Analysis method and system based on tobacco flow attraction model

Abstract

The invention relates to the technical field of data processing and analysis, in particular to an analysis method and system based on a tobacco flowing attraction model. The method comprises the following steps: acquiring sales data, production data, monthly return data and geographical assistance data of a target tobacco terminal; inputting the sales data and the production data into a target flow gravitation model to obtain a terminal gravitation value; performing anomaly detection on the terminal gravitational value to obtain a target abnormal gravitational value; predicting the monthly flow back data and the geographical auxiliary data through the hybrid prediction model to obtain a predicted flow amount; and performing abnormal flow analysis on the target abnormal gravitation value and the predicted flow amount to obtain an abnormal flow area range. The method can consider the mutual influence between the terminals, adapts to the dynamic change of the market, and is continuously self-optimized to adapt to the complex and changeable market environment.

Description

Technical Field

[0001] The present application relates to the technical field of data processing and analysis, and in particular to an analysis method and system based on a tobacco flow gravity model. Background Art

[0002] In related technologies, with the rapid development of the cigarette industry and increasingly fierce market competition, the problem of illegal tobacco circulation and sales across regions has become increasingly prominent, seriously affecting the healthy development of the industry. Existing regulatory methods include those that rely on manual inspections, anomaly detection based on sales data, flow analysis based on logistics data, or abnormal circulation detection using machine learning algorithms. However, existing regulatory methods often have the following problems: they are lagging, making it difficult to detect and prevent illegal circulation in a timely manner; they only focus on data from a single dimension and lack comprehensive consideration of multi-dimensional market factors; static regulatory methods are difficult to adapt to the rapidly changing market environment and the ever-evolving illegal circulation methods; they ignore the mutual influence between terminals and the spatial relationship of the market, making it impossible to accurately portray the overall market situation; and they lack effective anomaly handling and model optimization mechanisms, making it difficult to continuously improve the accuracy and reliability of detection.

[0003] In summary, the technical problems existing in the relevant technologies need to be improved. Summary of the Invention

[0004] The main purpose of the embodiments of the present application is to propose an analysis method and system based on a tobacco flow gravity model, which can take into account the mutual influence between terminals, adapt to market dynamics, and continuously self-optimize to adapt to a complex and changing market environment.

[0005] To achieve the above objectives, one aspect of the embodiments of the present application provides an analysis method based on a tobacco flow gravity model, the method comprising:

[0006] Obtain sales data, production data, monthly return data and geographic auxiliary data of target tobacco terminals;

[0007] Inputting the sales data and the production data into a target flow gravity model to obtain a terminal gravity value;

[0008] Performing anomaly detection on the terminal gravity value to obtain a target abnormal gravity value;

[0009] Predicting the monthly return flow data and the geographic auxiliary data using a hybrid prediction model to obtain a predicted flow amount;

[0010] The target abnormal gravity value and the predicted flow amount are subjected to abnormal flow analysis to obtain the abnormal flow area range.

[0011] In some embodiments, the method for constructing the target flow gravity model comprises the following steps:

[0012] Obtain the historical total sales and historical total production of the area to be tested;

[0013] Processing the historical sales volume and the historical production volume through an initial flow gravity model to obtain a terminal gravity value;

[0014] The initial flow gravity model is modified by using a particle model to obtain a target flow gravity model.

[0015] In some embodiments, the method of modifying the initial flow gravity model using a mass point model to obtain a target flow gravity model includes the following steps:

[0016] Constructing a plurality of mass points of unit mass with the target tobacco terminal as the center, wherein the mass points are distributed between the target tobacco terminal and other terminals;

[0017] Obtaining the particle distance between two adjacent particles;

[0018] In obtaining the particle distance, the particle farther from the target tobacco terminal is subject to the gravitational force of the particle closer to the target tobacco terminal;

[0019] Calculating the ratio of the gravity of the particle to the distance of the particle to obtain a particle ratio;

[0020] Obtaining a correction coefficient by comparing the mass point ratio with a first preset threshold and a second preset threshold;

[0021] The parameters in the initial flow gravity model are adjusted using the correction coefficient to obtain the target flow gravity model.

[0022] In some embodiments, performing abnormality detection on the terminal gravity value to obtain a target abnormal gravity value includes the following steps:

[0023] sorting the terminal gravity values from small to large;

[0024] Screening the sorted terminal gravity values through an anomaly judgment model to obtain preliminary anomaly gravity values;

[0025] plotting the preliminary abnormal gravity values into an abnormal gravity value distribution map;

[0026] The target abnormal gravity value is obtained by analyzing and screening the distribution of the preliminary abnormal gravity value in the abnormal gravity value distribution map.

[0027] In some embodiments, screening the sorted terminal gravity values using an abnormality judgment model to obtain preliminary abnormal gravity values includes the following steps:

[0028] Summarizing the terminal gravity values that are lower than a preset lower limit threshold to obtain a first abnormal gravity value;

[0029] Summarizing the terminal gravity values that are higher than a preset upper threshold value to obtain a second abnormal gravity value;

[0030] Summarizing the terminal gravity values whose change amplitudes compared with the previous period exceed a preset change rate threshold to obtain a third abnormal gravity value;

[0031] Summarize the terminal gravity values that exceed the average value of the surrounding terminal gravity values and exceed a preset difference threshold to obtain a fourth abnormal gravity value;

[0032] The first abnormal gravity value, the second abnormal gravity value, the third abnormal gravity value, and the fourth abnormal gravity value are aggregated to obtain the preliminary abnormal gravity value.

[0033] In some embodiments, predicting the monthly return flow data and the geographic auxiliary data using a hybrid prediction model to obtain a predicted flow volume includes the following steps:

[0034] Preprocessing the monthly flow-back data and the geographic auxiliary data;

[0035] Capturing the pre-processed monthly flow-back data and the geographic auxiliary data through a time series model to obtain first prediction data;

[0036] Performing multivariate feature integration on the pre-processed monthly flow return data, the geographic auxiliary data, and the first prediction data through a machine learning model to obtain second prediction data;

[0037] The first prediction data and the second prediction data are weighted averaged to obtain the predicted flow.

