Intelligent commodity display system based on model algorithm

Abstract

The invention relates to an intelligent commodity display system based on a model algorithm, and belongs to the field of new-generation information technology industry. Commodity display in the past is usually subjectively judged according to experience of store personnel, so that display effects are uneven, consumers are difficult to find desired commodities, the consumers cannot be retained, and greater economic benefits cannot be brought to enterprises. In order to solve the problem, the invention provides an automatic commodity display model algorithm based on overall planning and single store drawing, the model comprehensively considers the planning requirements of headquarters for store display, and combines the operating list, inventory data, sales data, commodity appearance size data and the like of store commodities to obtain the automatic commodity display model algorithm. And intelligent display is realized by comprehensively applying various algorithms and technical means. The system perfectly solves the past problems of the chain retail industry, does not need to increase the labor cost, can complete automatic display according to the commodities of each store, and truly achieves the purposes of cost reduction, efficiency improvement and thousand-store thousand-face.

Description

Technical Field

[0001] This invention is a management software for intelligent merchandise display. Its main function is to simplify the complex display process and realize personalized store strategies. Each store can customize a unique display plan according to its own characteristics. It is mainly used in industries such as chain supermarkets, convenience stores, chain pharmacies, maternal and infant products, and cosmetics. Background Technology

[0002] With the popularization of domestic chain retail stores, rising labor costs, and intensified competition, retail stores are generally experiencing declining operating profits. Optimizing product display, attracting customers, and improving staff efficiency have become important development strategies for retail stores. In the operation of convenience stores, product display is usually based on the subjective judgment of store staff's experience, as well as the company's internal sales strategies and sales data. However, due to the high turnover rate and inconsistent staff quality, the display results can be inconsistent, making it difficult for consumers to find the products they want and failing to retain customers. Currently, there are many display-related products on the market, but they only move the offline store display to the online display and do not require each store to make its own display. The headquarters will plan the display uniformly. However, due to the shortage of graphic designers and the large number of display images, the workload of graphic designers is huge. Moreover, graphic designers cannot understand the operating conditions of each store. The images they produce are only standard images and cannot achieve a unique look for each store. As a result, stores lose their individuality and reduce their competitiveness. If a unique look for each store were to be achieved, a large number of graphic designers would be needed, which would increase labor costs and does not conform to the current concept of cost reduction and efficiency improvement. Therefore, the automatic product display perfectly solves the above problems. It does not require additional labor costs, and the system can automatically display products according to each store, truly achieving cost reduction and efficiency improvement, and a unique look for each store. Summary of the Invention

