A power grid supply chain network distribution protocol inventory material demand prediction method based on a time series model

Abstract

The application discloses a power grid supply chain distribution network protocol inventory material demand prediction method based on a time series model, and belongs to the field of data processing. The method comprises the following steps: data collection and preprocessing; model dimension determination; model prediction; and model correction. The method introduces a time series model into the prediction of power grid supply chain demand, establishes a time series model suitable for the power grid supply chain distribution network protocol inventory material demand, mines the correlation law between a project and demand prediction by means of a big data analysis method and an artificial intelligence algorithm, forms a material demand characteristic portrait by using a data analysis method, and comprehensively and multi-angulary combs, analyzes, diagnoses and evaluates the material characteristic portrait through historical data cleaning, model establishment and event management, so as to improve the operation capacity of an algorithm model library and the data applicability, optimize the evaluation material demand prediction mode, and provide support for power grid supply chain material procurement and supply decision-making. The method can be widely used in the field of power grid supply chain distribution network protocol inventory material demand prediction and management.

Description

Technical Field

[0001] This invention belongs to the field of data processing, and in particular relates to a method for forecasting the demand of inventory materials for power grid supply chain distribution network agreements. Background Technology

[0002] Materials management is an important part of enterprise management. It involves the planning, organization, and control of the ordering, transportation, storage, and supply of various materials required in the production process.

[0003] Good material management is conducive to the rational use and conservation of materials, improving product quality, reducing production costs, accelerating capital turnover, and increasing corporate profits.

[0004] Power grid material planning management, as the front end of supply chain management, is the foundation for ensuring material supply.

[0005] Power grid material planning management is an important part of power grid enterprise management, and its management level directly affects the enterprise's efficiency and long-term development.

[0006] With the rapid development of the power industry and the advancement of smart grid construction, the management of the power grid supply chain has become increasingly important. Power grid material planning (also known as distribution network contract inventory or contract inventory), as a crucial component of the power grid supply chain, directly impacts the stable operation and cost control of the power grid through accurate demand forecasting.

[0007] Contractual inventory demand forecasting management involves predicting the scale of major materials and services required by the company in specific professional areas such as safe production, power grid development, and marketing services over a certain period in the future. Its purpose is to fully utilize historical data, combined with factors such as the company's annual budget control targets and project plans, and apply scientific and reasonable methods to derive realistic demand data, providing a basis for decision-making in annual work arrangements.

[0008] In recent years, although power companies have conducted preliminary research and practical work on annual demand forecasting for contracted inventory materials, some shortcomings and deficiencies still exist, and there is still a gap between their current achievements and the goals. Specifically, the forecasting accuracy of distribution network contracted inventory materials is not high enough, and the research on forecasting models and influencing factors needs further in-depth study. Due to the lack of big data analysis methods and scientific forecasting models, there is a significant deviation between demand plans and actual needs, resulting in excessively high or low contract matching execution rates in the subsequent supply phase, which to some extent affects the supply guarantee of power materials.

[0009] Patent application CN106600032A, published on April 26, 2017, discloses a "method and apparatus for forecasting inventory demand." The method includes: obtaining the planned investment amount for each project and the corresponding engineering attributes based on an initial investment plan; obtaining the material subcategories included in each project based on the engineering attributes; searching a pre-defined list of correspondences between material subcategories, engineering attributes, and prediction models to obtain a prediction model corresponding to each material subcategory; inputting the planned investment amount into the prediction model corresponding to each material subcategory to calculate the predicted demand for each material subcategory in each project; and summing the predicted demand for each material subcategory across different projects to calculate the total predicted demand for each material subcategory. This inventory demand forecasting method and apparatus improves the efficiency and accuracy of inventory forecasting by pre-constructing corresponding prediction models for different material subcategories under each project. However, since it only pre-builds corresponding prediction models for different material subcategories under each project, the mathematical modeling workload is large in actual implementation. The accuracy of the prediction model needs a verification process. Moreover, once external conditions (such as changes in the supply price of a single material or changes in the supply channel of a certain material) change, the previous modeling and prediction models need to be started from scratch. The entire project implementation process takes a long time from start to finish, which affects the user experience and effectiveness of actual users.