[0038] In some embodiments, performing abnormal flow analysis on the target abnormal gravity value and the predicted flow amount to obtain the abnormal flow area range includes the following steps:

[0039] Classifying the predicted flow into risk levels;

[0040] Marking the target anomaly gravity value on the regional distribution map;

[0041] Marking the predicted flow volume after risk level classification on the regional distribution map to obtain a predicted flow volume risk area;

[0042] A second distinction mark is performed on the overlapping portion of the predicted flow risk area and the target abnormal gravity value to obtain the range of the abnormal flow area.

[0043] In some embodiments, the method further includes performing abnormal processing on the obtained target abnormal gravity value to obtain abnormal processing data, including the following steps:

[0044] The target abnormal gravity value is corrected by a time series analysis method or multi-source data fusion correction to obtain abnormal processing data.

[0045] In some embodiments, it further includes:

[0046] Obtaining historical production data of the target tobacco terminal;

[0047] Analyzing the historical production data using a time series forecasting model to obtain a predicted total production volume;

[0048] Calculating the predicted total production volume using a gravity coefficient formula to obtain a gravity coefficient;

[0049] Adjusting the target flow gravity model by using the gravity coefficient to obtain an abnormal flow gravity prediction model;

[0050] The sales data and the production data are calculated using the abnormal flow gravity prediction model to obtain a predicted gravity value.

[0051] To achieve the above objectives, another aspect of the present application provides an analysis system based on a tobacco flow gravity model, the system comprising:

[0052] The first module is used to obtain sales data and production data of the target tobacco terminal;

[0053] The second module is used to input the sales data and the production data into a target flow gravity model to obtain a terminal gravity value;

[0054] The third module is used to perform anomaly detection on the terminal gravity value to obtain a target abnormal gravity value;

[0055] The fourth module is used to predict the monthly flow return data and the geographic auxiliary data by using a hybrid prediction model to obtain a predicted flow amount;

[0056] The fifth module is used to perform abnormal flow analysis on the target abnormal gravity value and the predicted flow amount to obtain the abnormal flow area range.

[0057] The embodiments of the present application include at least the following beneficial effects: This application provides an analysis method and system based on a tobacco flow gravity model. This solution processes sales data and production data using a target flow gravity model to obtain terminal gravity values. After performing anomaly detection on the terminal gravity values, target abnormal terminal gravity values are obtained. By analyzing the terminal gravity values and target abnormal gravity values, a more comprehensive and accurate portrayal of market trends can be achieved. This method can not only identify obvious abnormal flows, but also capture some hidden and complex illegal circulation patterns. During the processing of sales and production data, model parameters are continuously adjusted according to actual conditions to improve detection accuracy. The system also fully considers the mutual influence between terminals and the spatial relationship of the market to more accurately identify possible abnormal flows. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a flow chart of an analysis method based on a tobacco flow gravity model provided in an embodiment of the present application;

[0059] Figure 2 is a flow chart of a method for constructing a target flow gravity model provided in an embodiment of the present application;

[0060] Figure 3 yes Figure 1 Flowchart of step S300 in FIG.

[0061] Figure 4 It is a schematic diagram of an analysis system based on the gravity model of tobacco flow. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are merely examples of devices and methods consistent with some aspects of the embodiments of the present application.

[0063] It will be understood that the terms "first", "second", etc. used in this application may be used herein to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0064] The terms "at least one", "plurality", "each", "any", etc. used in this application include "at least one", "two" or more, "plurality" or "each", "any" or "any one", "each" or "any one" as used herein.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0066] Before explaining the embodiments of the present application in detail, some of the nouns and terms involved in the embodiments of the present application are first explained. The nouns and terms involved in the embodiments of the present application are subject to the following explanations.

[0067] Terminal gravity value: refers to the demand coefficient of tobacco products at terminals in each region. The larger the terminal gravity value, the greater the demand for tobacco products at terminals in that region.

[0068] Target abnormal gravity value: refers to the demand for the tobacco product in the terminal gravity value that is different from the normal market demand. The target abnormal gravity value is often the terminal gravity value of the illegal cross-regional circulation of tobacco products. The area where tobacco products are illegally circulated across regions can be inferred based on the target abnormal gravity value.

[0069] Predicted flow volume: refers to the prediction of changes in the flow volume of tobacco products in the future time period. Risk levels are divided by predicting changes in the flow volume values. Risk areas are determined based on the divided risk levels, and the intensity of supervision is adjusted accordingly.

[0070] Abnormal processing data: refers to the process of screening target abnormal gravity values through terminal gravity values, in which target abnormal gravity values that cannot reflect the illegal circulation of tobacco products across regions may be screened out. By obtaining abnormal processing data and conducting comprehensive analysis, more reliable target abnormal gravity values are obtained, thereby obtaining target abnormal gravity values that can better reflect the illegal circulation of tobacco products across regions.

[0071] Predicted gravity value: refers to the predicted gravity value obtained by predicting based on the acquired abnormal flow gravity prediction model. The predicted gravity value is used to judge the future market demand for the tobacco product, and the appropriate production quantity is adjusted according to the predicted gravity value.

[0072] Figure 1 is a flow chart of an analysis method based on a tobacco flow gravity model provided in an embodiment of the present application. Figure 1 The method may include but is not limited to steps S100 to S400:

[0073] Step S100: acquiring sales data, production data, monthly return data and geographic auxiliary data of the target tobacco terminal;

[0074] Step S200: input sales data and production data into the target flow gravity model to obtain the terminal gravity value;

[0075] Step S300: performing anomaly detection on the terminal gravity value to obtain a target abnormal gravity value;

[0076] Step S400: predicting the monthly return flow data and geographic auxiliary data using a hybrid prediction model to obtain predicted flow volume;

[0077] Step S500: performing abnormal flow analysis on the target abnormal gravity value and the predicted flow amount to obtain the abnormal flow area range.

