Intelligent operation platform of vending machine

Abstract

The invention relates to the technical field of intelligent operation, in particular to a vending machine intelligent operation platform which comprises a point location framework building module, a commodity intelligent matching and adjusting module and a commodity image generality analysis module, point location building rules are set through the point location framework building module, point location names are limited in the point location building process of a user, and the point location names are matched with the commodity image generality analysis module. When the point location is associated with the equipment, the name of the point location comprises the attribute of the place where the equipment is located, so that the user can select the commodity from the commodity information base according to the name of the point location when selecting the commodity from the commodity information base, and the user needs to process a large amount of complicated information when operating the service of the vending machine. According to the method, commodity types suitable for different point locations are easily forgotten, and a scene-based memory clue is created for a user by tightly binding the point location names with the attributes of the places where the equipment is located, so that the commodity information base can be quickly screened according to the point location names, and the commodity types related to the specific places can be accurately positioned.

Description

Technical Field

[0001] The present invention relates to the field of smart operation technology, and in particular to a smart operation platform for vending machines. Background Art

[0002] With the rapid development of social economy and the increasingly fast pace of people's lives, consumers have an increasingly strong demand for convenient shopping. As a device that provides 24-hour uninterrupted commodity sales services, vending machines have gradually been widely used in various places, such as schools, office buildings, shopping malls, transportation hubs, communities, etc. However, before the operation of traditional vending machines, the operator uploads multiple commodity information to the commodity information database, and then the user (specifically the commodity of the vending machine) logs in to the operation platform and selects the commodity suitable for the vending machine from the database, so that the customer can purchase the commodity in the vending machine;

[0003] However, in the existing smart operation platform for vending machines, operators will store a large number of products and rich product materials in the product information database in order to enable users to select products that meet their needs. However, due to the complexity and diversity of the actual attributes of the products, it is difficult for keywords to fully cover the characteristics of the products. Therefore, when users select products for vending machines through keywords, they still need to check the pictures of different products again to determine whether they are the products they expect. However, due to the excessive number of products in the product information database, it is impossible to analyze the user's expected products and recommend them to the user in time, which will reduce the user's satisfaction with the smart operation platform for vending machines. In view of this, we propose a smart operation platform for vending machines. Summary of the Invention

[0004] This invention aims to address the rapidly increasing demand for convenient shopping and the widespread use of vending machines in various locations as society develops. This has led to the emergence of intelligent operation platforms, hoping to leverage technology to improve efficiency and service. However, existing platforms, due to their complex product sources and limited screening mechanisms, contain a large number of products that do not meet user needs, making it difficult for users to select the products they desire for vending machines, resulting in reduced user satisfaction.

[0005] To achieve the above objectives, the present invention provides a smart vending machine operation platform, which includes a location architecture building module, a product intelligent matching and adjustment module, and a product image commonality analysis module, wherein:

[0006] The product intelligent matching and adjustment module collects the names of places with different attributes and the names of the required products, associates the place attributes with the required product names, and combines the product names into a set of required product names; matches the point names in the set of required product names corresponding to the point place names; marks the corresponding products, and prioritizes the marked products when selecting products;

[0007] The product image commonality analysis module perceives the point name A and enters the product information, calls out the product image set A in the product information, combines the product images of the perception point name A in the product information library into a product image set B, analyzes the commonalities of the product images in the product image set B, and calls out the product images that meet the commonalities in the product image set A as the first image to be entered as the product.

[0008] As a further improvement of this technical solution, the point architecture building module is used to establish a connection channel with the user's mobile device and create points using point creation technology;

[0009] The point architecture construction module contains multiple account passwords with different usage permissions. When a user establishes a connection channel with the point architecture construction module, the mobile device sends the login account password to the point architecture construction module through the communication network. The point architecture construction module senses the login account password sent by the user's mobile device, and compares the login account password with the account password stored in itself one by one. If the login account password is exactly the same as the account password stored in the point architecture construction module, the point architecture construction module sends a connection success response to the mobile device through the communication network to establish a connection channel. If the login account password is not exactly the same as the account password stored in the point architecture construction module, the point architecture construction module sends a connection failure response to the mobile device through the communication network and refuses to establish a connection channel.

