Multi-source consumption behavior feature fusion-based precise white spirit putting method and system

Abstract

The invention relates to the technical field of commodity supply chain management, and particularly provides a multi-source consumption behavior feature fusion-based liquor precise putting method and system, and the method comprises the following steps: S1, carrying out data input; s2, constructing a model; and S3, performing optimization. Compared with the prior art, the method can achieve the multi-source data fusion to enable the prediction error rate to be smaller than 9%, the weight model can respond to the market demand change in real time, the unsalable loss is reduced through block chain anti-counterfeiting and emergency reserve, and a personalized delivery strategy of cultural preference and climate characteristics is supported.

Description

Technical Field

[0001] The present invention relates to the technical field of commodity supply chain management, and specifically provides a precise liquor delivery method and system based on the fusion of multi-source consumer behavior characteristics. Background Art

[0002] The traditional dealer application volume is inflated, leading to inventory backlogs, significant regional consumption differences, and a demand forecasting error rate of over 20% during festival periods (critical periods such as Spring Festival / Mid-Autumn Festival).

[0003] The single historical sales volume forecasting model ignores the value of real-time click behavior, the static proportion allocation cannot adapt to the dynamic changes in the regional market, and there is a lack of a dealer credit assessment and anti-brush order verification mechanism.

[0004] How to accurately deliver liquor products to different regions at different times and stages using multiple dimensions as calculation factors is an urgent problem for those skilled in the art. Summary of the Invention

[0005] The present invention aims at the deficiencies of the above-mentioned existing technologies and provides a practical precise liquor delivery method based on the fusion of multi-source consumer behavior characteristics.

[0006] A further technical task of the present invention is to provide a rationally designed, safe and applicable precise liquor delivery system based on the fusion of multi-source consumer behavior characteristics.

[0007] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0008] The precise liquor delivery method based on the fusion of multi-source consumer behavior characteristics includes the following steps:

[0009] S1. Perform data input;

[0010] S2. Construct a model;

[0011] S3. Perform optimization.

[0012] Further, in step S1, the input indicators include the subscription volume, the number of people clicking on the mall, the regional transaction order volume, and the order popularity of the custom time;

[0013] The data processing method for the subscription volume is as follows:

[0014] Reliability = 0.6×(historical fulfillment volume / historical application volume) + 0.4×credit rating. The technical feature is to interface with the enterprise ERP system to obtain credit data in real time and automatically filter low-trust applications;

[0015] The data processing method for the number of people clicking on the mall is as follows:

[0016] Effective clicks = Σ(single-day clicks × 0.85^t × (1 + holiday compensation factor)),

[0017] The technical feature is to activate a 1.8 times compensation coefficient before the Spring Festival to identify potential explosive markets, and calculate the average transaction rate of the product in the store on the day of replenishment;

[0018] The data processing method of the regional transaction order volume is as follows:

[0019] Abnormal judgment = IF (single order volume > regional mean 3σOR price deviation > 40%),

[0020] The technical feature is to combine blockchain anti-counterfeiting data to intercept false transactions;

[0021] The data processing method of the custom time order heat is:

[0022] Heat value = Σ(order volume / (time interval+1)),

[0023] The technical feature is to support custom time ranges and dynamically capture changes in market trends.

[0024] Furthermore, in step S2, the dynamic weight allocation formula is:

[0025]

[0026] Capacity coefficient: based on base wine inventory and blending efficiency, formula: maximum output of the month = base wine inventory × 0.82;

[0027] Elasticity coefficient: price sensitivity function, high-end wine elasticity = 0.3-0.6, mass wine elasticity = 0.7-1.2;

[0028] Total delivery calculation:

[0029]

[0030] Furthermore, in step S3, the dynamic rebalancing mechanism:

[0031] When the actual sales volume in a certain area deviates from the predicted value by more than 15%, the weight is adjusted in real time:

[0032]

[0033] Risk hedging strategy:

[0034] 5% of total production capacity is reserved as emergency reserve, allocated according to regional risk level:

[0035]

[0036] The precise liquor delivery system based on the integration of multi-source consumption behavior characteristics first performs data input, then constructs a model, and finally conducts optimization.

[0037] Furthermore, when performing data input, the input metrics include the purchase volume, the number of people clicking on the mall, the number of regional transaction orders, and the order popularity within a custom time period.

[0038] The data processing method for the purchase volume is as follows:

[0039] Reliability = 0.6×(historical fulfillment volume / historical application volume) + 0.4×credit rating. The technical feature is to interface with the enterprise ERP system to obtain credit data in real time and automatically filter low-trust applications.

[0040] The data processing method for the number of people clicking on the mall is as follows:

[0041] Effective clicks = Σ(single-day click volume × 0.85^t × (1 + festival compensation factor)).

[0042] The technical feature is to activate a 1.8-fold compensation coefficient for a period of time before the Spring Festival to identify potential explosive markets and calculate the average transaction rate of this product in the store until the replenishment day.