[0078] In some embodiments, step S100 obtains sales data, production data, monthly return flow data, and geographic auxiliary data for the target tobacco terminal. The target tobacco terminal specifically refers to a cigarette terminal in the target area. The target area can be a city, a province, or a larger geographical area, depending on the actual application requirements. Sales data typically includes information such as sales revenue and sales volume for each terminal in different time periods, while production data includes information such as production amount and production volume for each terminal or production unit in the corresponding time period. Monthly return flow data refers to data related to tobacco products that should have been sold within a specific sales area but were returned to other areas due to various reasons in violation of regulations within a month, including information such as the quantity, brand, and direction of return flow. Geographic auxiliary data refers to the geographical distance between each area and the target tobacco terminal area, the regulatory path coefficient, and holiday marking data. The regulatory path coefficient is a quantitative indicator of the impact of regulatory measures on the target variable (such as the degree of tobacco market standardization, the incidence of violations, etc.) along a specific regulatory path. Monthly return flow data and geographic auxiliary data include regional location, flow volume (10,000 cigarettes), distance (km), regulatory coefficient, and holidays.

[0079] In some embodiments, in step S200, for the construction of the target flow gravity model, Figure 2 As shown, Figure 2 is a flow chart of a method for constructing a target flow gravity model provided in an embodiment of the present application, Figure 2 The method may include but is not limited to steps S610 to S630:

[0080] Step S610: Obtain the historical total sales volume and historical total production volume of the area to be detected;

[0081] Step S620: Process the historical sales volume and historical production volume using the initial flow gravity model to obtain a terminal gravity value;

[0082] Step S630: Correcting the initial flow gravity model using the particle model to obtain a target flow gravity model;

[0083] In some embodiments, in step S610, historical sales data and historical production data are obtained, and the total sales volume for each time period is calculated based on the obtained historical sales data, while the total production volume for each time period is calculated based on the historical production data. The purpose of this step is to integrate the original scattered data into summary data that can be directly used for model construction. Preferably, the total sales volume and total production volume can be calculated by weighted average using historical sales data and historical production data to better reflect the importance of different time periods. For example, recent data can be given a higher weight to make the model more in line with the current market conditions. Based on the calculated total sales volume and total production volume, a target flow gravity model is constructed. The core idea of this model is to regard each cigarette terminal as a point with a terminal gravity value, and the size of its terminal gravity value is affected by the sales volume and production volume of the terminal. By analyzing the interaction between these points, possible abnormal flow situations can be identified.

[0084] In some embodiments, in step S620, the initial flow gravity model includes a first formula, a second formula, and a third formula. The first formula is used to calculate the total sales volume and the total production volume to obtain the terminal non-cigarette gravity value; the second formula is used to calculate the terminal non-cigarette gravity value to obtain the terminal gravity; and the third formula is used to calculate the terminal gravity to obtain the terminal gravity value.

[0085] Specifically, the terminal non-cigarette gravity value is determined by a first formula, wherein the first formula is:

[0086]

[0087] Among them, F n The terminal non-cigarette gravity value of terminal n, C k represents the total sales volume of the terminal in the kth period, P nk It represents the total production of terminal n in the kth period, and α represents the influence coefficient of the gravity value of non-cigarettes.

[0088] The first formula considers the relationship between sales and production. By introducing the influence coefficient α, the degree of influence of sales and production on the terminal gravity value can be flexibly adjusted. Generally, the value of α ranges from 0.5 to 1.5, and the specific value can be adjusted according to actual circumstances. For example, when α is close to 1, sales and production have equal influence on the terminal gravity value; when α is greater than 1, production has a greater impact on the terminal gravity value; when α is less than 1, sales have a greater influence on the terminal gravity value.

[0089] The gravitational force of the terminal is determined by the second formula, where the second formula is:

[0090]

[0091] Among them, G N represents the gravitational force of terminal n, r n represents the gravitational radius of terminal n, F n It represents the terminal non-cigarette gravity value of terminal n in the first formula.

[0092] The second formula draws on the law of gravity in physics to relate the gravitational force of the terminal to the distance. n It can be understood as the range of a terminal's influence, which is usually determined by factors such as the terminal's size and location. For example, for a large retail terminal, its gravitational radius may be within 5-10 kilometers, while for a small terminal, its gravitational radius may be only 1-2 kilometers.

[0093] After calculating the gravity of each terminal, the gravity of each terminal on other terminals is calculated using the third formula. Taking into account the mutual influence between terminals, it is helpful to analyze the abnormal flow of non-smoke more comprehensively. The third formula is:

[0094] L n =∑(G i );

[0095] Among them, L n G represents the terminal gravitational force of each terminal on terminal n, that is, the terminal gravitational force of terminal n in the research area. i The third formula summarizes the influence of all terminals on a specific terminal and obtains the terminal gravity value of the terminal in the entire study area.

[0096] In some embodiments, in step S630, the initial flow gravity model is modified using a particle model to obtain a target flow gravity model. This includes: constructing a plurality of particle points of unit mass with the target tobacco terminal as the center, with the particle points distributed between the target tobacco terminal and other terminals; obtaining a particle distance between two adjacent particle points; obtaining a particle distance in which the particle point farther from the target tobacco terminal is subjected to a gravity from the particle point closer to the target tobacco terminal; calculating a ratio of the particle gravity to the particle distance to obtain a particle ratio; comparing the particle ratio with a first preset threshold and a second preset threshold to obtain a correction coefficient, and adjusting the parameters in the first formula using the correction coefficient to obtain the target flow gravity model.

[0097] Specifically, a number of unit mass particles are constructed with the terminal in the research object as the center of the circle. These particles are distributed between the terminal of the research object and other terminals. This processing method simplifies the complex terminal network into a particle system that can be mathematically processed. Then, among the two adjacent particles, the gravitational force F generated by the particle located farther from the research object terminal and the particle located closer to the research object terminal is obtained, and the distance r between the two particles is obtained. The ratio of the gravitational force F to the distance r is calculated, that is, The particle ratio reflects the intensity of the interaction between particles and is an important basis for correction.

[0098] The calculated mass point ratio is compared with the first preset threshold and the second preset threshold to obtain a correction coefficient. The correction coefficient is positively correlated with the non-cigarette gravity value influence coefficient of the first formula in the target flow gravity model. The non-cigarette gravity value influence coefficient is adjusted by obtaining the adjusted correction coefficient, thereby correcting the initial flow gravity model to obtain the target flow gravity model. Specifically:

[0099] when When β′=β×(1+δ1);

[0100] when When β′=β;

[0101] when When β′=β×(1-δ2);

[0102] Wherein, T1 represents the first preset threshold, T2 represents the second preset threshold, β represents the initial correction coefficient, β′ represents the adjusted correction coefficient, and δ1 and δ2 represent adjustment parameters.