[0010] As a further improvement of this technical solution, the point creation technology in the point architecture building module sets point creation rules: when creating a point, a line can be created only after the area is created, and a point can be created only after the line is created;

[0011] Each point has a corresponding point name, which is the attribute of the location of the equipment.

[0012] As a further improvement of the present technical solution, after the point architecture building module successfully establishes a connection channel with the mobile device, the point architecture building module senses the creation area, line and point request and device SN code sent by the mobile device through the connection channel;

[0013] After receiving the request to create areas, lines and points, the point architecture building module compares the request information to see if it is complete. If the request information is incomplete, the point architecture building module outputs a complete signal to the mobile device.

[0014] As a further improvement of this technical solution, the expression for comparing whether the request information is complete in the point architecture building module is as follows:

[0015] Perceive that creating areas, lines, and points in the point architecture building module requires m1, m2, and m3 mandatory information fields.rj (i=1,2,…,m1) indicates the required information field of the area, F lk (k=1,2,…,m2) represents the required information field of the line, F ps (s=1,2,…,m3) indicates the required information field of the point;

[0016] Define the field validity judgment function Valid(F) as follows:

[0017]

[0018] The regional information integrity judgment formula is:

[0019] The formula for judging line information integrity is:

[0020] The formula for judging the integrity of point information is:

[0021] When Validity r =1, Validity l =1 and Validity p =1, the request information is complete, and the point architecture building module creates the point.

[0022] As a further improvement to this technical solution, the names of places with different attributes and the names of the required products in the product intelligent matching adjustment module are collected by a web crawler. The web crawler simulates the behavior of a browser and automatically captures information on web pages according to rules. The specific steps are as follows:

[0023] Step 1: Use various commercial websites and industry forums as the starting URL of the web crawler. The web crawler sends HTTP requests to the servers of various commercial websites and industry forums to request the web page content corresponding to the specified URL;

[0024] Step 2: After receiving the request, the servers of various commercial websites and industry forums find the corresponding web page according to the requested URL and send the web page content in HTML format as a response to the crawler program;

[0025] Step 3: After receiving the web page content, the crawler program uses the HTML parser to parse the page and extract the names of places with different attributes and the names of the required products.

[0026] As a further improvement of the present technical solution, the string matching method in the commodity intelligent matching adjustment module is as follows: starting from the first character of the point name string, it is compared with the first character of the place name string;

[0027] If they are equal, continue to compare the next character, and so on, until all characters of the place name string are matched successfully. If all characters of the place name string can find corresponding equal characters in the point name string in sequence, then the place where the place name is collected is determined to be the place where the point is located.

[0028] As a further improvement to this technical solution, the product intelligent matching and adjustment module senses the desired product names in different locations, and again uses a string matching method to match the desired product names with the product names in the product information database, retrieves the matching product names in the product information database, and marks the compared products. When selecting products, the marked products are prioritized.

[0029] The priority sorting method is as follows: give the marked product a priority higher than the unmarked product. i Assign priority p i (p i =1), unmarked product a m Assign priority p j (p j =0); if two marked products appear, they are sorted randomly.

[0030] As a further improvement to the present technical solution, the product image commonality analysis module analyzes the commonality of product images through color histogram analysis. The color histogram divides the color space of the product image into several intervals at equal intervals, and counts the frequency of color occurrence in each interval to describe the color distribution of the product image.

[0031] If different product images are similar in color composition and distribution, the color histograms of the product images will also be similar. By comparing the color histograms of different product images, the commonalities of the product images in the product image set B can be found.

[0032] As a further improvement of the present technical solution, the product image commonality analysis module perceives the color histogram of each product in the product image set A, calculates the similarity between each color histogram and the color histogram of the product image set B in turn, and calls out the product image with the highest similarity as the first image of the product to be entered.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] In this vending machine smart operation platform, point creation rules are set through a point architecture building module. When a user creates a point, the point name is limited. When the point is associated with a device, the point name includes the location attribute of the device. When a user selects a product in the product information library, the user can select the product in the product information library according to the point name. Because users need to process a large amount of complex information when operating a vending machine business, they are prone to forget the types of products suitable for different points. By closely binding the point name with the location attribute of the device, a scene-based memory clue is created for the user, so that the product information library can be quickly filtered according to the point name, and the product category related to the specific location is accurately located.