[0010] The invention patent application CN111882262A, published on November 3, 2020, discloses "A Method for Accurately Predicting the Demand of Contractual Inventory Materials," which includes the following steps: Step 1: Obtaining the total investment amount of the production plan based on the production plan; Step 2: Compiling inventory records, detailing the remaining inventory reserves in the warehouse according to the items required by the production plan; Step 3: Listing the items required by the production plan and calculating the quantity of materials required for each item; Step 4: Subtracting the inventory quantity from the required items and calculating and summing the funds required for each item. This method helps reduce cost input and saves resources by calculating inventory quantities; it inputs all data into a computer program and uses a clustering algorithm for calculation, saving the results to an Elasticsearch database (ES). Simultaneously, it uses market price data for material procurement to accurately predict the actual demand for materials, avoiding significant discrepancies between actual and planned funding. However, this technical solution is based on inventory statistics. If the inventory data cannot be updated in a timely manner or is temporarily unavailable, the data source and implementation basis of the solution will be affected. Furthermore, it uses a "clustering" algorithm for calculation and saves the calculation results to ES (Elasticsearch, a distributed document non-relational database built on Apache Lucene, designed to handle large amounts of data and support complex data analysis; it can be widely used in full-text search, real-time analysis and other scenarios, and is particularly good at handling unstructured or semi-structured data). This places high demands on computer hardware and computing tools, and the computing power of the algorithm model library and the applicability of the database are somewhat affected.

[0011] Given current technological conditions and management levels, how to leverage modern big data analytics and artificial intelligence algorithms to uncover the correlation between projects and demand forecasting, optimize demand forecasting models using data analysis methods, and provide support for material procurement and supply decisions is a pressing technical problem that needs to be solved in actual power grid material planning and management. Summary of the Invention

[0012] The technical problem this invention aims to solve is to provide a method for forecasting the demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model. By leveraging modern big data analytics and artificial intelligence algorithms, it uncovers the correlation between projects and demand forecasting, uses data analysis methods to create a profile of material demand characteristics, and comprehensively analyzes, diagnoses, and evaluates these profiles from multiple perspectives through historical data cleaning, model building, and event management. This optimizes the demand forecasting model and provides support for material procurement and supply decisions.

[0013] The technical solution of this invention is: to provide a method for forecasting the demand for distribution network contract inventory materials in the power grid supply chain based on a time series model, including the forecasting and management of contract inventory demand, characterized in that the forecasting and management of contract inventory demand are carried out according to the following steps:

[0014] 1) Basic data collection and data preprocessing:

[0015] Basic data collection includes system interface data and protocol preparation data;

[0016] The system interface data should include at least: assessment material master data, project unit master data, project master data, warehouse master data, outbound data, and framework agreement bidding status;

[0017] The agreement preparation data should include at least the master data of the bidding package, the procurement catalog, the materials in the bidding package, the comparison between new and old materials, the project investment plan, and the rolling forecast comparison between new and old materials;

[0018] After the basic data collection is completed, data cleaning is performed, including operations such as removing negative values ​​and outliers, to complete the preprocessing of the basic data.

[0019] 2) Determine the dimensions:

[0020] Discover the correlation between projects and demand forecasts, use data analysis methods to form a profile of material demand characteristics, and use the outbound data after data cleaning and preprocessing as the package dimension, and the monthly summary quantity as the parameter to conduct profile analysis to evaluate the material characteristics.

[0021] 3) Model building:

[0022] By introducing time series models into the forecasting of electricity supply chain demand, and based on the profile analysis results of the assessed material characteristics, three types of model combinations are adopted: linear and nonlinear seasonal combination model, trend and seasonal autoregressive combination model, and discontinuous and seasonal combination model. Based on the analysis and calculation of electricity data, the models in the model library are selected accordingly, and the most suitable model combination and related parameters are determined through parameter tuning, thus completing the intelligent selection of material forecasting models.

[0023] 4) Model prediction:

[0024] The characteristics of the assessed materials are analyzed, and their characteristics are defined based on seasonality, volatility, continuity, and importance. A model is selected based on the characteristics of the assessed materials. Finally, the predicted values ​​of the assessed materials' standard package dimensions are output through the model.

[0025] 5) Model correction:

[0026] By comparing and analyzing the forecast results of material demand with the actual outbound data, the forecast deviation is statistically calculated and the accuracy is analyzed. Based on the correspondence between the forecast deviation index and the confidence level, the demand forecast results are applied to provide technical support for the procurement and supply decisions of the power grid supply chain.

[0027] Specifically, the method for forecasting the demand for distribution network agreement inventory materials in the power grid supply chain based on time series models introduces time series models into the forecasting of power grid supply chain demand, broadens the traditional time series algorithm modeling, and establishes a time series model suitable for the demand of distribution network agreement inventory materials in the power grid supply chain. By leveraging modern big data analysis methods and artificial intelligence algorithms, it mines the correlation between projects and demand forecasting, uses data analysis methods to form a profile of material demand characteristics, and comprehensively and from multiple perspectives sorts out, analyzes, diagnoses, and evaluates the material characteristic profile through historical data cleaning, model building, and event management. This improves the computing power and data applicability of the algorithm model library, optimizes the evaluation of material demand forecasting models, and provides support for power grid supply chain material procurement and supply decisions.