[0103] This segmented adjustment method can more flexibly respond to correction needs in different situations. Generally, T1 can be set to 1.2, T2 can be set to 0.8, and δ1 and δ2 can be set to 0.1. The specific values of these parameters can be fine-tuned according to the actual application scenario and historical data analysis results to obtain the best correction effect. For example, assuming that the initial correction coefficient β of a terminal is 1.0, the calculated The value is 1.3. Since 1.3 > T1 (assuming T1 = 1.2), β′ = β × (1 + δ1), and the adjusted β′ is 1.0 * (1 + 0.1) = 1.1. This means that the influence of this terminal is considered to be stronger than the initial estimate, so its correction factor needs to be appropriately increased. This approach ensures the flexibility and accuracy of the model, making the target flow gravity model closer to reality. Dynamically adjusting the correction model parameters based on actual conditions makes the target flow gravity model more accurately reflect the complex conditions in the real world, allowing the model to continuously optimize itself over time and improve its ability to identify and predict abnormal flow gravity.

[0104] The target flow gravity model innovatively combines the concept of gravity from physics with the specific characteristics of the cigarette industry, providing a new and effective tool for identifying and preventing illegal tobacco distribution. This approach not only considers the two key factors of sales and production but also incorporates flexible correction methods, making the model more adaptable to complex and volatile market environments. This model, constructed using this approach, allows cigarette companies to more accurately identify potential illegal distribution channels, enabling them to implement appropriate preventive and control measures, effectively maintaining market order and corporate profits.

[0105] In step S200, the sales data and production data are input into the constructed target flow gravity model to obtain the terminal gravity value.

[0106] In some embodiments, as Figure 3 As shown, Figure 3 yes Figure 1 Flowchart of step S300 in FIG. Figure 1 Step S300 includes but is not limited to steps S310 to S340:

[0107] Step S310: sorting the terminal gravity values from small to large;

[0108] Step S320: Screening the sorted terminal gravity values through the abnormality judgment model to obtain preliminary abnormal gravity values;

[0109] Step S330: plotting the preliminary abnormal gravity value into an abnormal gravity value distribution map;

[0110] Step S340: Analyze and screen the distribution of the preliminary abnormal gravity values in the abnormal gravity value distribution map to obtain the target abnormal gravity value.

[0111] In some embodiments, in step S310, the calculated terminal gravity values are sorted from smallest to largest. This sorting method helps to intuitively present the relative status and influence of each terminal in the entire study area. The sorting results can help managers quickly identify terminals that may require special attention.

[0112] In some embodiments, in step S320, the terminal gravity values sorted by the abnormality judgment model are screened to obtain preliminary abnormal gravity values, including: summarizing the terminal gravity values below the preset lower limit threshold to obtain a first abnormal gravity value; summarizing the terminal gravity values above the preset upper limit threshold to obtain a second abnormal gravity value; summarizing the terminal gravity values whose change amplitude exceeds the preset change rate threshold compared with the previous time period to obtain a third abnormal gravity value; summarizing the terminal gravity values that exceed the average value of the surrounding terminal gravity values and exceed the preset difference threshold to obtain a fourth abnormal gravity value; summarizing the first abnormal gravity value, the second abnormal gravity value, the third abnormal gravity value and the fourth abnormal gravity value to obtain a preliminary abnormal gravity value.

[0113] Specifically, the abnormal judgment model includes: the terminal gravity value is lower than the preset lower limit threshold, the terminal gravity value is higher than the preset upper limit threshold, the change in the terminal gravity value compared with the previous period exceeds the preset change rate threshold, and the difference between the terminal gravity value and the average terminal gravity value of the surrounding terminals exceeds the preset difference threshold.

[0114] For example, if a terminal's gravity value falls below a preset lower threshold, let's say the lower threshold is set at 50% of the average terminal gravity value for the region. This means that if a terminal's gravity value falls below half the regional average, there may be an anomaly and further investigation is warranted. This could mean that sales at that terminal are unusually low or that there are unreported sales.

[0115] For cases where the terminal gravity value exceeds a preset upper threshold, such as 200% of the average terminal gravity value in the region, if the terminal gravity value exceeds twice the regional average, this may indicate that the terminal has abnormally high sales or may be a transit point for the circulation of illegal tobacco.

[0116] The criteria for determining the magnitude of changes in the terminal gravity value are equally important. For example, a preset change rate threshold of 30% can be set. This means that if a terminal gravity value changes by more than 30% compared to the previous period, whether it is an increase or a decrease, it may indicate an abnormal situation. Such sudden changes may be due to drastic changes in market conditions or illegal activities.

[0117] The difference between a terminal's gravity value and the average gravity value of surrounding terminals is also an important criterion. For example, if the preset difference threshold is set to 50%, if the gravity value of a terminal differs by more than 50% from the average gravity value of its surrounding terminals, this may indicate that the terminal's operating conditions or market environment are significantly different from those of the surrounding terminals, requiring further investigation.

[0118] These anomaly gravity values are aggregated to obtain a preliminary anomaly gravity value. These specific judgment criteria and threshold settings make the anomaly detection process of the present invention more objective and operational. It should be understood that these thresholds are not fixed and can be adjusted based on specific application scenarios and historical data analysis results to achieve optimal anomaly detection results.

[0119] In some embodiments, in steps S330 to S340, to more intuitively display abnormalities, an abnormal gravity value distribution map is plotted based on the screened preliminary abnormal gravity values. Based on the abnormal gravity value distribution map, a further determination is made as to whether an abnormal value exists. This step serves as a secondary confirmation of the previous anomaly detection results, helping to reduce the possibility of misjudgment and improve the accuracy of anomaly detection. This visualization approach can help decision makers better understand the spatial distribution and concentration of abnormalities, thereby formulating more targeted control measures.

[0120] In some embodiments, after performing anomaly detection on the terminal gravity value and obtaining the target abnormal gravity value, it also includes performing abnormal processing on the obtained target abnormal gravity value to obtain abnormal processing data, wherein the data acquisition frequency is resampled according to the frequency of occurrence of the target abnormal gravity value, or the target abnormal gravity value is corrected by time series analysis method or multi-source data fusion correction to obtain abnormal processing data.

