Business demand data prediction method and device and nonvolatile storage medium

Abstract

The invention discloses a business demand data prediction method and device and a nonvolatile storage medium. The method comprises the following steps: obtaining historical work order data and historical environment data of a first level of a target service in a preset historical period, the historical work order data being a time sequence of a historical work order number, and the historical environment data being external environment data influencing a service demand of the target service; the first hierarchy is the highest hierarchy in a plurality of demand hierarchies of the target business; on the basis of the historical work order data and the historical environment data, the number of future work orders of the target business in multiple levels in a preset future period is predicted; and obtaining respective business demand prediction results of the plurality of levels based on the respective future work order numbers of the plurality of levels. According to the invention, the technical problem that the current prediction method depends on artificial experience judgment and subjective estimation and uses a simple statistical model or a single algorithm for prediction, resulting in large prediction result error is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis and prediction, in particular to a business demand data prediction method and device and a non-volatile storage medium. BACKGROUND

[0002] In the management of business expansion projects of power enterprises, the prediction of demand data of business projects, such as the prediction and allocation of investment budget, has always been a complex and critical task, which requires the accurate integration of multi-dimensional data such as work order quantity, investment amount, seasonal changes, etc. to ensure the efficient promotion of business and the optimal allocation of resources. However, the traditional budget prediction methods currently relied on, although certain efforts have been made in data processing, still have significant technical bottlenecks, which directly limit the accuracy and flexibility of budget management.

[0003] Specifically, the budget prediction of power business expansion projects, including new user access, equipment capacity expansion, etc., involves not only simple data aggregation, but also comprehensive consideration of different levels (such as provinces, cities, and counties) and types of business (such as high-voltage and low-voltage work orders). However, the existing technologies, such as statistical models and single time series algorithms, are usually based on the comparison of historical business size and the application of empirical coefficients, which ignores the multi-level characteristics and type diversity of business development, resulting in a serious disconnection between the prediction results and the actual scene, making it difficult to cope with the complexity and uncertainty of business development. In particular, it fails to capture the subtle fluctuations of city and county-level work orders, and lacks robustness in handling nonlinear relationships and periodic changes, thus making the accuracy of budget allocation and the efficiency of resource allocation low.

[0004] Currently, there is no effective solution to the above problems. SUMMARY

[0005] The embodiments of the present application provide a business demand data prediction method, device and non-volatile storage medium to at least solve the technical problem that the current prediction method relies on artificial experience judgment and subjective estimation, uses simple statistical models or single algorithms for prediction, resulting in large prediction error.

[0006] According to an aspect of some embodiments of the present application, a service demand data prediction method is provided, comprising: obtaining historical work order data of a target service at a first level in a preset historical period and historical environment data, wherein the historical work order data is a time series of historical work order quantity, the historical environment data is external environment data that has an impact on the service demand of the target service, and the first level is the highest level of a plurality of demand levels of the target service; predicting the future work order quantity of the target service at each of the plurality of levels in a preset future period based on the historical work order data and the historical environment data, wherein the plurality of levels includes the first level; and obtaining a service demand prediction result of each of the plurality of levels based on the future work order quantity of each of the plurality of levels.

[0007] Optionally, obtaining the historical work order data of the target service at the first level in the preset historical period comprises: obtaining initial work order data of the target service at the first level in the preset historical period; and performing abnormal correction processing on the initial work order data to obtain the historical work order data, wherein the abnormal correction processing includes replacement processing, deletion processing, and filtering processing, the replacement processing represents replacing abnormal data in the initial work order data with the mean of the remaining normal data, the remaining normal data is data in the initial work order data other than the abnormal data, the deletion processing represents deleting data that occurs repeatedly in the initial work order data, and the filtering processing represents white noise filtering processing of data in the initial work order data whose fluctuation frequency exceeds a preset frequency threshold.

[0008] Optionally, predicting the future work order quantity of the target service at each of the plurality of levels in the preset future period based on the historical work order data and the historical environment data comprises: predicting the future work order quantity of the target service at the first level in the preset future period based on the historical work order data and the historical environment data by using a plurality of preset prediction algorithms to obtain a first prediction result corresponding to each of the plurality of prediction algorithms; determining a weight corresponding to each of the plurality of prediction algorithms; performing weighted summation on the first prediction result corresponding to each of the plurality of prediction algorithms based on the weight corresponding to each of the plurality of prediction algorithms to obtain the future work order quantity of the first level; predicting the future work order quantity of the target service at a second level in the preset future period based on the future work order quantity of the first level, the historical work order data, and the historical environment data by using the plurality of prediction algorithms to obtain a second prediction result corresponding to each of the plurality of prediction algorithms, wherein the second level is a next level of service demand level of the first level; and continuing to repeat the above steps to predict the future work order quantity of the next level based on the future work order quantity of the previous level until the future work order quantity of each of the plurality of levels is obtained.

[0009] Optionally, the plurality of prediction algorithms includes a Prophet prophet prediction algorithm, a TAR threshold autoregressive algorithm, and an FFT fast Fourier transform algorithm.