[0028] Furthermore, the aforementioned assessment of material property profile analysis includes at least continuity analysis, fluctuation analysis, and seasonality analysis.

[0029] Furthermore, the prediction deviation is calculated as follows: (predicted outbound amount - actual outbound amount) / actual outbound amount.

[0030] Furthermore, based on the correspondence between prediction deviation indicators and confidence levels, and according to the prediction deviation results, the confidence levels of various materials are divided into four levels:

[0031] a. If the prediction error is less than or equal to 20%, the confidence level is I;

[0032] b. If 20% < prediction deviation <= 40%, the confidence level is II;

[0033] c. If 40% < prediction bias <= 80%, the confidence level is III;

[0034] d. Prediction bias > 80%, confidence level IV.

[0035] Specifically, for materials with confidence level I, the procurement plan for contract inventory is directly formulated by combining the batch demand coverage and the remaining framework agreement share, using the demand forecasting model to calculate the results and converting them into quantity and price.

[0036] For materials with confidence levels II and III, a joint agreement inventory procurement plan is formulated by combining the batch demand coverage and the remaining framework agreement share with the demand forecasting model calculation results and demand submissions, after quantity-price conversion.

[0037] For materials with a credibility level of IV, a manual reporting method is used directly, and a contract inventory procurement plan is formulated after converting quantity and price.

[0038] Specifically, the method of determining the most suitable model combination and related parameters for power data through parameter tuning includes at least the following:

[0039] 1) When selecting a model, consider indicators including goodness of fit and white noise test of residuals;

[0040] 2) Employ methods including least squares and maximum likelihood estimation to estimate model parameters using historical data;

[0041] 3) Employ testing methods including white noise test of residuals, cross-validation, and goodness-of-fit test of the model to statistically test the parameter estimation results and verify the accuracy and effectiveness of the model.

[0042] 4) Adjust and optimize the model based on the test results to improve the model's prediction accuracy and stability.

[0043] Furthermore, the preprocessing of the basic data includes:

[0044] 1) Data processing includes negative value removal and outlier handling;

[0045] Negative value removal: Determine if the current month is negative. If it is, assign 0 to the current month and sum it with the previous month to get the previous month's outbound value. If the sum is still negative, assign 0. Continue this process backwards until a non-negative value appears.

[0046] Outlier handling: Based on the monthly aggregated quantity of outbound data at the package level, the average of the outbound quantity of the month plus 3 times the standard deviation X is used as the boundary. Data exceeding X is considered outlier data, and an outlier alert is issued for business verification. If the outlier is determined to be system outlier data, it is fed back to the ERP system for adjustment and data resynchronization.

[0047] 2) Characteristic analysis and classification;

[0048] Using the monthly aggregated quantity of outbound data by package dimension after data cleaning and preprocessing as parameters, we conduct material characteristic profiling analysis;

[0049] Material property profiling analysis includes at least continuity analysis, fluctuation analysis, and seasonality analysis;

[0050] Seasonality and importance are categorized as high, medium, and low; continuity and volatility are categorized as high, relatively high, medium, relatively low, and low.

[0051] 3) Model building and correction;

[0052] For demand forecasting models, a combination of top-down forecasting logic and bottom-up forecasting model is used to construct and refine the models.

[0053] The aforementioned "top-down" forecasting logic includes using project investment plans and historical project material outbound data to analyze the correlation between project type, voltage level and major equipment and materials, and to sort out the corresponding patterns between project type and material demand; and comprehensively considering factors such as annual growth rate and project approval rate to decompose and derive the expected usage quantity and amount of each material sub-category.

[0054] The aforementioned "bottom-up" forecasting model utilizes data including historical material outbound data, standard packages, and comparison tables of new and old materials. Based on the analysis of material volatility, seasonality, importance, continuity, and project relevance, it employs multiple forecasting models capable of simultaneously handling continuous and intermittent demand for combined forecasting, thereby automating and intelligently selecting forecasting models. It uses simulation "if-else" technology, combined with demand forecasting results, to establish a library of intelligent material allocation strategies for different business scenarios, forming a "scenario-resource-strategy" operational model. Through an automatic optimization algorithm, it intelligently assigns weights to the forecasting results of different models, effectively improving the accuracy of the demand forecasting model.

[0055] The power grid supply chain distribution network protocol inventory material demand forecasting method based on time series model described in this invention adopts the simulation "if-else" technology, establishes a material intelligent allocation strategy library for different business scenarios, forms a "scenario-resource-strategy" operation model, and intelligently assigns weights to the prediction results of different prediction models through an automatic optimization algorithm, effectively improving the accuracy of the demand forecasting model.

