Industrial product dynamic display system with intelligent recommendation algorithm

Abstract

The invention discloses an industrial product dynamic display system with an intelligent recommendation algorithm, and the working process of the system comprises the steps: a data collection and storage stage: building an industrial product three-dimensional model database, and achieving the digital storage and management of product data; performing multi-dimensional classification labeling on the warehousing products, and constructing a structured product database; a user behavior analysis stage: collecting and recording product browsing behavior data of the user in real time; analyzing user browsing records based on the consulting quantity and the consulting time, and quantifying the preference degrees of the user for different types of products; in the intelligent recommendation stage, matched products are screened from the classification database for personalized display in combination with a product classification system and user preferences. According to the invention, the user experience can be improved. The method can be widely applied to the field of intelligent interaction.

Description

Technical Field

[0001] This invention relates to the field of intelligent interaction, and in particular to a dynamic display system for industrial products equipped with an intelligent recommendation algorithm. Background Technology

[0002] At present, the development of technologies such as intelligent interactive multimedia display systems has also provided technical support for the dynamic display of industrial products.

[0003] In current product display systems, users are often presented with a default product each time they enter the system. The system lacks the ability to intelligently filter product categories to match the user's preferences, resulting in a poor user experience and potentially discouraging users from exploring other products from the same company. Therefore, providing product recommendations based on user interests to improve conversion rates is a significant technical challenge that urgently needs to be addressed by those skilled in the art. Summary of the Invention

[0004] In view of this, in order to solve the technical problem that existing industrial product display systems cannot intelligently filter recommended products, resulting in a poor user experience, this invention proposes an industrial product dynamic display system with an intelligent recommendation algorithm. The system includes the following steps:

[0005] In the data acquisition and storage phase, a 3D model database of industrial products is established to realize the digital storage and management of product data; multi-dimensional classification and labeling of products entering the warehouse are carried out to build a structured product database;

[0006] During the user behavior analysis phase, user product browsing behavior data is collected and recorded in real time; user browsing records are analyzed based on the number of views and the time of view to quantify the degree of user preference for different categories of products;

[0007] In the intelligent recommendation stage, products are selected from the category database and displayed in a personalized manner, based on the product classification system and user preferences.

[0008] In some embodiments, during the data acquisition and storage phase, octree partitioning is used to achieve model lightweighting, and the WebSocket long connection protocol is used to achieve real-time synchronization.

[0009] In some embodiments, the user behavior analysis phase also considers price factors.

[0010] Based on the above solution, the present invention provides an industrial product dynamic display system with intelligent recommendation algorithm. For each user's independent account, the system identifies their historical browsing records and calculates the product types and price ranges preferred by the customer. The system proactively adjusts the default products displayed on the default display page to match the user's preferences, thereby enhancing their willingness to learn more and their willingness to purchase. Attached Figure Description

[0011] Figure 1 This is a flowchart of the workflow of an industrial product dynamic display system with intelligent recommendation algorithm according to the present invention. Detailed Implementation

[0012] In addition to the problem mentioned in the background technology that the default product display of current industrial product display systems ignores the user's actual intention, the following problems also exist: when loading models of large industrial equipment, lag issues are prone to occur, resulting in poor display effects and user experience; cross-terminal collaboration is not possible, which means that after the PC is updated, the mobile device may not be updated in sync, resulting in slower mobile consultations, and vice versa.

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] It should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0015] It should be understood that the terms "system," "apparatus," "unit," and / or "module" used in this application are a method of distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0016] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "a," and / or "the" are not specifically singular and may include the plural. Generally, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements. An element defined by the phrase "comprising an..." does not exclude the presence of other identical elements in the process, method, product, or apparatus that includes the element.

[0017] In the description of the embodiments of this application, "a plurality of" refers to two or more. The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0018] Furthermore, flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, the steps can be processed in reverse order or simultaneously. Additionally, other operations can be added to these processes, or one or more steps can be removed from them.

[0019] Reference Figure 1 This is a schematic diagram of the workflow of an optional example of the dynamic display system for industrial products with intelligent recommendation algorithm proposed in this invention. The system can be applied to computer equipment, and the processing method proposed in this embodiment may include, but is not limited to, the following steps:

[0020] Step S1: Upload the 3D model of the industrial product to the database;

[0021] Step S2: Classify the industrial products in the database to obtain a classified database;

[0022] Step S3: Obtain the user's browsing history;

[0023] Step S4: Based on the user's browsing history, generate product preferences by taking into account factors such as the number of views, the viewing time, and user actions.