[0010] Optionally, determining the weights corresponding to each of the multiple prediction algorithms includes: using multiple prediction algorithms, based on historical work order data and historical environment data, predicting the number of future work orders at the first level of the target business within a preset test period, and obtaining the test prediction results corresponding to each of the multiple prediction algorithms; obtaining the actual number of work orders at the first level of the target business within the preset test period, as the actual collection result; calculating the sum of squared errors between the test prediction results and the actual collection results corresponding to each of the multiple prediction algorithms, as the prediction error results corresponding to each of the multiple prediction algorithms; calculating the reciprocal of the prediction error results corresponding to each of the multiple prediction algorithms; and determining the weights corresponding to each of the multiple prediction algorithms based on the reciprocals of the prediction error results corresponding to each of the multiple prediction algorithms.

[0011] Optionally, based on the future work order quantity of each of the multiple levels, the business demand forecast results for each of the multiple levels are obtained, including: obtaining the work order unit demand data of the target business within a preset historical period; calculating the product of the future work order quantity and the work order unit demand data for each of the multiple levels to obtain the initial demand forecast results for each of the multiple levels; calculating the optimized demand forecast difference for each of the multiple levels based on the initial demand forecast results for each of the multiple levels; and determining the business demand forecast results for each of the multiple levels based on the initial demand forecast results for each of the multiple levels and the optimized demand forecast difference for each of the multiple levels.

[0012] According to another aspect of the present invention, a business demand data prediction device is also provided, comprising: an acquisition module, configured to acquire historical work order data and historical environment data of a target business at a first level within a preset historical period, wherein the historical work order data is a time series of the number of historical work orders, the historical environment data is external environment data that affects the business demand of the target business, and the first level is the highest level among multiple demand levels of the target business; a first prediction module, configured to predict the future number of work orders of the target business at each of multiple levels within a preset future period based on the historical work order data and the historical environment data, wherein the multiple levels include the first level; and a second prediction module, configured to obtain business demand prediction results for each of the multiple levels based on the future number of work orders for each of the multiple levels.

[0013] According to another aspect of the present invention, a non-volatile storage medium is also provided, the non-volatile storage medium including a stored program, wherein, when the program is running, it controls the device where the non-volatile storage medium is located to execute any of the above-described business demand data prediction methods.

[0014] According to another aspect of the present invention, a computer device is also provided, the computer device including a processor, the processor being configured to run a program, wherein the program executes any of the above-described business demand data prediction methods during runtime.

[0015] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements any of the above-described business requirement data prediction methods.

[0016] In this embodiment of the invention, a business demand data prediction method is adopted. This method acquires historical work order data and historical environment data of the target business at the first level within a preset historical period. The historical work order data is a time series of the number of historical work orders, and the historical environment data is external environment data that affects the business demand of the target business. The first level is the highest level among multiple demand levels of the target business. Based on the historical work order data and historical environment data, the future number of work orders for each of the multiple levels of the target business within a preset future period is predicted, where the multiple levels include the first level. Based on the future number of work orders for each of the multiple levels, the business demand prediction results for each of the multiple levels are obtained. This achieves the goal of multi-dimensional prediction of business demand data, thereby improving the accuracy and robustness of prediction and optimizing resource allocation efficiency. Furthermore, it solves the technical problem that current prediction methods rely on manual experience and subjective estimation, using simple statistical models or single algorithms for prediction, resulting in large prediction errors. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0018] Figure 1 A hardware structure block diagram of a computer terminal for implementing a business demand data prediction method is shown.

[0019] Figure 2 This is a flowchart illustrating the business demand data prediction method provided in an embodiment of the present invention;

[0020] Figure 3 This is a schematic diagram of the architecture of an integrated prediction model provided by an optional embodiment of the present invention;

[0021] Figure 4 This is a structural block diagram of a business demand data prediction device provided according to an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] According to an embodiment of the present invention, a business demand data prediction method embodiment 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.

[0025] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal for implementing a business demand data prediction method is shown. Figure 1 As shown, the computer terminal 10 may include one or more processors (shown as 102a, 102b, ..., 102n in the figure) (the processor may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the sameFigure 1 The different configurations shown.

[0026] It should be noted that the aforementioned one or more processors and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10. As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0027] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the business demand data prediction method in this embodiment of the invention. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned business demand data prediction method of the application. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0028] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10.

[0029] Figure 2 This is a flowchart illustrating the business demand data prediction method provided by an embodiment of the present invention, as shown below. Figure 2 As shown, the method includes the following steps:

[0030] Step S201: Obtain historical work order data and historical environment data of the first level of the target business within a preset historical period. The historical work order data is a time series of the number of historical work orders, and the historical environment data is external environment data that affects the business needs of the target business. The first level is the highest level among the multiple demand levels of the target business.