[0056] The present invention provides a method for forecasting the demand of power grid supply chain distribution network protocol inventory materials based on time series models. In terms of data modeling and algorithm engineering, the method introduces time series models into the forecasting of power grid supply chain demand, broadens the traditional time series algorithm modeling, incorporates more efficient models that can automatically optimize parameters, and uses a combination of algorithm model libraries to improve the forecasting accuracy, thereby enhancing the computing power and data applicability of the algorithm model library.

[0057] At the system operation level, taking into account the remaining share of the contracted inventory, the remaining inventory in transit and in stock, a "one-stop" online operation mode for submitting bidding and procurement plans and supply plans is formed. The decision-making and instructions are deployed and connected in the "cloud" to realize online decision support, and improve the front-end and back-end collaboration capabilities by monitoring the supply and demand of materials in real time.

[0058] Simulation of materials dynamically improves resource optimization. By using simulation "if-else" technology and combining demand forecasting results, a library of intelligent material allocation strategies is established for different business scenarios, forming a "scenario-resource-strategy" operation model to adapt to the optimal material strategy, thereby achieving intelligent overall planning of material allocation and further improving the level of material services.

[0059] Compared with the prior art, the advantages of the present invention are:

[0060] 1. The technical solution of this invention, based on existing research on material demand forecasting projects, significantly improves the accuracy of demand forecasting, mainly in terms of data modeling and algorithm engineering. It broadens the traditional time series algorithm modeling by incorporating more efficient models that can automatically optimize parameters, including Prophet, Auto-Arima, TBATS, etc., effectively improving the computing power and data applicability of the algorithm model library, and thus improving the prediction accuracy of the combined application of the algorithm model library.

[0061] 2. The technical solution of this invention improves the front-end and back-end collaboration capabilities by monitoring the supply and demand of materials in real time. At the system operation level, it comprehensively considers the remaining share of the agreement inventory, the remaining in-transit and in-stock inventory, and forms a "one-stop" online operation mode for submitting bidding and procurement plans and supply plans. It connects decision-making and instructions through "cloud" deployment and realizes online decision support.

[0062] 3. The technical solution of this invention improves the level of resource optimization by simulating the dynamics of materials; it uses the simulation "if-else" technology, combined with the results of demand forecasting, to establish a library of intelligent material allocation strategies for different business scenarios, forming a "scenario-resource-strategy" operation model, adapting to the optimal material strategy, thereby realizing intelligent overall planning of material allocation and further improving the level of material services. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the overall path of the prediction method of the present invention;

[0064] Figure 2 This is a schematic diagram of the "top-down" prediction logic in this invention;

[0065] Figure 3 This is a schematic diagram of the "bottom-up" prediction model in this invention. Detailed Implementation

[0066] The invention will now be further described with reference to the accompanying drawings.

[0067] Time series data is a collection of data points arranged chronologically, used to predict future values. Time series models predict future demand trends based on past observations, helping businesses make more accurate demand forecasts and plans by analyzing seasonality, trends, and periodicity in historical data.

[0068] In supply chain demand forecasting, commonly used time series models include moving average models, exponential smoothing models, autoregressive moving average models (ARMA), autoregressive integral moving average models (ARIMA), and seasonal autoregressive integral moving average models (SARIMA).

[0069] The moving average model is a method for predicting future demand by using the average of data from several past periods. It can smooth out outliers, reduce prediction errors, and is suitable for situations where demand changes relatively smoothly. However, the moving average model is less effective at predicting demand when it exhibits significant fluctuations or seasonal variations.

[0070] Exponential smoothing is a method that smooths past observations by assigning different weights. By gradually decreasing the weights, it focuses more on recent observations and can better adapt to changing demand trends. Exponential smoothing is suitable for situations where the demand curve has a linear trend.

[0071] The Autoregressive Moving Average (ARMA) model is a time-series-based stochastic process model that combines the characteristics of autoregression and moving average. ARMA models can capture the autocorrelation of time series and the correlation of moving averages, making them effective at modeling multivariate time series.

[0072] The Autoregressive Moving Average (ARIMA) model is an extension of the ARMA model, introducing differencing operations to handle non-stationary time series. ARIMA models can handle trends in time series and also account for seasonality. For situations with significant seasonal variations in demand, ARIMA models can provide relatively accurate forecasts.

[0073] The Seasonal Autoregressive Integral Moving Average (SARIMA) model is a further extension of the ARIMA model, specifically designed for handling seasonal time series. SARIMA models can account for seasonal trends and cyclical variations, providing more accurate forecasts when demand exhibits significant seasonality.