[0024] Step S5: Based on the categorized database, make recommendations and displays according to the product preferences.

[0025] In some feasible embodiments, step S1 specifically includes:

[0026] An octree partitioning algorithm based on spatial partitioning is used to process the polygonal mesh of the 3D model of industrial products, achieving a 30%-50% compression of the number of vertices and faces. The compressed model data is then saved to the database.

[0027] Taking an axial wind turbine blade assembly as an example: Assuming its geometric topology data consists of a blade surface mesh composed of 12,000 triangular facets, approximately 8,500 facets are used to construct the key geometric structure of the blade, and approximately 3,500 facets are used to describe non-core features of secondary details. In the octree segmentation process, the system first uses the blade assembly boundingbox as the root node (the spatial coordinate range is set to X:0-500mm, Y:0-500mm, Z:0-300mm), and generates 8 child nodes in the first segmentation. Among them, child node A1 (the spatial coordinate range is, for example, X:100-200mm, Y:200-300mm, Z:50-150mm) contains blade clusters numbered 1-50. Since the total number of triangular faces in this cluster reaches 600,000 (an average of 12,000 faces per blade, so 50 would be 50 * 12,000 = 600,000), the same applies to A2-A8. Then, a second segmentation is triggered: still taking A1 as an example, A1 is subdivided into 8 micro-nodes, each containing about 5-6 blades, reducing the total number of faces to about 80,000-100,000. A2-A8 are also segmented in the same way. After that, a third and fourth segmentation will continue until the eighth segmentation is completed or the number of faces per blade in a micro-node is ≤1800 (retaining 70% of the core topological features), at which point the segmentation stops, ultimately achieving the compression of the blade group model data.

[0028] The origin of the rule of ≤1800 facets per leaf (retaining 70% of core topological features): In the original model, core features account for about 70% of the total features (8500 / 12000≈70%). This is the benchmark for "retaining 70% of core topological features". It completely retains the most critical 70% (about 6000) of the 8500 core features. However, through a mesh simplification algorithm (merging adjacent facets and deleting coplanar redundancy), it is compressed to 1800 facets (compression ratio of about 70%). In the end, although the number of 1800 facets is only 15% of the original (1800 / 12000), it completely covers the key information of the original 70% of core topological features. Therefore, it is called "retaining 70% of core topological features".

[0029] In some feasible embodiments, step S1 further includes:

[0030] The WebSocket long connection protocol is used to achieve real-time synchronization of model data across PCs, mobile devices, and VR devices.

[0031] When the PC, mobile, and VR devices start up, they all send a WebSocket connection request to the server. After the server responds, three continuous "communication channels" are formed between the three devices and the server. For example, when uploading a 3D model file of an industrial pump on the PC, the PC will tell the server through the WebSocket channel: "I have uploaded a new model with the ID Pump2025". At the same time, it will send a command to the mobile and VR devices through WebSocket: "Please load the model with the ID Pump2025". After receiving the message, the mobile and VR devices will immediately download the model from the cloud and display it. The whole process can be said to be almost synchronous.

[0032] In some feasible embodiments, step S2 specifically includes:

[0033] For different companies, product categories are set accordingly, such as water pump equipment as A, industrial furnace equipment as B, water filtration equipment as C, membrane manufacturing equipment as D, etc.

[0034] In some feasible embodiments, step S4 specifically includes:

[0035] Its formula is expressed as follows:

[0036] Max[α i *β i ], i = 1, ..., C

[0037]

[0038] β i =ω1·N i +ω2·D i +ω3·S i +ω4·Q i

[0039]

[0040] The above formula means that α i With β i In combination, product preferences are selected from different types of industrial products. For example, among α1*β2, α2*β2, and α3*β3, which one is the largest? For example, if α2*β2 is the largest, it means that this customer is most interested in products of type i2.