[0031] In this step, to build an accurate predictive model for business demand data (such as budget), the first step is to comprehensively collect historical work order data and historical environmental data for the target business within a preset historical period, starting from the highest level, the first level, namely the provincial level. Historical work order data specifically represents a time series of provincial work orders, recording the total number of work orders at each point in time (e.g., each month) within a specific past period (e.g., the past 16 months), including but not limited to the number of work orders for new user access, equipment capacity expansion, etc. Changes in these work order numbers reflect the fluctuating characteristics of business demand. Historical environmental data covers various external environmental factors that influence the target business demand, such as holidays, policy announcements, temperature changes, and economic development levels in the time series. This data helps the model identify and understand non-internal driving factors of business demand, such as the impact of high summer temperatures on electricity demand and the stimulating effect of policy adjustments on new installations.

[0032] The process of acquiring this data involves automated data collection and processing. It involves extracting data from multiple sources, such as online systems. Raw data (e.g., work order IDs, timestamps, amounts) is parsed into structured formats (CSV or database tables) and stored in a temporary buffer. Historical data collection extends beyond internal work order records to external data sources, such as holiday calendars and economic indicator reports, accessed via API interfaces, further enriching the model's input dimensions. The combination of historical work order data and historical environmental data provides a solid foundation for building a multi-level model. The highest-level provincial data is particularly important because it provides an overview of the province's business scale and trends, offering macro-level guidance for more refined city and county-level forecasts. Simultaneously, by identifying and analyzing external environmental data, the model can learn how to make accurate predictions in complex and ever-changing environments, thereby improving the overall robustness and adaptability of the forecast. This data collection and processing step is a crucial prerequisite for ensuring the effective operation of the subsequent three-level, nine-category integrated forecasting model, laying the data foundation for achieving multi-dimensional and accurate forecasting.

[0033] Step S202: Based on historical work order data and historical environment data, predict the number of future work orders for the target business at multiple levels within a preset future period, where the multiple levels include the first level.

[0034] In this step, firstly, by collecting and organizing work order records and related environmental factors such as holidays, policy adjustments, and seasonal changes within a specific historical period (e.g., the past 16 months) of the target business, a multi-dimensional dataset containing time-series characteristics is constructed. This dataset not only reflects the internal dynamics of business demand but also incorporates external variables affecting demand, providing comprehensive learning samples for the prediction model. Subsequently, using this dataset, the prediction model adopts a multi-level prediction architecture, with the highest level, the first level, being the provincial-level work order prediction. Figure 3This is a schematic diagram of the architecture of an integrated prediction model provided by an optional embodiment of the present invention, such as... Figure 3 As shown, the model analyzes the time series of historical work order quantities at the provincial level and combines it with historical environmental data to learn the fluctuation patterns of work order volume over time and its correlation with external factors. It then predicts the quantity trends of various types of work orders (such as new high-voltage installations and low-voltage capacity expansions) at the provincial level within a preset future period (e.g., the next 16 months), including summary, work order type, and demand type levels. This also provides a foundation and guidance for forecasting at the next level, such as the city and county levels. The forecasting process can comprehensively utilize multiple algorithms, such as the Prophet model to handle seasonality and holiday effects, the Threshold Autoregressive (TAR) model to capture nonlinear changes, and the Fast Fourier Transform (FFT) to identify periodic patterns. The algorithm weights are dynamically adjusted using an inverse error weighting method to achieve accurate predictions of work order quantities at different levels and for different business types.

[0035] Step S203: Based on the number of future work orders at each of the multiple levels, obtain the business demand forecast results for each of the multiple levels.

[0036] In this step, an integrated model is constructed to analyze the periodic, trend-based changes and nonlinear relationships in the time series, obtaining a predicted number of work orders for each level within a preset future period. These predicted work order numbers are then transformed into specific business demand forecasts. For example, for high-voltage new installation work orders, the forecast not only reflects the number of work orders that may be processed in the future but also relates to the actual needs behind each work order, such as equipment procurement and line construction. By calculating the average cost or investment requirements of historical work orders, the budget amount corresponding to future work orders is derived. This transformation mechanism takes into account the specificity and differences of different levels and business types, ensuring an accurate match between budget allocation and actual business needs. It provides data support for resource planning for business expansion projects at all levels, achieving a leap from abstract work order forecasting to specific business demand forecasting, and serves as a key bridge connecting budget forecasting and business execution.

[0037] Through the above steps, the goal of multi-dimensional prediction of business demand data is achieved, thereby improving the accuracy and robustness of prediction and optimizing resource allocation efficiency. This solves the technical problem that current prediction methods rely on human experience and subjective estimation, and use simple statistical models or single algorithms for prediction, resulting in large errors in the prediction results.

[0038] As an optional embodiment, obtaining historical work order data of the first level of the target business within a preset historical period includes: obtaining initial work order data of the first level of the target business within a preset historical period; performing anomaly correction processing on the initial work order data to obtain historical work order data, wherein the anomaly correction processing includes replacement processing, deletion processing, and filtering processing. The replacement processing represents replacing the abnormal data in the initial work order data with the average of the remaining normal data, and the remaining normal data is the data in the initial work order data other than the abnormal data. The deletion processing represents deleting the duplicate data in the initial work order data. The filtering processing represents performing white noise filtering on the data in the initial work order data whose fluctuation frequency exceeds a preset frequency threshold.