[0074] In addition to the models mentioned above, there are other time series models that can be used to forecast supply chain demand, such as the Seasonal Exponential Smoothing (SES) model and the Grey Model (GM). These models have different characteristics and applicable scenarios, and companies can choose the appropriate model for forecasting based on their own needs and data characteristics.

[0075] The technical solution of this invention utilizes modern big data analysis methods and artificial intelligence algorithms to uncover the correlation between projects and demand forecasting. It uses data analysis methods to create a profile of material demand characteristics, and through historical data cleaning, model building, and event management, it comprehensively and from multiple perspectives sorts out, analyzes, diagnoses, and evaluates the material characteristic profile, optimizes the demand forecasting model, and provides support for material procurement and supply decisions.

[0076] The technical solution of this invention introduces time series models into the forecasting of electricity supply chain demand, broadening the scope of traditional time series algorithm modeling by incorporating more efficient models with automatic parameter optimization capabilities, including Prophet, Auto-Arima, and TBATS. This effectively enhances the computational power and data applicability of the algorithm model library. Through the mining and analysis of historical data, a time series model suitable for the demand of distribution network agreement inventory materials in the power grid supply chain is established, and the model is used for demand forecasting. This method has high forecasting accuracy and stability, and can provide strong support for power grid material management.

[0077] like Figure 1 As shown, the overall method path of the technical solution of the present invention is as follows:

[0078] (1) Data processing:

[0079] Data collection is divided into two types: system integration data and agreement preparation data. System integration data mainly includes: material master data, project unit master data, project master data, warehouse master data, outbound data, and framework agreement bidding status. Agreement preparation data includes bid package master data, procurement catalog, bid package materials, comparison of new and old materials, project investment plan, and rolling forecast comparison of new and old materials. Simultaneously, data cleaning is performed, removing negative values ​​and outliers, completing the preprocessing of the basic data.

[0080] (2) Determine the dimensions:

[0081] Using the monthly aggregated (quantity) dimensions of the outbound data after data cleaning and preprocessing as parameters, material characteristic profiling analysis is performed. This analysis mainly includes continuity analysis, fluctuation analysis, and seasonality analysis. Based on the results of the material characteristic profiling analysis, an intelligent material forecasting model is selected.

[0082] (3) Model building:

[0083] The basic data are all time series data. For bottom-up forecasting, three types of model combinations are used: linear and nonlinear seasonal combination model, trend and seasonal autoregressive combination model, and discontinuous and seasonal combination model. Based on the analysis and calculation of power data, the models in the model library will be selected accordingly, and the most suitable model combination and related parameters will be determined through parameter tuning.

[0084] (4) Model prediction:

[0085] The characteristics of the materials are analyzed, and their characteristics are defined based on seasonality, volatility, continuity, and importance. Then, a model is selected based on these characteristics. Finally, the model outputs predicted values ​​for the standard package dimension.

[0086] (5) Model correction:

[0087] By comparing and analyzing demand forecast results with actual outbound data, the forecast deviation is statistically calculated to analyze accuracy. Based on the correspondence between forecast deviation indicators and confidence levels, the demand forecast results are applied practically.

[0088] Specifically, the above forecast deviation = (forecasted outbound amount - actual outbound amount) / actual outbound amount;

[0089] During implementation, based on the correspondence between prediction deviation indicators and confidence levels, and according to the prediction deviation results, the confidence levels of various materials can be divided into four categories:

[0090] a. Prediction deviation <= 20%, confidence level I:

[0091] For materials with a confidence level of 1, the contract inventory procurement plan is directly formulated by combining the batch demand coverage and the remaining framework agreement share, using the demand forecasting model to calculate the results and converting the quantity and price.

[0092] b. 20% < prediction deviation <= 40%, confidence level II;

[0093] 40% < prediction bias <= 80%, confidence level III.

[0094] For materials with confidence levels of 2 and 3, the procurement plan for contract inventory will be jointly formulated by combining the batch demand coverage and the remaining framework agreement share with the demand forecasting model calculation results and demand submissions, after quantity-price conversion.

[0095] c, Level IV: Prediction bias > 80%;

[0096] For materials with a credibility level of 4, a manual reporting method is adopted, and a contract inventory procurement plan is formulated after converting quantity and price.

[0097] Furthermore, in this technical solution, the data processing (also known as the implementation means) includes the following steps:

[0098] (1) Data processing:

[0099] Negative value removal: Determine if the current month is negative. If it is, assign 0 to the current month and sum it with the previous month to get the previous month's outbound value. If the sum is still negative, assign 0. Continue this process until a non-negative value appears.

[0100] Outlier Handling: Based on the monthly summary (quantity) of outbound data by package dimension, the average outbound quantity for the current month plus 3 standard deviations (X) is used as the boundary. Data exceeding X is considered outlier, and an outlier alert is issued for business verification. Outliers are assessed; if determined to be system anomalies, feedback is provided to the ERP system for adjustment and data resynchronization.