[0043] The data processing method for the number of regional transaction orders is as follows:

[0044] Abnormality determination = IF(single-order volume > 3σ of the regional average OR price deviation > 40%).

[0045] The technical feature is to combine blockchain anti-counterfeiting data to intercept false transactions.

[0046] The data processing method for the order popularity within a custom time period is as follows:

[0047] Popularity value = Σ(order volume / (time interval + 1)).

[0048] The technical feature is to support a custom time range and dynamically capture market trend changes.

[0049] Furthermore, when constructing the model, the dynamic weight allocation formula:

[0050]

[0051] Production capacity coefficient: Based on the base liquor inventory and blending efficiency, the formula is: the maximum monthly production = base liquor inventory × 0.82.

[0052] Elasticity coefficient: The price sensitivity function, the elasticity of high-end liquor = 0.3 - 0.6, and the elasticity of popular liquor = 0.7 - 1.2.

[0053] Calculation of the total delivery volume:

[0054]

[0055] When optimizing, the dynamic rebalancing mechanism:

[0056] When the deviation between the actual sales volume and the predicted value in a certain region > 15%, trigger real-time adjustment of weights:

[0057]

[0058] Risk hedging strategy:

[0059] Reserve 5% of the total production capacity as emergency reserve, allocated according to the regional risk level:

[0060]

[0061] Compared with the prior art, the liquor precise placement method and system based on multi-source consumption behavior feature fusion of the present invention have the following prominent beneficial effects:

[0062] The multi-source data fusion of the present invention makes the prediction error rate < 9%. The weight model can respond to the changes in market demand in real time, reduce the backlog loss through blockchain anti-counterfeiting and emergency reserve, and support personalized placement strategies for cultural preferences (aroma type / package) and climate characteristics (temperature / humidity). Specific implementation manners

[0063] In order to enable those skilled in the art of the present technology to better understand the solution of the present invention, the present invention will be further described in detail below in conjunction with specific implementation manners. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

[0064] The following gives a preferred embodiment:

[0065] The liquor precise placement method based on multi-source consumption behavior feature fusion in this embodiment has the following steps:

[0066] S1. Perform data input;

[0067]

[0068]

[0069] S2. Construct a model;

[0070] Dynamic weight allocation formula:

[0071]

[0072] Production capacity coefficient: Based on the base liquor inventory and blending efficiency (formula: maximum monthly production = base liquor inventory × 0.82);

[0073] Elasticity coefficient: Price sensitivity function (elasticity of high-end liquor = 0.3 - 0.6, elasticity of mass-market liquor = 0.7 - 1.2);

[0074] Calculation of total delivery volume:

[0075]

[0076] S3. Optimize;

[0077] Dynamic rebalancing mechanism:

[0078] When the deviation between the actual sales volume and the predicted value in a certain region > 15%, trigger real-time adjustment of weights:

[0079]

[0080] Risk hedging strategy:

[0081] Reserve 5% of the total production capacity as emergency reserve and allocate it according to the regional risk level:

[0082]

[0083] Based on the above method, in the liquor precise delivery system based on the integration of multi-source consumer behavior characteristics in this embodiment, first, data input is performed; then, the model is constructed; finally, optimization is carried out.

[0084] Among them, when performing data input, the input indicators include the subscription volume, the number of people clicking on the mall, the regional transaction order volume, and the order popularity at the custom time;

[0085] The data processing method for the subscription volume is:

[0086] Reliability = 0.6 × (historical fulfillment volume / historical application volume) + 0.4 × credit rating. The technical feature is to dock with the enterprise ERP system to obtain credit data in real time and automatically filter low-trust applications;

[0087] The data processing method for the number of people clicking on the mall is:

[0088] Effective clicks = Σ (daily click volume × 0.85^t × (1 + festival compensation factor)),

[0089] The technical feature is to activate a 1.8-fold compensation coefficient for a period of time before the Spring Festival to identify potential explosive markets, and finally calculate the average transaction rate of this product in the store on the replenishment day;

[0090] The data processing method for the regional transaction order volume is:

[0091] Abnormal determination = IF (single order quantity > 3σ of regional average OR price deviation > 40%),

[0092] The technical feature is to combine blockchain anti-counterfeiting data to intercept false transactions;

[0093] The data processing method for customizing the order heat of time is as follows:

[0094] Heat value = Σ (order quantity / (time interval + 1)),

[0095] The technical feature is to support a custom time range and dynamically capture market trend changes.