[0121] Specifically, resampling includes increasing or decreasing the frequency of data collection. For example, if a terminal is found to have frequent abnormal values, you can consider increasing the data collection frequency of the terminal in order to more closely monitor its changes. On the contrary, if the data in a certain area has been very stable, you can appropriately reduce the sampling frequency to save resources. Data correction includes using trend extrapolation to perform data correction or multi-source data fusion correction. For example, when performing trend extrapolation, you can use time series analysis methods such as exponential smoothing or ARIMA models to predict reasonable terminal gravity values based on historical data, and replace the target abnormal gravity value with a reasonable terminal gravity value. Multi-source data fusion correction uses other relevant data sources, such as sales data from surrounding areas, market research reports, etc., to conduct comprehensive analysis to obtain more reliable data.

[0122] If resampling is selected, after the resampling and collection frequency adjustment, sales and production data from the target tobacco terminal are reacquired. This data is then fed into the target flow gravity model to obtain terminal gravity values and perform anomaly detection on these values. This cycle continues until the preset data quality requirements are met. If data correction is selected, sales and production data from the target tobacco terminal are acquired based on the obtained anomaly-processed data. This data is then fed into the target flow gravity model to obtain terminal gravity values and perform anomaly detection on these values. This cycle continues until the preset data quality requirements are met. This iterative optimization approach continuously improves the accuracy and reliability of the model. Data quality requirements include: the proportion of outliers does not exceed 5% of the total sample size, the target flow gravity model processing error is within 10%, and the target flow gravity model parameter change is less than 1% after three consecutive iterations. These specific indicators can be adjusted according to actual application needs to balance model accuracy and computational efficiency.

[0123] In some embodiments, in step S400, monthly return flow data and geographic auxiliary data are predicted using a hybrid prediction model to obtain predicted flow volume. This includes: preprocessing the monthly return flow data and geographic auxiliary data; capturing the preprocessed monthly return flow data and geographic auxiliary data using a time series model to obtain first predicted data; integrating multivariate features of the preprocessed monthly return flow data, geographic auxiliary data, and the first predicted data using a machine learning model to obtain second predicted data; and performing a weighted average of the first predicted data and the second predicted data to obtain predicted flow volume.

[0124] Specifically, preprocessing includes missing value processing, outlier correction, feature engineering, and data standardization. More specifically, missing value processing includes filling null values with 0 to assume that there is no flow or no record; outlier correction needs to be combined with the context of regulatory policies. For example, the flow volume in November 2024 is 160,000, which is marked as a special event; feature engineering includes time features, geographic features, regulatory path features, lag features, and moving averages. Among them, time features include month, quarter, and whether it is a holiday; geographic features are the straight-line distance to the target tobacco terminal area calculated based on geopy; regulatory path features are regulatory coefficients assigned according to the flow path; lag features include generating lag values of flow volume for the past 1-3 months; and moving average includes calculating a 6-month moving average to capture long-term trends; data standardization includes Min-Max standardization of numerical features to eliminate dimensionality effects.

[0125] The hybrid forecasting model includes a time series model and a machine learning model. The time series model is used to capture seasonality, trends, and holiday effects in the data, while the machine learning model is used to integrate multivariate features and handle nonlinear relationships. Specifically, the time series model is fed with single-city time series data in the format of date and flow volume, along with holiday labels (such as Spring Festival and Mid-Autumn Festival). Prophet is used to configure the time series model, with annual periodicity analysis enabled and weekly periodicity ignored. Custom holiday data is used, with the holidays_df parameter accounted for the impact of special dates, and fitting is performed using the df_city data to explore long-term trends in the data and patterns influenced by holidays. This generates the first forecast data, which includes forecasted flow volume for the next 12 months and seasonal decomposition results (such as trends and holiday effects). The machine learning model is fed with basic features, dynamic features, and the output of the time series model (the first forecast data). Basic features include distance, regulatory coefficient, month, and holiday; dynamic features include lag values and moving averages; and the output of the time series model includes forecasted future flow volume and its confidence interval. For the machine learning model configuration, we used XGBRegressor for regression analysis, setting the number of trees to 200, the maximum depth to 5, the learning rate to 0.05, the sample ratio to 0.8, and the objective function to square error. We used the training data sets X_train and y_train for fitting. We used time series cross-validation to prevent future data leakage, using evaluation metrics including mean squared error, mean absolute error, and R². The machine learning model then outputted a second prediction. The first and second predictions were weighted averaged to obtain the predicted flow. The weighted average assigns weights based on the validation set performance.

[0126] The hybrid forecasting model generates predicted liquidity. The predicted liquidity is compared with the same period in previous years to calculate the year-on-year change. Risk levels are then assigned based on this year-on-year change. For example, if the predicted liquidity is 398 and the year-on-year change is +12%, the risk level is considered very high. The predicted liquidity is visualized using a time series decomposition plot and a feature importance plot. The time series decomposition plot shows trends, seasonality, and holiday effects, while the feature importance plot displays the contribution of each feature in the XGBoost model to the prediction.

[0127] The risk level classification includes:

[0128] Very high risk: predicted liquidity > historical mean + 2σ;

[0129] High risk: historical mean + σ < predicted liquidity ≤ historical mean + 2σ;

[0130] Medium risk: historical mean - σ < predicted liquidity ≤ historical mean + σ;

[0131] Low risk: predicted liquidity ≤ historical mean - σ;

[0132] Here, σ represents the grade parameter obtained based on experience.

[0133] Multiple experiments with the hybrid prediction model show that distance and the regulatory coefficient have the greatest impact on predicted mobility, followed by holidays. The hybrid prediction model generates a list of high-risk cities based on monthly updated data, enabling real-time monitoring. By adjusting the regulatory coefficient, we predict changes in predicted mobility under varying levels of regulation, enabling regulatory simulations.