[0101] (2) Characteristic analysis and classification:

[0102] Using the monthly summary (quantity) of outbound data after data cleaning and preprocessing as parameters, a material characteristic profile analysis is performed. The material characteristic profile analysis mainly includes continuity analysis, volatility analysis, and seasonality analysis. Seasonality and importance are respectively divided into high, medium, and low, while continuity and volatility are divided into high, relatively high, medium, relatively low, and low.

[0103] (3) Model building correction:

[0104] like Figure 2 As shown in the figure, in this technical solution, the demand forecasting model adopts a combination of "top-down" forecasting logic and "bottom-up" forecasting model.

[0105] a. Top-down forecasting logic:

[0106] The main approach involves using project investment plans and historical project material outflow data to analyze the correlation between project type, voltage level, and major equipment and materials, and to identify the corresponding patterns between project type and material demand. It also takes into account factors such as annual growth rate and project approval rate to break down the expected usage quantity and amount for each material subcategory.

[0107] b. Bottom-up prediction model:

[0108] The bottom-up forecasting model in this technical solution mainly utilizes historical material outbound data, standard packages, and comparison tables of new and old materials. Based on the analysis of material characteristics such as volatility, seasonality, importance, continuity, and project relevance, it employs multiple forecasting models capable of simultaneously handling continuous and intermittent demand for combined forecasting, thereby automating and intelligently selecting forecasting models. It uses simulation "if-else" technology, combined with demand forecasting results, to establish an intelligent material allocation strategy library for different business scenarios, forming a "scenario-resource-strategy" operational model. Through an automatic optimization algorithm, it intelligently assigns weights to the forecasting results of different models, effectively improving the accuracy of the demand forecasting model.

[0109] Furthermore, a brief introduction to the "if-else technique" (also known as the if-else structure) in this technical solution is as follows:

[0110] The if-else structure is a flow control statement in Python that determines the next action of the program based on the truth value of a condition. Its basic syntax is as follows:

[0111] ```

[0112] if condition:

[0113] Code Block 1

[0114] else:

[0115] Code Block 2

[0116] ```

[0117] The condition is an expression that evaluates to either True or False. If the condition is True, code block 1 will be executed; otherwise, code block 2 will be executed.

[0118] In Python, indentation is used to represent code blocks, so code block 1 and code block 2 must have the same indentation.

[0119] Below is a simple example demonstrating how to use an if-else structure to determine the parity of a number based on user input:

[0120] ```

[0121] num = int(input("Please enter a number:"))

[0122] if num%2==0:

[0123] print("This is an even number.")

[0124] else:

[0125] print("This is an odd number.")

[0126] ```

[0127] In this example, the user enters a number. If the result of the number modulo 2 is 0, code block 1 will be executed (i.e., the entered number is even); otherwise, code block 2 will be executed (i.e., the entered number is odd).

[0128] Compared to the control expressions and multi-branch expressions in the case statement, the conditional expressions in the if_else structure are more intuitive.

[0129] like Figure 3 As shown below, several examples are given of how to use simulation "if-else" technology to establish a resource intelligent allocation strategy library for different business scenarios, forming a "scenario-resource-strategy" operation model. Through automatic optimization algorithms, the prediction results of different prediction models are intelligently weighted, effectively improving the accuracy of demand prediction models:

[0130] 1. If seasonality = medium to high; then model selection = a combination of linear and nonlinear seasonal models;

[0131] 2. If continuity = none; volatility = medium to high; project specificity = low to medium; then model selection = a combination of discontinuous and seasonal models;

[0132] 3. If seasonality = none; volatility = medium; project specialization = high; importance = low; then model selection = autoregressive combination model of trend and seasonality;

[0133] 4.……

[0134] Since flow control in Python programming is an existing technology, its specific implementation steps and processes will not be detailed here.

[0135] The advantages of this technical solution are:

[0136] 1) Improve prediction accuracy by combining algorithm model libraries:

[0137] Based on existing research on material demand forecasting projects, the accuracy of demand forecasting has been significantly improved, mainly in terms of data modeling and algorithm engineering. The traditional time series algorithm modeling has been broadened, and more efficient models with automatic parameter optimization capabilities have been incorporated, including Prophet, Auto-Arima, and TBATS, which effectively enhances the computing power and data applicability of the algorithm model library.

[0138] 2) Real-time monitoring of material supply and demand enhances front-end and back-end collaboration capabilities:

[0139] At the system operation level, taking into account the remaining share of the contracted inventory, the remaining inventory in transit and in stock, a "one-stop" online operation mode for submitting bidding and procurement plans and supply plans is formed, and the decision-making and instructions are deployed and connected in the "cloud" to realize online decision support.