[0096] When constructing the model, the dynamic weight distribution formula:

[0097]

[0098] Production capacity coefficient: Based on the base liquor inventory and blending efficiency, the formula: the maximum monthly production = base liquor inventory × 0.82;

[0099] Elasticity coefficient: Price sensitivity function, high-end liquor elasticity = 0.3 - 0.6, mass-market liquor elasticity = 0.7 - 1.2;

[0100] Total delivery volume calculation:

[0101]

[0102] When optimizing, the dynamic rebalancing mechanism:

[0103] When the deviation between the actual sales volume and the predicted value in a certain region > 15%, trigger real-time weight adjustment:

[0104]

[0105] Risk hedging strategy:

[0106] Reserve 5% of the total production capacity as an emergency reserve and allocate it according to the regional risk level:

[0107]

[0108] The above specific implementation manners are only specific cases of the present invention. The patent protection scope of the present invention includes but is not limited to the above specific implementation manners. Any technical solution that conforms to the above specific implementation manners of the present invention and any appropriate changes or substitutions made by those of ordinary skill in the relevant technical field shall fall within the patent protection scope of the present invention.

[0109] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A precise liquor delivery method based on the fusion of multi-source consumption behavior characteristics, characterized in that: The steps are as follows: S1. Input data; S2, construct the model; S3. Optimize.

2. The liquor precision delivery method based on multi-source consumption behavior feature fusion according to claim 1 is characterized in that: In step S1, the input indicators include the purchase volume, the number of mall clicks, the regional transaction order volume and the custom time order popularity; The data processing method of the subscription amount is as follows: Credibility = 0.6×(historical fulfillment volume / historical application volume)+0.4×credit rating. The technical feature is to connect to the enterprise ERP system to obtain credit data in real time and automatically filter out the least credible applications. The data processing method of the mall click number is: Effective clicks = Σ(single-day clicks × 0.85^t × (1 + holiday compensation factor)), The technical feature is to activate a 1.8 times compensation coefficient before the Spring Festival to identify potential explosive markets, and calculate the average transaction rate of the product in the store on the day of replenishment; The data processing method of the regional transaction order volume is as follows: Abnormal judgment = IF (single order volume > regional mean 3σOR price deviation > 40%), The technical feature is to combine blockchain anti-counterfeiting data to intercept false transactions; The data processing method of the custom time order heat is: Heat value = Σ(order volume / (time interval+1)), The technical feature is to support custom time ranges and dynamically capture changes in market trends.

3. The liquor precision delivery method based on multi-source consumption behavior feature fusion according to claim 2 is characterized in that: In step S2, the dynamic weight allocation formula is: Capacity coefficient: based on base wine inventory and blending efficiency, formula: maximum output of the month = base wine inventory × 0.82; Elasticity coefficient: price sensitivity function, high-end wine elasticity = 0.3-0.6, mass wine elasticity = 0.7-1.2; Total delivery calculation:

4. The liquor precision delivery method based on multi-source consumption behavior feature fusion according to claim 3 is characterized in that: In step S3, the dynamic rebalancing mechanism: When the actual sales volume in a certain area deviates from the predicted value by more than 15%, the weight is adjusted in real time: Risk hedging strategy: 5% of total production capacity is reserved as emergency reserve, allocated according to regional risk level:

5. The precise liquor delivery system based on the fusion of multi-source consumption behavior characteristics is characterized by: First, data is input; then, the model is built; finally, optimization is performed.

6. The liquor precision delivery system based on multi-source consumption behavior feature fusion according to claim 5 is characterized in that: When inputting data, the input indicators include the purchase volume, the number of mall clicks, the regional transaction volume and the custom time order popularity; The data processing method of the subscription amount is as follows: Credibility = 0.6×(historical fulfillment volume / historical application volume)+0.4×credit rating. The technical feature is to connect to the enterprise ERP system to obtain credit data in real time and automatically filter out the least credible applications. The data processing method of the mall click number is: Effective clicks = Σ(single-day clicks × 0.85^t × (1 + holiday compensation factor)), The technical feature is to activate a 1.8 times compensation coefficient before the Spring Festival to identify potential explosive markets, and calculate the average transaction rate of the product in the store on the day of replenishment; The data processing method of the regional transaction order volume is as follows: Abnormal judgment = IF (single order volume > regional mean 3σOR price deviation > 40%), The technical feature is to combine blockchain anti-counterfeiting data to intercept false transactions; The data processing method of the custom time order heat is: Heat value = Σ(order volume / (time interval+1)), The technical feature is to support custom time ranges and dynamically capture changes in market trends.

7. The liquor precision delivery system based on multi-source consumption behavior feature fusion according to claim 6 is characterized in that: When building a model, the dynamic weight allocation formula is: Capacity coefficient: based on base wine inventory and blending efficiency, formula: maximum output of the month = base wine inventory × 0.82; Elasticity coefficient: price sensitivity function, high-end wine elasticity = 0.3-0.6, mass wine elasticity = 0.7-1.2; Total delivery calculation:

8. The liquor precision delivery system based on multi-source consumption behavior feature fusion according to claim 7 is characterized in that: When optimizing, the dynamic rebalancing mechanism: When the actual sales volume in a certain area deviates from the predicted value by more than 15%, the weight is adjusted in real time: Risk hedging strategy: 5% of total production capacity is reserved as emergency reserve, allocated according to regional risk level:

Images