[0134] In some embodiments, the structure for deploying time series models and machine learning models is described. Specifically, for the time series model, the Prophet library is used to construct the model. First, the Prophet class is imported, followed by instantiating model_prophet. During structural deployment, yearly_seasonality is enabled to capture annual cyclical patterns; weekly_seasonality is set to False to ignore the influence of weekly cycles; and the holidays_df is passed in using the holidays parameter to account for the impact of custom holidays. Finally, the fit method is called to train the model using the df_city data, allowing the model to learn the data characteristics and patterns, completing the overall structural deployment. For the machine learning model, the XGBRegressor from the XGBoost library is used to construct the model. During deployment, the number of trees is set to 200 to control model complexity and fitting ability; the maximum tree depth is set to 5 to avoid overfitting; the learning rate is set to 0.05 to adjust the step size of each iteration; the sample sampling ratio is set to 0.8, and a random portion of samples is selected for training; the objective function is the squared error, which is used for regression tasks. Finally, the model is fitted using the training sets X_train and y_train to allow the model to learn the data characteristics and patterns, completing the structural deployment.

[0135] In some embodiments, in step S500, the target abnormal gravity value and the predicted flow are subjected to abnormal flow analysis to obtain the abnormal flow area range. Specifically, the predicted flow is divided into risk levels; the target abnormal gravity value is marked on the regional distribution map; the predicted flow after risk level division is marked with a first difference mark on the regional distribution map to obtain the predicted flow risk area; the part where the predicted flow risk area and the target abnormal gravity value overlap is marked with a second difference mark to obtain the abnormal flow area range. Specifically, the regional distribution map can be drawn using geographic information system (GIS) software, and the appropriate map projection method and scale are selected according to actual needs to clearly display the geographical features of the area. Different colors, symbols or charts are used to intuitively represent different numerical ranges when performing the first difference mark and the second difference mark.

[0136] In some embodiments, a gravity field distribution map is drawn based on the target flow gravity model, and the terminal gravity values are annotated on the gravity field distribution map. Based on the gravity field distribution map, the status of the terminals in the study area is analyzed. Terminal status includes out-of-favor and favored states, and each terminal is determined to be out-of-favor or favored. Terminals with terminal gravity values below 0 are defined as out-of-favor terminals, while terminals with terminal gravity values above 0 are defined as favored terminals. This classification method is simple and intuitive, making it easy to operate and understand in practice.

[0137] After determining the terminal's status, corresponding advertising recommendations are made. For terminals that have fallen out of favor, it's recommended to send advertising recommendations from favored terminals to them. This approach aims to leverage the influence of favored terminals to drive business development for those that have fallen out of favor. Conversely, for favored terminals, advertising recommendations targeted at those that have fallen out of favor are sent to favored terminals. This strategy aims to expand the influence of favored terminals while helping to improve the performance of those that have fallen out of favor.

[0138] In some embodiments, when the number of pet terminals is lower than a preset number, a recommendation to adjust the operating strategy of each terminal in the study area will be triggered. This mechanism is set up to ensure that the entire study area can maintain a healthy competitive situation and prevent excessive concentration or monopoly. For example, the preset number can be set to 30% of the total number of terminals. If the proportion of pet terminals is lower than this threshold, it may mean that the market is too concentrated or the overall environment is not healthy enough. In this case, some measures may need to be taken, such as increasing support for small and medium-sized terminals, or adjusting pricing strategies, etc., to promote the healthy development of the market.

[0139] Overall, this application not only effectively identifies and handles anomalies but also provides specific business recommendations based on the model's results. This comprehensive solution, from data analysis to practical application, provides a powerful tool for market management and operational decision-making in the cigarette industry. By continuously applying and optimizing the target flow gravity model, we can better understand market dynamics, effectively prevent illegal circulation, and develop more precise market strategies.

[0140] In some embodiments, the method further includes: obtaining historical production data of a target tobacco terminal; calculating the historical production data using the fourth formula or a time series prediction model to obtain a predicted total production; calculating the predicted total production using a gravity coefficient formula to obtain a gravity coefficient; adjusting the target flow gravity model using the gravity coefficient to obtain an abnormal flow gravity prediction model; and calculating sales data and production data using the abnormal flow gravity prediction model to obtain a predicted gravity value. The abnormal flow gravity prediction model not only enables analysis of current market conditions but also enables reasonable prediction of future trends, providing an important basis for the company's long-term strategic decision-making.

[0141] Specifically, historical production data for each terminal in the study area is obtained for each time period. This historical data forms the basis for forecasting and typically covers a time span of 12 to 36 months to capture any seasonal fluctuations and long-term trends. Based on this historical production data, the predicted total production for each terminal is calculated using the fourth formula. The fourth formula is:

[0142]

[0143] Among them, P n represents the predicted total production of terminal n, P nk represents the historical production data for terminal n in period k, where K represents the total number of periods. The fourth formula is suitable for relatively stable market environments. However, in practice, more complex forecasting models, such as exponential smoothing or ARIMA models, can be selected to improve forecast accuracy, depending on the specific situation.

[0144] Based on the predicted total production, the terminal gravity coefficient is adjusted through the gravity coefficient formula. The gravity coefficient formula is:

[0145]

[0146] Among them, β n represents the gravitational coefficient of terminal n, P′ n represents the predicted total production of terminal n, P nrepresents the rated production volume of terminal n, and γ represents the adjustment parameter. When the predicted production volume exceeds the rated production volume, the attraction coefficient of that terminal is increased; otherwise, it is decreased. The adjustment parameter γ typically ranges from 0.5 to 1.5 and can be adjusted based on actual conditions. For example, when γ = 1, the change in the attraction coefficient is proportional to the change in production volume; when γ < 1, the change in the attraction coefficient is relatively gradual; and when γ > 1, the change in the attraction coefficient is amplified.

[0147] Assume that the rated production volume of a terminal is P n is 1000 units, and the total production volume P′ is predicted based on historical data. n is 1200 units, and the adjustment parameter γ is set to 1. Substituting the specific value into the gravity coefficient formula, we get This means that because the forecasted production volume exceeds the rated production volume, the terminal's gravity coefficient has been adjusted to 1.2. This indicates that the terminal's influence is expected to increase in the future, as it is expected to produce more products, thereby increasing its market appeal. Based on the abnormal flow gravity prediction model, the forecasted gravity value for each terminal in the future is calculated. This dynamic adjustment method enables the model to better adapt to market changes and improves the accuracy and reliability of forecasts.