[0039] Optionally, the program can first automatically extract initial work order data for the target business from data sources such as online systems. This data covers the quantity, type, and time series of historical work orders, forming a preliminary dataset. Subsequently, anomaly correction processing is performed on this initial work order data to improve data quality and ensure the accuracy of model training. Anomaly correction processing can include: replacement processing, where the mean of normal data from the same period is calculated and used to replace abnormal data caused by sudden events (such as a sharp decrease in work order volume) to smooth data fluctuations; deletion processing, removing duplicate records caused by system upgrades or data entry errors to prevent the model from learning incorrect work order trends; and filtering processing, applying white noise filtering algorithms, such as moving average filtering, to smooth high-frequency fluctuating work orders (such as metering equipment failure work orders) and reduce white noise interference. Through these processing steps, the resulting historical work order data is cleaner and more reliable, not only accurately reflecting the periodic changes and long-term trends of business needs but also effectively reducing the impact of outliers and noise on the prediction model. This provides a solid data foundation for subsequently building a multi-dimensional integrated prediction model, ensuring the accuracy and robustness of the prediction results.

[0040] As an optional embodiment, based on historical work order data and historical environment data, the number of future work orders for the target business at multiple levels within a preset future period is predicted, including: using multiple preset prediction algorithms, based on historical work order data and historical environment data, to predict the number of future work orders for the target business at the first level within the preset future period, obtaining a first prediction result corresponding to each of the multiple prediction algorithms; determining the weights corresponding to each of the multiple prediction algorithms; weighted summing of the first prediction results corresponding to each of the multiple prediction algorithms based on their respective weights, to obtain the number of future work orders at the first level; using multiple prediction algorithms, based on the number of future work orders at the first level, historical work order data, and historical environment data, to predict the number of future work orders for the target business at the second level within the preset future period, obtaining a second prediction result corresponding to each of the multiple prediction algorithms, wherein the second level is the next level of business demand level below the first level; continuing to repeat the above steps, predicting the number of future work orders at the next level based on the number of future work orders at the previous level, until multiple levels are traversed to obtain the number of future work orders for each of the multiple levels.

[0041] Optionally, by integrating multiple prediction algorithms such as Prophet, Threshold Autoregression (TAR), and Fast Fourier Transform (FFT), and based on historical work order data at the provincial level (first level) and historical environmental data including information such as holidays and policy changes, preliminary predictions of the future number of work orders are first generated, i.e., the first prediction results. Subsequently, the weights of each algorithm are dynamically calculated using the inverse error weighting method to ensure the model's adaptability and accuracy. Based on the algorithm weights, the first prediction results are weighted and summed to obtain the predicted value of the future number of work orders at the provincial level. This process is then progressively delegated to the city level (second level) and county level. The prediction at each level takes the prediction results of the previous level, its own historical work order data, and historical environmental data as input to produce a prediction of the future number of work orders, i.e., the second prediction results, and so on. The dynamic adjustment of weights and the progressive weighted summation of prediction results ensure that the prediction model can effectively cope with data anomalies and business fluctuations, achieving multi-level and multi-dimensional prediction of the number of work orders from macro to micro perspectives. This top-down forecasting mechanism not only improves the accuracy of forecasts but also ensures a close match between budget allocation and business needs, providing strong support for the refined management of power grid companies' business expansion projects.

[0042] As an optional embodiment, several prediction algorithms include the Prophet prediction algorithm, the TAR threshold autoregressive algorithm, and the FFT fast Fourier transform algorithm.

[0043] Optionally, the selection of multiple prediction algorithms aims to capture and analyze the characteristics of work order data from different dimensions to achieve more accurate budget forecasting. The Prophet prediction algorithm excels at handling time-series data with significant seasonality and holiday effects, decomposing trend terms, seasonal terms, and holiday effects to fit long-term patterns and effectively predict the fluctuation trends of work orders for business expansion projects over time. The TAR threshold autoregressive algorithm is used to detect and predict nonlinear abrupt changes in the data, such as sudden increases or decreases in the number of work orders caused by policy changes or market shocks, enhancing the model's responsiveness to abnormal events. The FFT (Fast Fourier Transform) algorithm focuses on identifying periodic patterns in time series, such as the periodic increase in work orders corresponding to seasonal electricity peaks, accurately grasping periodic business demand through frequency domain analysis of the signal. These algorithms complement each other, dynamically adjusting weights through an inverse error weighting method to ensure the stability and accuracy of the prediction model when facing complex fluctuations and abnormal data, representing a key technological combination for achieving multi-level, multi-dimensional accurate prediction.

[0044] As an optional embodiment, determining the weights corresponding to each of the multiple prediction algorithms includes: using multiple prediction algorithms, based on historical work order data and historical environment data, predicting the number of future work orders at the first level of the target business within a preset test period, and obtaining the test prediction results corresponding to each of the multiple prediction algorithms; obtaining the actual number of work orders at the first level of the target business within the preset test period, as the actual collection result; calculating the sum of squared errors between the test prediction results and the actual collection results corresponding to each of the multiple prediction algorithms, as the prediction error results corresponding to each of the multiple prediction algorithms; calculating the reciprocal of the prediction error results corresponding to each of the multiple prediction algorithms; and determining the weights corresponding to each of the multiple prediction algorithms based on the reciprocals of the prediction error results corresponding to each of the multiple prediction algorithms.