[0140] 3) Simulate dynamic improvement of material resources to enhance resource optimization:

[0141] By employing the "if-else" simulation technology and combining it with demand forecasting results, a library of intelligent material allocation strategies is established for different business scenarios. This forms a computational model of "scenario-resource-strategy," which adapts to the optimal material strategy, thereby achieving intelligent coordination of material allocation and further improving the level of material services.

[0142] This invention can be widely used in the field of power grid supply chain distribution network agreement inventory material demand forecasting and decision management.

Claims

1. A method for forecasting demand for distribution network contract inventory materials in a power grid supply chain based on a time series model, including forecasting and managing contract inventory demand, characterized by: The forecasting and management of contract inventory demand are carried out according to the following steps: 1) Basic data collection and data preprocessing: Basic data collection includes system interface data and protocol preparation data; The system interface data should include at least: assessment material master data, project unit master data, project master data, warehouse master data, outbound data, and framework agreement bidding status; The agreement preparation data should include at least the master data of the bidding package, the procurement catalog, the materials in the bidding package, the comparison between new and old materials, the project investment plan, and the rolling forecast comparison between new and old materials; After the basic data collection is completed, data cleaning is performed, including operations such as removing negative values ​​and outliers, to complete the preprocessing of the basic data. 2) Determine the dimensions: Discover the correlation between projects and demand forecasts, use data analysis methods to form a profile of material demand characteristics, and use the outbound data after data cleaning and preprocessing as the package dimension, and the monthly summary quantity as the parameter to conduct profile analysis to evaluate the material characteristics. 3) Model building: By introducing time series models into the forecasting of electricity supply chain demand, and based on the profile analysis results of the assessed material characteristics, three types of model combinations are adopted: linear and nonlinear seasonal combination model, trend and seasonal autoregressive combination model, and discontinuous and seasonal combination model. Based on the analysis and calculation of electricity data, the models in the model library are selected accordingly, and the most suitable model combination and related parameters are determined through parameter tuning, thus completing the intelligent selection of material forecasting models. 4) Model prediction: The characteristics of the assessed materials are analyzed, and their characteristics are defined based on seasonality, volatility, continuity, and importance. A model is selected based on the characteristics of the assessed materials. Finally, the predicted values ​​of the assessed materials' standard package dimensions are output through the model. 5) Model correction: By comparing and analyzing the forecast results of material demand with the actual outbound data, the forecast deviation is statistically calculated and the accuracy is analyzed. Based on the correspondence between the forecast deviation index and the confidence level, the demand forecast results are applied to provide technical support for the procurement and supply decisions of the power grid supply chain.

2. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 1, characterized in that: The proposed time-series model-based method for forecasting the demand of distribution network agreement inventory materials in the power grid supply chain introduces time-series models into the forecasting of power grid supply chain demand, broadening the traditional time-series algorithm modeling and establishing a time-series model suitable for the demand of distribution network agreement inventory materials in the power grid supply chain. By leveraging modern big data analysis methods and artificial intelligence algorithms, it uncovers the correlation between projects and demand forecasting, uses data analysis methods to form a profile of material demand characteristics, and comprehensively and from multiple perspectives, analyzes, diagnoses, and evaluates the material characteristic profile through historical data cleaning, model building, and event management. This enhances the computational power and data applicability of the algorithm model library, optimizes the evaluation of material demand forecasting models, and provides support for power grid supply chain material procurement and supply decisions.

3. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 1, characterized in that: The aforementioned assessment of material properties profile analysis includes at least continuity analysis, fluctuation analysis, and seasonality analysis.

4. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 1, characterized in that: The prediction deviation is calculated as follows: (predicted outbound amount - actual outbound amount) / actual outbound amount.

5. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 4, characterized in that: Based on the correspondence between prediction deviation indicators and confidence levels, and according to the prediction deviation results, the confidence levels of various materials are divided into four levels: a. If the prediction error is less than or equal to 20%, the confidence level is I; b. If 20% < prediction deviation <= 40%, the confidence level is II; c. If 40% < prediction bias <= 80%, the confidence level is III; d. Prediction bias > 80%, confidence level IV.

6. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 5, characterized in that, For materials with confidence level I, the contract inventory procurement plan is directly formulated by combining the batch demand coverage and the remaining framework agreement share, using the demand forecasting model to calculate the results and converting the quantity and price. For materials with confidence levels II and III, a joint agreement inventory procurement plan is formulated by combining the batch demand coverage and the remaining framework agreement share with the demand forecasting model calculation results and demand submissions, after quantity-price conversion. For materials with a credibility level of IV, a manual reporting method is used directly, and a contract inventory procurement plan is formulated after converting quantity and price.

7. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 1, characterized in that: The method of determining the most suitable model combination and related parameters for power data through parameter tuning includes at least the following: 1) When selecting a model, consider indicators including goodness of fit and white noise test of residuals; 2) Employ methods including least squares and maximum likelihood estimation to estimate model parameters using historical data; 3) Employ testing methods including white noise test of residuals, cross-validation, and goodness-of-fit test of the model to statistically test the parameter estimation results and verify the accuracy and effectiveness of the model. 4) Adjust and optimize the model based on the test results to improve the model's prediction accuracy and stability.

8. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 1, characterized in that: The preprocessing of the basic data includes: 1) Data processing includes negative value removal and outlier handling; Negative value removal: Determine if the current month is negative. If it is, assign 0 to the current month and sum it with the previous month to get the previous month's outbound value. If the sum is still negative, assign 0. Continue this process backwards until a non-negative value appears. Outlier handling: Based on the monthly aggregated quantity of outbound data at the package level, the average of the outbound quantity of the month plus 3 times the standard deviation X is used as the boundary. Data exceeding X is considered outlier data, and an outlier alert is issued for business verification. If the outlier is determined to be system outlier data, it is fed back to the ERP system for adjustment and data resynchronization. 2) Characteristic analysis and classification; Using the monthly aggregated quantity of outbound data by package dimension after data cleaning and preprocessing as parameters, we conduct material characteristic profiling analysis; Material property profiling analysis includes at least continuity analysis, fluctuation analysis, and seasonality analysis; Seasonality and importance are categorized as high, medium, and low; continuity and volatility are categorized as high, relatively high, medium, relatively low, and low. 3) Model building and correction; For demand forecasting models, a combination of "top-down" forecasting logic and "bottom-up" forecasting model is used to construct and modify the models. The aforementioned "top-down" forecasting logic includes using project investment plans and historical project material outbound data to analyze the correlation between project type, voltage level and major equipment and materials, and to sort out the corresponding patterns between project type and material demand; and comprehensively considering factors such as annual growth rate and project approval rate to decompose and derive the expected usage quantity and amount of each material sub-category. The aforementioned "bottom-up" forecasting model utilizes data including historical material outbound data, standard packages, and comparison tables of new and old materials. Based on the analysis of material volatility, seasonality, importance, continuity, and project relevance, it employs multiple forecasting models capable of simultaneously handling continuous and intermittent demand for combined forecasting, thereby automating and intelligently selecting forecasting models. It uses simulation "if-else" technology, combined with demand forecasting results, to establish a library of intelligent material allocation strategies for different business scenarios, forming a "scenario-resource-strategy" operational model. Through an automatic optimization algorithm, it intelligently assigns weights to the forecasting results of different models, effectively improving the accuracy of the demand forecasting model.

9. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 1, characterized in that: The proposed time-series model-based power grid supply chain distribution network protocol inventory material demand forecasting method adopts simulation "if-else" technology, establishes a material intelligent allocation strategy library for different business scenarios, forms a "scenario-resource-strategy" operation model, and intelligently assigns weights to the prediction results of different prediction models through an automatic optimization algorithm, effectively improving the accuracy of the demand forecasting model.

10. The method for forecasting demand for distribution network protocol inventory materials in the power grid supply chain based on a time series model as described in claim 1, characterized in that: The proposed method for forecasting the demand for distribution network protocol inventory materials in the power grid supply chain based on time series models introduces time series models into the forecasting of power grid supply chain demand in terms of data modeling and algorithm engineering. This broadens the traditional time series algorithm modeling, incorporates more efficient models that can automatically optimize parameters, and uses a combination of algorithm model libraries to improve forecasting accuracy, thereby enhancing the computing power and data applicability of the algorithm model library. The power grid supply chain distribution network agreement inventory demand forecasting method based on time series model, at the system operation level, comprehensively considers the remaining share of agreement inventory, remaining in transit and in stock inventory, forming a "one-stop" online operation mode for bidding and procurement plan submission and supply plan, and connects decision-making and instructions through "cloud" deployment, realizes online decision support, and improves front-end and back-end collaboration capabilities by monitoring material supply and demand in real time. The proposed time-series model-based power grid supply chain distribution network protocol inventory material demand forecasting method simulates the dynamic improvement of material resource optimization. By using the simulation "if-else" technology and combining the demand forecasting results, it establishes a material intelligent allocation strategy library for different business scenarios, forming a "scenario-resource-strategy" operation model to adapt to the optimal material strategy, thereby realizing intelligent overall planning of material allocation and further improving the material service level.

Images