[0148] In some embodiments, the method further includes: obtaining historical consumption data, historical sales data, and historical production data from the target tobacco terminal; quantifying the linear relationship between the historical consumption data, historical sales data, and historical production data using the Pearson correlation coefficient to obtain a correlation; measuring the parity of the historical consumption data, historical sales data, and historical production data by calculating the number of consecutive rising or falling months; measuring the dispersion of the historical consumption data, historical sales data, and historical production data by calculating the coefficient of variation; and analyzing the correlation, parity, and dispersion to obtain abnormal production status. This multi-dimensional data provides a rich source of information for analysis. Generally, historical consumption data, historical sales data, and historical production data should cover a time span of at least 12 months to capture possible seasonal variations. Historical consumption data represents the amount of money consumers spend on goods, focusing on reflecting their purchasing power and demand. Historical sales data represents the revenue generated by a company from selling goods, focusing more on its sales performance and market presence.

[0149] Specifically, the Pearson correlation coefficient is used to quantify the linear relationship between historical consumption data, historical sales data, and historical production data. If the correlation coefficient between consumption and sales is close to 1, while the correlation between production and them is low, this may indicate the presence of inventory backlog or sales from abnormal channels.

[0150] Parity is measured by counting the number of consecutive months of increases or decreases, while dispersion is measured by calculating the coefficient of variation (standard deviation divided by the mean). These metrics can help identify unusual patterns in the data. For example, if sales data for a particular terminal exhibits high dispersion, while consumption and production figures remain relatively stable, this may indicate irregular large-scale transactions at that terminal.

[0151] Abnormal production status is determined by analyzing correlations, parity, and dispersion. Specifically, abnormal production status is divided into two categories: stable production and abnormal production. Stable production is further divided into benign production and hidden abnormal production, while abnormal production is further divided into obvious production anomalies and hidden production anomalies. For example, when the correlation between historical consumption data, historical sales data, and historical production data is stable, and the parity and dispersion are within the normal range, it can be determined to be benign production. If the correlation is stable but the dispersion is slightly high, it may be hidden abnormal production. When the correlation changes significantly, or the dispersion of a certain indicator is abnormally high, it can be determined to be an obvious production anomaly. When both the correlation and dispersion are in a critical state, it may be a hidden production anomaly. This multi-level classification method can more meticulously characterize the production status of the terminal, helping to promptly identify potential problems and implement appropriate management measures.

[0152] In some embodiments, the application of the analysis method based on the tobacco flow gravity model provides a powerful support tool for market management of the cigarette industry.

[0153] Specifically, consumption and production data from cigarette terminals in the target area are obtained. This data forms the foundation of the entire analysis process and can typically be obtained through a company's sales and production systems, as well as through market research. By inputting sales and production data into the constructed target flow gravity model, terminal gravity values are obtained. Anomalies are detected on these terminal gravity values to obtain target abnormal gravity values. Abnormal flow analysis is then performed on these terminal gravity values and target abnormal gravity values to determine abnormal flow trends. Applications based on abnormal flow trends include abnormal flow analysis, market strategy development, inventory optimization, risk assessment, resource allocation, and performance evaluation.

[0154] Specifically, abnormal flow analysis: By analyzing the distribution and changing trends of each terminal's gravity value, abnormal flow patterns within the study area can be identified. For example, if a terminal's gravity value suddenly rises sharply while surrounding terminals' gravity values simultaneously decrease, this may indicate illegal cross-regional product transfers. By promptly identifying these abnormal patterns, companies can take appropriate measures to prevent and combat illegal distribution channels. Market strategy formulation: Based on the distribution of terminal gravity values, companies can develop differentiated marketing strategies. For example, for terminals with high terminal gravity values, they can consider expanding product offerings or providing more promotional support; while for terminals with low terminal gravity values, they may need to improve service quality or adjust their product mix. Inventory optimization: Based on forecasts of future terminal gravity value changes, companies can conduct more scientific inventory management. For example, for terminals predicted to see an increase in terminal gravity values, they can appropriately increase inventory levels; for terminals predicted to see a decrease in terminal gravity values, they can consider reducing inventory to avoid capital tie-up. Risk assessment: Combining the results of abnormal production status analysis, companies can conduct a comprehensive assessment of the operating risks of each terminal. For example, for terminals that are determined to have hidden production anomalies, it may be necessary to strengthen supervision and auditing to prevent potential violations. Resource allocation: Based on the terminal gravity value and abnormal status, enterprises can more reasonably allocate various resources, such as marketing funds, human resources, etc. For example, more resources can be invested in terminals with high growth potential but low current terminal gravity values to promote their rapid development. Performance evaluation: Changes in terminal gravity values are used as one of the important indicators of performance evaluation. For example, the increase in terminal gravity values can be included in the KPI system of terminal managers to motivate them to take effective measures to enhance the market influence of the terminal.

[0155] Through these specific applications, companies can comprehensively improve the efficiency and accuracy of market management. For example, after applying this method, a tobacco company successfully identified five terminals with abnormal flows, promptly blocked a key illegal distribution channel, and increased the company's legal sales by 3% the following year. Another company, by optimizing inventory management, reduced inventory turnover days from 45 to 35 days, significantly reducing capital tied up. Overall, the analysis method based on the tobacco flow gravity model provides the cigarette industry with a comprehensive, systematic, and efficient market management tool. By comprehensively considering multiple factors such as consumption, production, and sales, and combining it with advanced data analysis techniques, this method can help companies better grasp market dynamics, effectively prevent illegal distribution, and optimize resource allocation, thereby gaining a competitive advantage in the fierce market.

[0156] like Figure 4 As shown, an analysis system based on a tobacco flow gravity model includes:

[0157] The first module is used to obtain sales data and production data of the target tobacco terminal;

[0158] The second module is used to input sales data and production data into the target flow gravity model to obtain the terminal gravity value;

[0159] The third module is used to perform anomaly detection on the terminal gravity value and obtain the target abnormal gravity value;

[0160] The fourth module is used to predict the monthly flow return data and geographical auxiliary data through a hybrid prediction model to obtain the predicted flow volume;

[0161] The fifth module is used to perform abnormal flow analysis on the target abnormal gravity value and the predicted flow volume to obtain the abnormal flow area range.