[0045] Optionally, the weights of each prediction algorithm are determined by comparing the prediction results with the actual number of work orders. This aims to quantify the prediction accuracy of each algorithm and thus optimize the overall prediction performance of the model. First, prediction algorithms such as Prophet, TAR, and FFT are used to predict the future number of work orders at the provincial level (first tier) within a set test period (e.g., the latest 12 months) based on historical work order and environmental data, obtaining the corresponding test prediction results for each algorithm. Then, the actual number of work orders executed at the provincial level within the same test period is collected as a comparison benchmark. Next, the sum of squared errors between the prediction results of each algorithm and the actual number of work orders is calculated, i.e., the prediction error result, such as TAR error sum = Σ(predicted value - actual value)². To convert the deviation into a basis for weight allocation, the reciprocal of the prediction error result is calculated; a larger reciprocal indicates higher accuracy of the algorithm's prediction. Finally, an inverse error weighting method is used to allocate weights, where each weight is proportional to the reciprocal of the prediction error. This weight is calculated by the ratio of the reciprocal of each algorithm's individual prediction error to the sum of the reciprocals of all algorithms' prediction errors. This ensures that the model can automatically adjust the contribution of each algorithm based on its historical performance, improving the robustness and accuracy of the predictions. This weight determination mechanism is an indispensable part of the model training and prediction process. It dynamically reflects the contribution of each algorithm to the overall prediction effect and is a key step in achieving multi-dimensional, multi-level accurate predictions. Through this mechanism, the model can intelligently select the optimal prediction strategy based on the characteristics of historical data, providing more accurate business demand predictions for subsequent budget allocation.

[0046] As an optional embodiment, based on the future work order quantity of each of the multiple levels, the business demand forecast results of each of the multiple levels are obtained, including: obtaining the work order unit demand data of the target business within a preset historical period; calculating the product of the future work order quantity and the work order unit demand data of each of the multiple levels to obtain the initial demand forecast results of each of the multiple levels; calculating the optimized demand forecast difference of each of the multiple levels based on the initial demand forecast results of each of the multiple levels; and determining the business demand forecast results of each of the multiple levels based on the initial demand forecast results of each of the multiple levels and the optimized demand forecast difference of each of the multiple levels.

[0047] Optionally, after acquiring and processing the future work order quantities at each level, the program analyzes the work order unit demand data of the target business within the historical period, i.e., the average unit price (such as investment amount, manpower, materials, etc.) corresponding to each work order, and calculates the product of the future work order quantity and unit demand data at each level to obtain the initial demand forecast results for each level. For example, the program can query the project investment data of the previous year, calculate the work order unit price, and calculate the work order unit price as follows: work order unit price = actual investment amount of the project in the previous year / corresponding work order quantity. Multiplying the predicted work order quantity by the work order unit price yields the initial investment amount.

[0048] Subsequently, an optimized demand forecasting gap mechanism is introduced. Based on the input-output performance indicators of various cities, the initial forecast is adjusted to better reflect actual business needs and reduce over- or under-allocation of resources. This gap is automatically calculated by comparing the average efficiency of each city with the provincial average to determine the amount of budget that needs to be increased or decreased. Finally, the initial demand forecast results are combined with the optimized demand forecast gap to determine the business demand forecast results at each level, achieving multi-dimensional and multi-level accurate forecasting from the provincial to the county level, providing a scientific basis for the investment budget management of business expansion projects. For example, the program compares the input-output performance coefficients (such as economic benefits per unit of investment) of various cities. For units that are lower than the provincial average, the budget is deducted according to the difference ratio (maximum 5%): Deduction amount = Initial budget × min{5%, provincial average - city coefficient). The total deduction amount is redistributed according to the proportion of the city's performance coefficient. The redistributed amount_city k = total deduction amount × (city k coefficient / Σ city coefficient), and the final budget = initial budget + sum of redistributed amount - total deduction amount. The program generates a table of work order volume and investment budget for the next 16 months (which may include all results from the forecasting model, such as the total number of work orders and the number of work orders of various types), and exports it by province, city, and county level.

[0049] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0050] Through the above description of the embodiments, those skilled in the art can clearly understand that the business demand data prediction method according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0051] According to embodiments of the present invention, an apparatus for implementing the above-described business demand data prediction method is also provided. Figure 4 This is a structural block diagram of a business demand data prediction device provided according to an embodiment of the present invention, such as... Figure 4As shown, the device includes: an acquisition module 41, a first prediction module 42, and a second prediction module 43. The device will be described below.

[0052] The acquisition module 41 is used to acquire historical work order data and historical environment data of the first level of the target business within a preset historical period. The historical work order data is a time series of the number of historical work orders, and the historical environment data is external environment data that affects the business needs of the target business. The first level is the highest level among the multiple demand levels of the target business.

