic data. The first target calibration model is obtained by analyzing the multi-source initial operating data sample using time series regression.

3. The method according to claim 2, characterized in that, The target calibration model is invoked to calibrate the multi-source initial running data, resulting in multi-source target running data, including: The second target calibration model is invoked to calibrate the multi-source initial operating data to obtain the target competition data. The second target calibration model includes at least one of the following: a demand price elasticity model, a competition intensity assessment model, and a value attribute assessment model. The demand price elasticity model is obtained by fitting the demand quantity of the vehicle and the price information of the vehicle. The competition intensity assessment model is obtained by fitting the price information and the average price information of the same type of vehicle. The value attribute assessment model is obtained by fitting the difference between the price information and the average price information.

4. The method according to claim 2, characterized in that, The target calibration model is invoked to calibrate the multi-source initial running data, resulting in multi-source target running data, including: The third objective calibration model is invoked to calibrate the multi-source initial operating data to obtain the objective efficiency data. The third objective calibration model includes at least one of the following: an economies of scale model, a learning curve model, a supply chain efficiency assessment model, or a time decay model. The economies of scale model is obtained by fitting a power function to the vehicle's output and cost. The learning curve model is obtained by fitting the vehicle's cost and total cumulative output. The supply chain efficiency assessment model is obtained by fitting the relationship between the vehicle's cost and the vehicle's delivery risk. The time decay model is obtained by fitting an exponential relationship between the time change over different time periods and the vehicle's cost.

5. The method according to claim 2, characterized in that, The first target calibration model is invoked to calibrate the multi-source initial operating data, resulting in the target economic data, including: The first target calibration model is invoked to calibrate the multi-source initial operating data, resulting in first target economic data, second target economic data, third target economic data, and fourth target economic data. The first target economic data represents the inflation rate, the second target economic data represents exchange rate fluctuations, the third target economic data represents interest rate changes, and the fourth target economic data represents the target GDP growth rate. The inflation rate measures the upward pressure on vehicle costs, the exchange rate fluctuations reflect the cost elasticity of the vehicle, the interest rate changes affect the assessment of vehicle costs, and the target GDP growth rate affects the scale effect corresponding to the vehicle's costs.

6. The method according to claim 3, characterized in that, The second target calibration model is invoked to calibrate the multi-source initial running data, resulting in the target competition data, including: The demand price elasticity model is invoked to calibrate the multi-source initial operating data to obtain first target competition data, wherein the first target competition data is used to represent the price elasticity coefficient of the vehicle's cost; The competition intensity assessment model is invoked to calibrate the multi-source initial operating data to obtain second target competition data, wherein the second target competition data is used to represent the competition intensity coefficient of the vehicle's cost; The value assessment attribute model is invoked to calibrate the multi-source initial operating data, resulting in third target competition data; wherein, the third target competition data is used to represent the premium coefficient of the vehicle's cost; The method further includes: performing technical evaluation on the multi-source initial operating data to obtain evaluation results; and calibrating the multi-source initial operating data based on the evaluation results to obtain fourth target competition data, wherein the fourth target competition data is used to represent the technical generation gap coefficient of the vehicle cost.

7. The method according to claim 4, characterized in that, The third target calibration model is invoked to calibrate the multi-source initial operating data, resulting in the target efficiency data, including: The scale economy model is invoked to calibrate the multi-source initial operating data to obtain first target efficiency data, wherein the first target efficiency data is used to represent the scale effect coefficient of the vehicle cost; The learning curve model is invoked to calibrate the multi-source initial operating data to obtain second target efficiency data, wherein the second target efficiency data is used to represent the learning curve coefficient of the vehicle cost; The supply chain efficiency assessment model is invoked to calibrate the multi-source initial operating data to obtain third target efficiency data, wherein the third target efficiency data is used to represent the supply chain efficiency coefficient of the vehicle cost; The time decay model is invoked to calibrate the multi-source initial operating data to obtain fourth target efficiency data, wherein the fourth target efficiency data is used to represent the time decay coefficient of the vehicle cost.

8. The method according to claim 1, characterized in that, The error includes a first error and a second error, wherein the first error represents the mean absolute error and the second error represents the root mean square error; the error threshold includes a first error threshold and a second error threshold; and the method further includes: In response to the first error being less than the first error threshold and the second error being less than the second error threshold, the analysis result is determined to be that the error satisfies the error threshold.

9. The method according to claim 8, characterized in that, In response to the analysis result indicating that the error does not meet the error threshold, the multi-source target operating data is adjusted, including: In response to the first error being greater than or equal to the first error threshold, or the second error being greater than or equal to the second error threshold, it is determined that the error does not meet the error threshold, and the multi-source target operation data is adjusted within the target time period.

10. A vehicle data adjustment device, characterized in that, include: An acquisition unit is used to acquire multi-source initial operating data of the vehicle, wherein the multi-source initial operating data is used to represent the degree of impact of operating data from different sources on the cost of the vehicle; The calibration unit is used to call the target calibration model to calibrate the multi-source initial operating data to obtain multi-source target operating data. The target calibration model is trained using multi-source initial operating data samples from historical time periods. The multi-source initial operating data samples are used to represent the degree of influence of operating data from different sources on the cost of vehicle samples. The verification unit is used to perform error analysis on the multi-source initial running data and the multi-source target running data to obtain the analysis result, wherein the analysis result is used to indicate whether the error between the multi-source initial running data and the multi-source target running data meets the error threshold. An adjustment unit is configured to adjust the multi-source target operating data in response to the analysis result indicating that the error does not meet the error threshold, wherein the accuracy of the adjusted multi-source target operating data is higher than that of the original multi-source target operating data.

11. A processor, characterized in that, The processor is used to run a program, wherein the program executes the method according to any one of claims 1 to 9 when it runs.

12. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 9.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform the method according to any one of claims 1 to 9.

14. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 9.

Images