[0162] It can be understood that the contents of the above method embodiments are all applicable to the present system embodiments, the functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0163] The embodiments described in the embodiments of this application are intended to more clearly illustrate the technical solutions of the embodiments of this application and do not constitute a limitation on the technical solutions provided by the embodiments of this application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0164] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0165] Those skilled in the art will appreciate that all or some of the steps in the methods, systems, and functional modules / units in the devices disclosed above may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0166] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0167] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0168] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0169] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, 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, which is stored in a storage medium and includes multiple instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: various media that can store programs, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0170] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but are not intended to limit the scope of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and essence of the present invention should be within the scope of the present invention.

Claims

1. An analysis method based on a tobacco flow gravity model, characterized in that: The method comprises: Obtain sales data, production data, monthly return data and geographic auxiliary data of target tobacco terminals; Inputting the sales data and the production data into a target flow gravity model to obtain a terminal gravity value; Performing anomaly detection on the terminal gravity value to obtain a target abnormal gravity value; Predicting the monthly return flow data and the geographic auxiliary data using a hybrid prediction model to obtain a predicted flow amount; The target abnormal gravity value and the predicted flow amount are subjected to abnormal flow analysis to obtain the abnormal flow area range.

2. The method according to claim 1, characterized in that The method for constructing the target flow gravity model comprises the following steps: Obtain the historical total sales and production of the area to be tested; Processing the historical sales volume and the historical production volume through an initial flow gravity model to obtain a terminal gravity value; The initial flow gravity model is modified by using a particle model to obtain a target flow gravity model.

3. The method according to claim 2, characterized in that The method of correcting the initial flow gravity model by using the mass point model to obtain the target flow gravity model includes the following steps: Constructing a plurality of mass points of unit mass with the target tobacco terminal as the center, wherein the mass points are distributed between the target tobacco terminal and other terminals; Obtaining the particle distance between two adjacent particles; In obtaining the particle distance, the particle farther from the target tobacco terminal is subject to the gravitational force of the particle closer to the target tobacco terminal; Calculating the ratio of the gravity of the particle to the distance of the particle to obtain a particle ratio; Obtaining a correction coefficient by comparing the mass point ratio with a first preset threshold and a second preset threshold; The parameters in the initial flow gravity model are adjusted using the correction coefficient to obtain the target flow gravity model.

4. The method according to claim 1, wherein The abnormality detection of the terminal gravity value to obtain the target abnormal gravity value includes the following steps: sorting the terminal gravity values from small to large; Screening the sorted terminal gravity values through an anomaly judgment model to obtain preliminary anomaly gravity values; plotting the preliminary abnormal gravity values into an abnormal gravity value distribution map; The target abnormal gravity value is obtained by analyzing and screening the distribution of the preliminary abnormal gravity value in the abnormal gravity value distribution map.

5. The method according to claim 4, characterized in that The method of screening the sorted terminal gravity values by using an abnormality judgment model to obtain a preliminary abnormal gravity value includes the following steps: Summarizing the terminal gravity values that are lower than a preset lower limit threshold to obtain a first abnormal gravity value; Summarizing the terminal gravity values that are higher than a preset upper threshold value to obtain a second abnormal gravity value; Summarizing the terminal gravity values whose change amplitudes compared with the previous period exceed a preset change rate threshold to obtain a third abnormal gravity value; Summarize the terminal gravity values that exceed the average value of the surrounding terminal gravity values and exceed a preset difference threshold to obtain a fourth abnormal gravity value; The first abnormal gravity value, the second abnormal gravity value, the third abnormal gravity value, and the fourth abnormal gravity value are aggregated to obtain the preliminary abnormal gravity value.

6. The method according to claim 1, characterized in that The monthly return flow data and the geographic auxiliary data are predicted using a hybrid prediction model to obtain the predicted flow volume. The following steps are included: Preprocessing the monthly flow-back data and the geographic auxiliary data; Capturing the pre-processed monthly flow-back data and the geographic auxiliary data through a time series model to obtain first prediction data; Performing multivariate feature integration on the pre-processed monthly flow return data, the geographic auxiliary data, and the first prediction data through a machine learning model to obtain second prediction data; The first prediction data and the second prediction data are weighted averaged to obtain the predicted flow.

7. The method according to claim 1, characterized in that The abnormal flow analysis is performed on the target abnormal gravity value and the predicted flow amount to obtain the abnormal flow area range, including the following steps: Classifying the predicted flow into risk levels; Marking the target anomaly gravity value on the regional distribution map; Marking the predicted flow volume after risk level classification on the regional distribution map to obtain a predicted flow volume risk area; A second distinction mark is performed on the overlapping portion of the predicted flow risk area and the target abnormal gravity value to obtain the range of the abnormal flow area.

8. The method according to claim 1, characterized in that The method further includes performing abnormal processing on the obtained target abnormal gravity value to obtain abnormal processing data, including the following steps: The target abnormal gravity value is corrected by a time series analysis method or multi-source data fusion correction to obtain abnormal processing data.

9. The method according to claim 1, characterized in that Also includes: Obtaining historical production data of the target tobacco terminal; Analyzing the historical production data using a time series forecasting model to obtain a predicted total production volume; Calculating the predicted total production volume using a gravity coefficient formula to obtain a gravity coefficient; Adjusting the target flow gravity model by using the gravity coefficient to obtain an abnormal flow gravity prediction model; The sales data and the production data are calculated using the abnormal flow gravity prediction model to obtain a predicted gravity value.

10. An analysis system based on a tobacco flow gravity model, characterized in that: The system comprises: The first module is used to obtain sales data and production data of the target tobacco terminal; The second module is used to input the sales data and the production data into a target flow gravity model to obtain a terminal gravity value; The third module is used to perform anomaly detection on the terminal gravity value to obtain a target abnormal gravity value; The fourth module is used to predict the monthly flow return data and the geographic auxiliary data by using a hybrid prediction model to obtain a predicted flow amount; The fifth module is used to perform abnormal flow analysis on the target abnormal gravity value and the predicted flow amount to obtain the abnormal flow area range.

Images