[0003] To address the aforementioned issues, this invention provides an automatic merchandise display model algorithm based on overall planning and single-store output. This model comprehensively considers the headquarters' planning requirements for store displays, while also incorporating the store's merchandise operation list, inventory data, sales data, and merchandise appearance and size data. By comprehensively utilizing multiple algorithms and technologies, it achieves personalized displays for each store. The overall process of this invention is as follows: I. Overall Headquarters Planning 1. Headquarters Merchandise Planning: Planning merchandise from different dimensions, including the operational dimensions of merchandise, the labels of merchandise, and the upper and lower limits of merchandise display, in order to have overall control over merchandise and refine the merchandise management philosophy; 2. Planning in the dimension of merchandise operation: From a macro perspective, merchandise is divided into regional levels, and from a micro perspective, merchandise is divided into structural categories, business districts, key products, strategic products, and store-specific products, etc., to ensure a more reasonable merchandise structure that can not only meet the individual needs of stores, but also enable the headquarters to have overall control over merchandise. 3. Product Tag Planning: Create a variety of product tags, and configure product tags flexibly to enable tag selection and rule setting in various scenarios; 4. Product display upper and lower limit planning: Reasonably plan the upper and lower limits of product display, set the upper and lower limits of product display, and achieve a reasonable proportion of product resources while ensuring the integrity of the category structure. II. Headquarters Rules and Guidelines The headquarters formulates plans that conform to the corporate culture, display standards, and marketing strategies. Through refinement in multiple dimensions, these plans are developed into actionable rules to achieve the company's goals. 1. Rule Division: Based on the different store areas, shelf types, and operating conditions, the same product category is divided into rules to maximize the rationality of the rules; 2. Block Space Guidance Planning: Rationally plan the display length of different labels on the shelf, calculate the optimal length of labels and the number of products based on the algorithm, and combine the current company display requirements and marketing strategies to plan the optimal label block ratio; 3. Block Label Guidance Planning: Different display rules are set for different label blocks, such as brand rules, position order rules, height priority rules, sales priority rules, ingredient rules, and other different rule combinations, so that each block can be adapted to the most suitable rules. 4. Product Standard Guidance Planning: For special products, special displays can be specified, including product location, product layout, product and block combination, and product and product combination; 5. Rule Allocation: Based on the algorithm logic of rule division and the logic of display template establishment, different rules are adapted to different display templates, so that each display image has its most suitable display rules. III. Smart Store Grouping By intelligently analyzing data such as store sales, profit, gross profit, average order value, best-selling products, and consumer shopping behavior, stores are tagged and grouped into different levels and price ranges, laying a solid foundation for intelligent product selection. IV. Intelligent Store Display Analyze the headquarters' display rules and implement them in every step of the display process; 1. Confirm the number of product SKUs in a block: Confirm the number of product SKUs in the block based on the rules configured in the block, the product data under the category to which the block belongs, and the product tags; 2. Block width confirmation: Calculate the block width based on the configured block size and the width of the shelf; 3. Display method confirmation: Analyze the block display rules, confirm whether the display method is vertical or horizontal, and then place the SKUs of the block products on the shelves in order according to the display method; 4. Rule-based product SKU confirmation: Based on the block rules, tags, product tags, and block products, confirm the number of product SKUs under each rule in the block, and divide the block rules into the corresponding number of grids according to the number of SKUs; 5. Rule Placement Confirmation: Confirm which label of product should be placed in each cell according to the block rules; 6. Product Location Confirmation: Place the products in the corresponding compartments according to the product labels; 7. Confirmation of product display area and stack number: Based on the set number of display areas and stack number of products under the block rules, display the corresponding number of display areas and stack number of products under the block rules. If the product is not set, the data of the block rules will be used. If the product is set, the data set by the product will be used first. 8. Product Area Expansion: Determine if there are any blocks with too few products, resulting in insufficient shelf display. If such blocks exist, check if there are any blocks above, below, left, or right with too many products to display. If so, supplement the display of products according to the priority principle of the blocks above, below, left, and right. If it is not possible to supplement all products, the remaining products will not be displayed. 9. Product Expansion: Obtain the maximum expansion area and expansion base number set for the current block. Sort the products according to the comprehensive ratio of sales volume, sales revenue, and gross profit (sales volume: sales revenue: gross profit = 3:3:4). Prioritize expanding the products with the highest expansion area. First, expand by one expansion base number. If there is still space remaining after all products have been expanded, continue to expand according to this rule until expansion is no longer possible (the remaining space is not enough to fully display the products that need to be expanded by one expansion base number). Products that have reached the maximum number of expansion areas will no longer be expanded. 10. Shelf Height Adjustment: After all the products are displayed, adjust the shelf height from bottom to top. The shelf height should be the height of the tallest product on the current shelf + 2 cm to allow space for retrieving the products. V. Intelligent Display Analysis, Detection, and Reminder The effectiveness of the intelligent display is evaluated to check for any flaws and whether the current display meets basic aesthetic requirements. 1. Shelf full display rate detection: Calculate the full display rate of a single shelf section: Product width in linear meters (sum of the widths of all products on the shelf) / Shelf length in linear meters (sum of the widths of each shelf panel) * 100% Determining if shelf fullness is reasonable: Fullness of ≥90% is considered reasonable. 2. Shelf Height Adaptation Detection: Detects whether the current shelf height distribution is reasonable; Shelf height: Data from the shelving configuration settings; Product shelf thickness: The height of each product shelf (product height: take the tallest product on each shelf as the calculation value) plus the shelf thickness. By comparing the height of the shelf with the height of the product shelf, determine whether the current product can make full use of the width of the shelf; Matching: If the height of the product shelf is the same as the height of the shelf, then the shelf height is the height of the tallest product on each shelf; Exceeding: If the height of the product shelf exceeds the height of the shelf, first calculate the excess height of the product (product shelf height - shelf height). The display requirement is to minimize the number of products exceeding the shelf. Set the shelf height from bottom to top according to the height of the tallest product on each shelf until the top shelf. The height of the top shelf is the remaining height. Disadvantage: If the height of the product shelf is lower than the height of the shelf, the shelf height should first be the height of the tallest product on each shelf. Then, the height of the product plus the shelf height should be subtracted from the shelf height to get the remaining height. The remaining height should be divided by the number of shelves to calculate the height that each shelf should be increased. Finally, the most suitable height for each shelf can be obtained. 3. Oversized Product Detection: The sum of the widths of the products on the shelf is compared with the width of the shelf. If the width of the product on the shelf is greater than the width of the shelf, the product is oversized. 4. Unlisted Product Detection: Detects data on unlisted products among the products associated with the current display image; The effect achieved by the invention