[0053] The first prediction module 42, connected to the acquisition module 41, is used to predict the number of future work orders for the target business at multiple levels within a preset future period based on historical work order data and historical environment data, wherein the multiple levels include the first level.

[0054] The second prediction module 43, connected to the first prediction module 42, is used to obtain the business demand prediction results for each of the multiple levels based on the number of future work orders for each level.

[0055] It should be noted that the acquisition module 41, the first prediction module 42, and the second prediction module 43 mentioned above correspond to steps S201 to S203 in the embodiments. Multiple modules and their corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiments. It should also be noted that the above modules, as part of the device, can run on the computer terminal 10 provided in the embodiments.

[0056] Embodiments of the present invention may provide a computer device. Optionally, in this embodiment, the computer device may be located in at least one of a plurality of network devices in a computer network. The computer device includes a memory and a processor.

[0057] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the business demand data prediction method and apparatus in this embodiment of the invention. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the aforementioned business demand data prediction method. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to a computer terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0058] The processor can invoke information and application programs stored in the memory via a transmission device to perform the following steps: acquiring historical work order data and historical environment data of the first level of the target business within a preset historical period, wherein the historical work order data is a time series of the number of historical work orders, and the historical environment data is external environment data that affects the business needs of the target business; the first level is the highest level among multiple demand levels of the target business; based on the historical work order data and historical environment data, predicting the future number of work orders for each of the multiple levels of the target business within a preset future period, wherein the multiple levels include the first level; and based on the future number of work orders for each of the multiple levels, obtaining the business demand prediction results for each of the multiple levels.

[0059] Optionally, the processor may also execute program code for the following steps: obtaining historical work order data of the first level of the target business within a preset historical period, including: obtaining initial work order data of the first level of the target business within a preset historical period; performing anomaly correction processing on the initial work order data to obtain historical work order data, wherein the anomaly correction processing includes replacement processing, deletion processing and filtering processing, the replacement processing is characterized by replacing the abnormal data in the initial work order data with the average of the remaining normal data, the remaining normal data being the data in the initial work order data excluding the abnormal data, the deletion processing is characterized by deleting duplicate data in the initial work order data, and the filtering processing is characterized by performing white noise filtering processing on data in the initial work order data whose fluctuation frequency exceeds a preset frequency threshold.

[0060] Optionally, the processor may also execute program code with the following steps: based on historical work order data and historical environment data, predict the number of future work orders for the target business at multiple levels within a preset future period, including: using multiple preset prediction algorithms, based on historical work order data and historical environment data, predicting the number of future work orders for the first level of the target business within a preset future period, obtaining the first prediction result corresponding to each of the multiple prediction algorithms; determining the weights corresponding to each of the multiple prediction algorithms; based on the weights corresponding to each of the multiple prediction algorithms, weighted summing the first prediction results corresponding to each of the multiple prediction algorithms to obtain the number of future work orders at the first level; using multiple prediction algorithms, based on the number of future work orders at the first level, historical work order data, and historical environment data, predicting the number of future work orders for the target business at the second level within a preset future period, obtaining the second prediction result corresponding to each of the multiple prediction algorithms, wherein the second level is the next level of business demand level below the first level; continuing to repeat the above steps, predicting the number of future work orders at the next level based on the number of future work orders at the previous level, until multiple levels are traversed to obtain the number of future work orders for each of the multiple levels.

[0061] Optionally, the processor may also execute program code that includes multiple prediction algorithms, such as the Prophet prediction algorithm, the TAR threshold autoregressive algorithm, and the FFT fast Fourier transform algorithm.

[0062] Optionally, the processor may also execute program code for the following steps: determining the weights of each of the multiple prediction algorithms, including: using multiple prediction algorithms, based on historical work order data and historical environment data, predicting the number of future work orders at the first level of the target business within a preset test period, and obtaining the test prediction results corresponding to each of the multiple prediction algorithms; obtaining the actual number of work orders at the first level of the target business within the preset test period, as the actual collection result; calculating the sum of squared errors between the test prediction results and the actual collection results corresponding to each of the multiple prediction algorithms, as the prediction error results corresponding to each of the multiple prediction algorithms; calculating the reciprocal of the prediction error results corresponding to each of the multiple prediction algorithms; and determining the weights of each of the multiple prediction algorithms based on the reciprocals of the prediction error results corresponding to each of the multiple prediction algorithms.

[0063] Optionally, the processor may also execute program code that performs the following steps: based on the future work order quantity of each of the multiple levels, obtain the business demand forecast results for each of the multiple levels, including: obtaining the work order unit demand data of the target business within a preset historical period; calculating the product of the future work order quantity and the work order unit demand data for each of the multiple levels to obtain the initial demand forecast results for each of the multiple levels; calculating the optimized demand forecast difference for each of the multiple levels based on the initial demand forecast results for each of the multiple levels; and determining the business demand forecast results for each of the multiple levels based on the initial demand forecast results and the optimized demand forecast difference for each of the multiple levels.