[0041] Where C represents the number of categories of industrial products, α i This represents the weighted total number of times a user browses industrial products of category i; This indicates the number of times a user browses industrial products of category i within one week; This indicates the number of times a user browses industrial products of category i within a week to a month; This indicates the number of times a user browses industrial products of category i within one to two months; This indicates the number of times a user browses industrial products of category i within 2 to 3 months; Ω1, Ω2, Ω3, and Ω4 represent the time weights for the corresponding time periods; β i The interaction depth factor represents the depth of interaction a user makes while browsing industrial products of type i; N i This indicates the total number of times a user viewed product category i in the past 3 months; D i This indicates the total number of times a user downloaded technical documentation while browsing product category i within the past three months; S i This indicates the total number of times a user shared content while browsing product category i within the past 3 months; Q i This represents the total number of times a user made customer service inquiries while browsing product category i in the past 3 months; ω1, ω2, ω3, and ω4 represent the weight coefficients of the corresponding operations.

[0042] Let N be the total number of customer historical browsing records. Then, the total number of browsing records for water pumps is NA, the total number of historical browsing records for industrial furnaces is NB, and so on, with NC, ND, and so on. At the same time, when calculating the total, the time factor of customer browsing products must also be considered. Therefore, a time weight Ω is added to the historical browsing records, with recently viewed product records having a higher weight and older viewed product records having a lower weight.

[0043] For example, browsing within the past week has a weight of 1.0, browsing from one week to one month has a weight of 0.7, browsing from one to two months has a weight of 0.5, and browsing from more than two months has a weight of 0.3. When calculating the total browsing volume for different device types, the total number of browsing history records for different time periods needs to be multiplied by the corresponding weight Ω, and then summed to calculate the corresponding N, making the algorithm more reflective of the customer's current focus and needs. Then, these calculated final browsing volumes for each type of product are sorted from largest to smallest. Assuming the sorting order is: NA > NB > NC > ND, it is determined that the customer has viewed water pumps with the highest number of views (NA), so the default display device can be automatically locked to water pumps, thus quickly attracting the user's attention.

[0044] After determining a product category based on the above results, such as water pump equipment as exemplified here, let's assume: "The prices of the products most frequently viewed by customers (i.e., water pump equipment) are C1, C2, C3...C100". Sort these prices by size, assuming we get: C1 < C2 < C3 = C4 < ... < C100. Set the quantile threshold to 10% to 90%, and then remove the smallest 10 digits and the largest 10 digits (i.e., the extreme prices on both sides) in C1 < C2 < C3 = C4 < ... < C100. This gives us the prices from C11 to C90.

[0045] Reassign the prices of C11 to C90 in ascending order (C11 < C12 < C13 ... < C90) so that C11 = C1, C12 = C2 ... C90 = C80;

[0046] Calculate the average value Avg of the values ​​from C1 to C80 using the following formula:

[0047]

[0048] Here, n = 80;

[0049] The standard deviation σ is calculated from the values ​​of C1 to C80 using the following formula:

[0050]

[0051] Where N = n = 80, μ = Avg, x i =C i .

[0052] Ultimately, the customer's ideal price range Avg±σ can be obtained.

[0053] In some feasible embodiments, in step S4 or S5, to further ensure accuracy, a price elasticity coefficient E is introduced to measure the customer's sensitivity to price changes in water pump equipment, specifically including:

[0054] The formula for calculating the price elasticity coefficient E is:

[0055]

[0056] Here, %ΔQ represents the percentage change in customer browsing time for water pump equipment, and %ΔP represents the percentage change in the price of water pump equipment.

[0057] If |E| > 1, it indicates that the customer is price-sensitive. When determining the default product to display, prioritize products with prices around Avg-σ. If |E| < 1, it indicates that the customer is not price-sensitive. The price range can be appropriately widened, and more distinctive or high-performance flagship products can be selected (pre-determine which are flagship products), with prices within the range of Avg±σ.

[0058] For example: Water pump A costs 150,000 and users browse for 2 minutes; water pump B costs 180,000 and users browse for 1 minute; water pump C costs 120,000 and users browse for 3 minutes. Based on water pump A, the price elasticity coefficient Eab for water pump B is %ΔQ / %ΔP = %[(1-2)min / 2min] / %[(18-15) / 15] = |-50% / 20%| = 2.5, |Eab| > 1, indicating that customers are price-sensitive. Based on water pump A, the price elasticity coefficient Eac for water pump C is %ΔQ / %ΔP = %[(3-2) / 2] / %[(12-15) / 15] = |50% / -20%| = 2.5, |Eac| > 1, further indicating that customers are price-sensitive, and lower prices can increase user attention. Therefore, in this case, a high-performance product with a price close to "Avg-σ" is recommended.