[0004] The automatic merchandise display system of this invention provides a more efficient and accurate solution for optimizing store merchandise display in the retail industry, achieving the following significant effects: Personalized Displays for Every Store: The system utilizes model algorithms to analyze the headquarters' overall display requirements and plans, combining this with the store's inventory list to automatically generate a customized display for each store, truly achieving a scientific and rational personalized display for every store. Optimized Display Results: The intelligent display system can adjust the placement and method of product display based on real-time data analysis, store product characteristics, and sales strategies to achieve optimal product presentation. Reduced manpower input: Traditional store displays require store managers to have high professional qualities and keen observation skills, while common display software on the market requires a lot of manpower. Intelligent display systems can automatically complete the display of goods, realize a thousand stores with a thousand different displays, while reducing labor costs and lowering the requirements for personnel professional qualities, thus achieving true cost reduction and efficiency improvement. Attracting consumers' attention: By using scientific display methods to showcase product features and create a pleasant shopping environment and atmosphere, we can attract consumers' curiosity and make it easier for them to discover and select the products they need, thereby maximizing the attraction of consumers' attention. Increased sales volume and profit margin: Intelligent display systems can use algorithms to output the most suitable display method for products, the most suitable products for bundled sales, and the display positions of products with different gross profit, sales volume, and sales volume, making it easier for customers to find their favorite products and stimulating consumers' desire to buy, prompting consumers to increase their purchase volume, thereby increasing sales volume and profit margin. Maximizing store revenue: The intelligent display system calculates the sales performance of the same product under different locations, different display layouts, and different left and right products based on information such as the previous product display position, number of rows, and left and right products. It then determines the optimal product display and combination to maximize store revenue. In summary, the intelligent merchandise display system of this invention provides retail stores with a precise and reasonable merchandise display solution. Through highly integrated automation functions and a scientific calculation model, it significantly improves store revenue and reduces labor costs.

Claims

1. A model-based intelligent merchandise display system, characterized by: (1) Display logic definition module: Based on the category classification results of retail enterprises and for different categories, the display rule elements and the weight of each element are visualized on the system. (2) Store intelligent grouping module: Based on the sales performance and product category structure of retail stores, the module groups stores according to the sales, profit, space ratio and business district attributes, using a similarity algorithm based on major categories; (3) Automatic display result drawing module: Based on the sales and profit differences brought by each product, as well as the product size, volume and proportion of display resources, the module automatically calculates the product's display position, number of display surfaces and quantity on the shelf and generates display drawings according to the display logic described in (1); (4) Effect evaluation module: The module scores the display results generated in (3) from the dimensions of display resource utilization rate and compliance rate of the display logic rules described in (1).

2. The similarity algorithm as described in claim 1 (2), characterized in that... Cluster analysis (K-Means clustering algorithm) is performed on all stores of a retail enterprise under the same category to identify store groups with similar structures and performance.