[0064] This invention provides a method for predicting business demand data. It acquires historical work order data and historical environment data of the first level within a preset historical period for a target business. The historical work order data is a time series of the number of historical work orders, and the historical environment data is external environment data affecting the business demand of the target business. The first level is the highest level among multiple demand levels of the target business. Based on the historical work order data and historical environment data, it predicts the future number of work orders for each of the multiple levels of the target business within a preset future period, where the multiple levels include the first level. Based on the future number of work orders for each of the multiple levels, it obtains the business demand prediction results for each of the multiple levels, achieving the goal of multi-dimensional prediction of business demand data. This improves the accuracy and robustness of prediction, optimizes resource allocation efficiency, and solves the technical problem that current prediction methods rely on manual experience and subjective estimation, using simple statistical models or single algorithms, resulting in large prediction errors.

[0065] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a non-volatile storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0066] Embodiments of the present invention also provide a non-volatile storage medium. Optionally, in this embodiment, the aforementioned non-volatile storage medium can be used to store the program code executed by the business demand data prediction method provided in the above embodiments.

[0067] Optionally, in this embodiment, the non-volatile storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0068] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: obtaining historical work order data and historical environment data of the target service at the first level within a preset historical period, wherein the historical work order data is a time series of the number of historical work orders, the historical environment data is external environment data that affects the business requirements of the target service, and the first level is the highest level among multiple requirement levels of the target service; based on the historical work order data and historical environment data, predicting the future number of work orders for each of the multiple levels of the target service within a preset future period, wherein the multiple levels include the first level; and obtaining the business requirement prediction results for each of the multiple levels based on the future number of work orders for each of the multiple levels.

[0069] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: obtaining historical work order data of the first level of the target service within a preset historical period, including: obtaining initial work order data of the first level of the target service within a preset historical period; performing anomaly correction processing on the initial work order data to obtain historical work order data, wherein the anomaly correction processing includes replacement processing, deletion processing and filtering processing, the replacement processing represents replacing the abnormal data in the initial work order data with the average of the remaining normal data, the remaining normal data being the data in the initial work order data other than the abnormal data, the deletion processing represents deleting the duplicate data in the initial work order data, and the filtering processing represents performing white noise filtering processing on the data in the initial work order data whose fluctuation frequency exceeds a preset frequency threshold.

[0070] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: predicting the number of future work orders for the target business at multiple levels within a preset future period based on historical work order data and historical environment data, including: predicting the number of future work orders for the target business at the first level within a preset future period based on historical work order data and historical environment data using multiple preset prediction algorithms, obtaining a first prediction result corresponding to each of the multiple prediction algorithms; determining the weights corresponding to each of the multiple prediction algorithms; weighted summing the first prediction results corresponding to each of the multiple prediction algorithms based on their respective weights, obtaining the number of future work orders at the first level; predicting the number of future work orders for the target business at the second level within a preset future period based on the number of future work orders at the first level, historical work order data, and historical environment data using multiple prediction algorithms, obtaining a second prediction result corresponding to each of the multiple prediction algorithms, wherein the second level is the next level of business demand level below the first level; continuing to repeat the above steps, predicting the number of future work orders at the next level based on the number of future work orders at the previous level, until multiple levels are traversed to obtain the number of future work orders at each of the multiple levels.

[0071] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: multiple prediction algorithms including the Prophet prediction algorithm, the TAR threshold autoregressive algorithm, and the FFT fast Fourier transform algorithm.

[0072] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: determining the weights corresponding to each of the multiple prediction algorithms, including: using multiple prediction algorithms, based on historical work order data and historical environment data, predicting the number of future work orders of the target service at the first level within a preset test period, and obtaining the test prediction results corresponding to each of the multiple prediction algorithms; obtaining the actual number of work orders of the target service at the first level within the preset test period, as the actual collection result; calculating the sum of squared errors between the test prediction results and the actual collection results corresponding to each of the multiple prediction algorithms, as the prediction error results corresponding to each of the multiple prediction algorithms; calculating the reciprocal of the prediction error results corresponding to each of the multiple prediction algorithms; and determining the weights corresponding to each of the multiple prediction algorithms based on the reciprocals of the prediction error results corresponding to each of the multiple prediction algorithms.

[0073] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: obtaining business demand prediction results for multiple levels based on the future work order quantity for each level, including: acquiring work order unit demand data for the target business within a preset historical period; calculating the product of the future work order quantity and work order unit demand data for each level to obtain initial demand prediction results for each level; calculating optimized demand prediction differences for each level based on the initial demand prediction results for each level; and determining business demand prediction results for each level based on the initial demand prediction results and the optimized demand prediction differences for each level.

[0074] Embodiments of the present invention also provide a computer program product, including a computer program. Optionally, in this embodiment, when the computer program is executed by a processor, it can: acquire historical work order data and historical environment data of a target business at a first level within a preset historical period, wherein the historical work order data is a time series of the number of historical work orders, the historical environment data is external environment data that affects the business needs of the target business, and the first level is the highest level among multiple demand levels of the target business; based on the historical work order data and the historical environment data, predict the future number of work orders for each of the multiple levels of the target business within a preset future period, wherein the multiple levels include the first level; and based on the future number of work orders for each of the multiple levels, obtain the business demand prediction results for each of the multiple levels.