[0059] Finally, by combining the user's preferred ideal price range (Avg±σ) with the price elasticity coefficient (E), we can derive a reference conclusion for products that align with the user's current focus and price sensitivity. Based on this conclusion, we select water pump equipment products from the product library that correspond to the user's preferences and use them as the default products displayed on the user's page.

[0060] A dynamic display system for industrial products with intelligent recommendation algorithms, comprising:

[0061] The database construction module is used to execute steps S1 and S2;

[0062] The user analysis module is used to execute steps S3 and S4;

[0063] The recommended module is used to execute step S5.

[0064] The content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0065] A dynamic display device for industrial products equipped with intelligent recommendation algorithms:

[0066] At least one processor;

[0067] At least one memory for storing at least one program;

[0068] When the at least one program is executed by the at least one processor, the at least one processor implements the processing method of an industrial product dynamic display system with intelligent recommendation algorithm as described above.

[0069] The content of the above system embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above system embodiments, and the beneficial effects achieved are also the same as those achieved by the above system embodiments.

[0070] A storage medium storing processor-executable instructions, which, when executed by a processor, are used to implement the processing method of an industrial product dynamic display system with an intelligent recommendation algorithm as described above.

[0071] The content of the above system embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above system embodiments, and the beneficial effects achieved are also the same as those achieved in the above system embodiments.

[0072] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A dynamic display system for industrial products with an intelligent recommendation algorithm, characterized in that, Its workflow includes the following steps: Upload the 3D model of the industrial product to the database; The industrial products in the database are classified to obtain a classified database; Get user's browsing history; Based on the user's browsing history, product preferences are generated using factors such as the number of views, the viewing time, and the user's actions. Based on the categorized database, recommendations are displayed according to the product preferences.

2. The industrial product dynamic display system with intelligent recommendation algorithm according to claim 1, characterized in that, The formula for generating product preferences based on factors such as the number of views, the time spent viewing, and user actions is as follows: Max[a i *b i ],i=1,...,C b i =ω1·N i +ω2·D i +ω3·S i +ω4·Q i Where C represents the number of categories of industrial products, α i This represents the weighted total number of times a user browses industrial products of category i; This indicates the number of times a user viewed industrial products of category i within the first time period. This indicates the number of times a user viewed industrial products of category i during the second time period. This indicates the number of times a user viewed industrial products of category i during the third time period; This indicates the number of times a user browses industrial products of type i during the fourth time period; Ω1, Ω2, Ω3, and Ω4 represent the time weights of the corresponding time periods; β i The interaction depth factor represents the depth of interaction a user makes while browsing industrial products of type i; N i This indicates the total number of times a user has viewed product category i within a preset time period; D i This indicates the total number of times a user downloaded technical documents while browsing product category i within a preset time period; S i This indicates the total number of times a user shared content while browsing product category i within a preset time period; Q i This represents the total number of times a user made customer service inquiries while browsing products of type i within a preset time period; ω1, ω2, ω3, and ω4 represent the weight coefficients of the corresponding operations.

3. The industrial product dynamic display system with intelligent recommendation algorithm according to claim 2, characterized in that, Also includes: The product preferences are filtered and sorted according to price.

4. The industrial product dynamic display system with intelligent recommendation algorithm according to claim 2, characterized in that, Also includes: Calculate the user's ideal price range based on the product preference price.

5. The industrial product dynamic display system with intelligent recommendation algorithm according to claim 4, characterized in that, Also includes: By introducing a price elasticity coefficient, the user's ideal price range is adjusted to obtain the adjusted price range; Recommendations will be displayed based on the adjusted price range.

6. The industrial product dynamic display system with intelligent recommendation algorithm according to claim 1, characterized in that, The step of uploading the 3D model of the industrial product to the database specifically includes: Obtain 3D models of industrial products; Based on octree segmentation, the polygonal mesh of the 3D model is processed to obtain compressed model data; The compressed model data is saved to the database.

7. The industrial product dynamic display system with intelligent recommendation algorithm according to claim 1, characterized in that, In the step of uploading the 3D model of the industrial product to the database, the WebSocket long connection protocol is used.

8. A dynamic display device for industrial products with an intelligent recommendation algorithm, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the industrial product dynamic display system with intelligent recommendation algorithm as described in any one of claims 1-7.

Images