[0035] At the same time, the commodity intelligent matching and adjustment module analyzes the commodities required for the corresponding equipment according to the point name. When the user selects commodities for the equipment, the commodity intelligent matching and adjustment module increases the priority of the commodities required for the point in the commodity information library, so that the user can quickly find the required commodities in the commodity information library. When the user selects commodities for the equipment, he does not have to blindly browse and filter in the massive commodity information library, which helps to improve the user's satisfaction and loyalty to the vending machine smart operation platform, increase the frequency of user use and dependence on the platform, and promote the sustainable development of the platform business.

[0036] In addition to the above-described objects, features and advantages, the present invention has other objects, features and advantages. The present invention will be further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is the overall module principle diagram of the present invention;

[0038] Figure 2 Schematic diagram of the module for the point structure of the present invention;

[0039] Figure 3 Schematic diagram of the point architecture building module and mobile equipment of the present invention;

[0040] Figure 4 This is a schematic diagram of the commodity intelligent matching and adjustment module of the present invention;

[0041] Figure 5 This is a schematic diagram of the commodity image commonality analysis module of the present invention.

[0042] The meaning of each number in the figure is:

[0043] 100. Point structure building module; 200. Product intelligent matching and adjustment module; 300. Product image commonality analysis module. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0045] Example 1

[0046] refer to Figure 1-Figure 5 , a vending machine intelligent operation platform, including a point structure building module 100, a product intelligent matching adjustment module 200 and a product image commonality analysis module 300;

[0047] The point structure building module 100 is used to establish a connection channel with the user's mobile device and create points using point creation technology;

[0048] The point structure construction module 100 stores multiple account passwords with different usage permissions. When the user establishes a connection channel with the point structure construction module 100, the mobile device sends the login account password to the point structure construction module 100 through the communication network. The point structure construction module 100 senses the login account password sent by the user's mobile device, and compares the login account password with the account password stored in itself one by one. If the login account password is exactly the same as the account password stored in the point structure construction module 100, the point structure construction module 100 sends a connection success response to the mobile device through the communication network, thereby establishing a connection channel. If the login account password is not exactly the same as the account password stored in the point structure construction module 100, the point structure construction module 100 sends a connection failure response to the mobile device through the communication network and refuses to establish a connection channel.

[0049] The corresponding expressions are as follows:

[0050] The account name received is A input The received password is encrypted to P input The i-th account name stored in the point structure building module 100 is A i , the corresponding encryption password is P i , where i = 1, 2, ..., n, n is the number of stored account password groups;

[0051] Define the matching function Match(A input ,P input ,A i ,P i )as follows:

[0052]

[0053] The formula for determining whether the account password is correct is:

[0054] When Correct=1, it means that the account password is correct, and the point architecture building module 100 establishes a connection channel with the user's mobile device; when Correct=0, it means that the account password is incorrect, and the point architecture building module 100 rejects the connection.

[0055] Point creation technology in the point architecture building module 100: Setting point creation rules: When creating points, it is necessary to create areas before creating routes, and then create points after creating routes. This sequence of creating areas, routes, and then points ensures that the point creation process in the point architecture building module 100 has a clear logical order, just like building a house with an overall plan (area), then a frame structure (route), and finally specific rooms (points). This helps users operate in an orderly manner and reduces errors and confusion.

[0056] Each point has a corresponding point name, which is the attribute of the place where the equipment is located, such as supermarket, shopping mall, etc. When users select products in the product information library, they can select products in the product information library according to the point name, avoiding blind search. Since the point name reflects the attributes of the place, for example, at the "supermarket" point, users can quickly find products suitable for supermarket sales, and at the "shopping mall" point they can find products that are more in line with the shopping mall's positioning, which improves the efficiency and accuracy of product selection, saves user time, and improves user experience.

[0057] After the point architecture building module 100 successfully establishes a connection channel with the mobile device, the point architecture building module 100 senses the creation area, line and point request sent by the mobile device through the connection channel (the creation area, line and point request includes the area name, line name, point name, geographical location, surrounding environment, location, etc.) and the device SN code;

[0058] After receiving the request to create areas, lines and points, the point architecture building module 100 compares whether the request information is complete (for example, confirming whether the area information contains the necessary name and approximate range, whether the line information is associated with a valid area, whether the point information contains an accurate geographic location, etc.). If the request information is incomplete, the point architecture building module 100 outputs a supplementary completion signal to the mobile device.