[0075] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0076] In the above embodiments of the present invention, 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.

[0077] 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.

[0078] 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.

[0079] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0080] 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 non-volatile storage medium. Based on this understanding, the technical solution of the present invention, 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 the present invention. 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.

[0081] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting business demand data, characterized in that, include: Obtain historical work order data and historical environment data of the first level of the target business within a preset historical period. The historical work order data is a time series of the number of historical work orders, and the historical environment data is external environment data that affects the business requirements of the target business. The first level is the highest level among the multiple requirement levels of the target business. Based on the historical work order data and the historical environment data, the number of future work orders for the target business at multiple levels within a preset future period is predicted respectively, wherein the multiple levels include the first level; Based on the number of future work orders for each of the multiple levels, the business demand prediction results for each of the multiple levels are obtained.

2. The method according to claim 1, characterized in that, The acquisition of historical work order data at the first level of the target business within a preset historical period includes: Obtain the initial work order data of the first level of the target service within the preset historical period; The initial work order data is subjected to anomaly correction processing to obtain the historical work order data. The anomaly correction processing includes replacement processing, deletion processing, and filtering processing. The replacement processing means replacing the abnormal data in the initial work order data with the average of the remaining normal data. The remaining normal data refers to the data in the initial work order data other than the abnormal data. The deletion processing means deleting the duplicate data in the initial work order data. The filtering processing means filtering the data in the initial work order data whose fluctuation frequency exceeds a preset frequency threshold using white noise.

3. The method according to claim 1, characterized in that, The step of predicting the future number of work orders for the target business at multiple levels within a preset future period based on the historical work order data and the historical environment data includes: Using multiple preset prediction algorithms, based on the historical work order data and the historical environment data, the number of future work orders at the first level of the target business within the preset future period is predicted, and the first prediction result corresponding to each of the multiple prediction algorithms is obtained. Determine the weights corresponding to each of the multiple prediction algorithms; Based on the weights corresponding to the multiple prediction algorithms, the first prediction results corresponding to the multiple prediction algorithms are weighted and summed to obtain the number of future work orders at the first level. Based on the number of future work orders at the first level, the historical work order data, and the historical environment data, the number of future work orders at the second level of the target business within the preset future period is predicted by the multiple prediction algorithms, and the second prediction result corresponding to each of the multiple prediction algorithms is obtained, wherein the second level is the next level of business demand level below the first level. Continue repeating the above steps, predicting the future work order quantity of the next level based on the future work order quantity of the previous level, until all levels have been traversed to obtain the future work order quantity of each level.

4. The method according to claim 3, characterized in that, The multiple prediction algorithms include the Prophet prediction algorithm, the TAR threshold autoregressive algorithm, and the FFT fast Fourier transform algorithm.

5. The method according to claim 3, characterized in that, Determining the weights corresponding to each of the plurality of prediction algorithms includes: Using the multiple prediction algorithms, based on the historical work order data and the historical environment data, the number of future work orders for the target business at the first level within a preset test period is predicted, and the test prediction results corresponding to each of the multiple prediction algorithms are obtained. The actual number of work orders at the first level of the target service within the preset test period is obtained as the actual collection result. The sum of squared errors between the test prediction results and the actual acquisition results corresponding to each of the multiple prediction algorithms is calculated and used as the prediction error results corresponding to each of the multiple prediction algorithms. Calculate the reciprocal of the prediction error result corresponding to each of the multiple prediction algorithms; The weights of each prediction algorithm are determined based on the reciprocals of their respective prediction error results.

6. The method according to any one of claims 1 to 5, characterized in that, The process of obtaining business demand forecast results for each of the multiple levels based on the future number of work orders for each level includes: Obtain the work order unit demand data of the target business within the preset historical period; Calculate the product of the future work order quantity and the work order unit demand data for each of the multiple levels to obtain the initial demand forecast results for each of the multiple levels; Based on the initial demand forecast results of each of the multiple levels, the optimized demand forecast difference of each of the multiple levels is calculated respectively; Based on the initial demand forecast results of each of the multiple levels and the difference between the optimized demand forecasts of each of the multiple levels, the business demand forecast results of each of the multiple levels are determined.

7. A business demand data prediction device, characterized in that, include: The acquisition module is used to acquire historical work order data and historical environment data of the first level of the target business within a preset historical period. The historical work order data is a time series of the number of historical work orders, and the historical environment data is external environment data that affects the business requirements of the target business. The first level is the highest level among the multiple requirement levels of the target business. The first prediction module is used to predict the number of future work orders for the target business at multiple levels within a preset future period based on the historical work order data and the historical environment data, wherein the multiple levels include the first level. The second prediction module is used to obtain the business demand prediction results for each of the multiple levels based on the number of future work orders for each level.

8. A non-volatile storage medium, characterized in that, The non-volatile storage medium includes a stored program, wherein, when the program is executed, it controls the device where the non-volatile storage medium is located to execute the business demand data prediction method according to any one of claims 1 to 6.

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

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

Images