[0059] The expression for comparing whether the request information is complete is as follows:

[0060] In the perception point architecture building module 100, creating areas, lines, and points requires m1, m2, and m3 mandatory information fields. rj (j=1,2,…,m1) indicates the required information field of the region, F lk(k=1,2,…,m2) represents the required information field of the line, F ps (s=1,2,…,m3) indicates the required information field of the point;

[0061] Define the field validity judgment function Valid(F) as follows:

[0062]

[0063] The regional information integrity judgment formula is:

[0064] The formula for judging line information integrity is:

[0065] The formula for judging the integrity of point information is:

[0066] When Validity r =1, Validity l =1 and Validity p =1, the request information is complete, and the point architecture building module 100 creates the point.

[0067] When the requested information is complete, the point architecture building module 100 stores the received area, line and point information in sequence according to the point creation rules. During the storage process, a hierarchical association relationship is established between the area, line and point, and the device SN code is associated with the point, so that during the subsequent use of the device, the user can obtain data such as the device's usage and operating status by logging in with the account and password.

[0068] In existing vending machines, after users create locations and associate devices with them, a product information database is created to meet operational needs. This database stores a large number of products and product materials, including images, selling prices, costs, and specifications. Without a clear mechanism to prioritize different products, users will waste a significant amount of time selecting products, reducing operational efficiency and impacting user experience.

[0069] The commodity intelligent matching adjustment module 200 uses a web crawler to collect the names of places with different attributes and the names of the required commodities, and combines them into a set of required commodity names; uses a string matching method to compare the collected place names with the location names to determine the required commodity names at the location;

[0070] The device senses the products and product names in the product information database. Before selecting a product, the user matches the product in the product information database with the required product name set (using string matching) and marks the corresponding product. When selecting a product, the marked product is prioritized.

[0071] For users of the vending platform, the product intelligent matching and adjustment module 200 can quickly see the products that meet the needs of the location and prioritize them when selecting products for the location equipment, which greatly saves product selection time, avoids blind searching among massive products, and provides a convenient and efficient product selection experience.

[0072] The working principle of the web crawler in the product intelligent matching and adjustment module 200 is to simulate the browser behavior and automatically crawl information on the web page according to the rules. The specific steps are as follows:

[0073] Step 1: Use various commercial websites and industry forums as the starting URL of the web crawler. The web crawler sends HTTP requests to the servers of various commercial websites and industry forums to request the web page content corresponding to the specified URL;

[0074] Step 2: After receiving the request, the servers of various commercial websites and industry forums find the corresponding web page according to the requested URL and send the web page content in HTML format as a response to the crawler program;

[0075] Step 3: After receiving the web page content, the crawler program uses an HTML parser to parse the page and extract the names of places with different attributes and the names of the products required for places with different attributes.

[0076] Because consumers in places with different attributes have different product demands, web crawlers can collect the product names needed by places with different attributes, accurately identifying the product demand characteristics of each place. For example, schools may have a greater demand for products such as stationery, books, and snacks, while hospitals may have a higher demand for medicines, health products, and convenience foods. This lays the foundation for providing accurate product recommendations based on the attributes of the location.

[0077] Different attribute places in the commodity intelligent matching adjustment module 200 are, for example, supermarket places, commercial pedestrian streets, hospitals, etc.; required commodity names are, for example, dairy products, men's clothing, medicines, etc.;

[0078] Specifically, the string matching method in the commodity intelligent matching adjustment module 200 is as follows: starting from the first character of the point name string, it is compared with the first character of the venue name string;

[0079] If they are equal, continue to compare the next character, and so on, until all characters of the place name string are matched successfully. If all characters of the place name string can find corresponding equal characters in the point name string in sequence, then the place corresponding to the collection place name is determined to be the place of the point. The corresponding expression is as follows:

[0080] Divide into point name character strings according to the point name: S = s1s2...s n , where s n Indicates the nth character in the location name and divides the location name string into T = t1t2...t n , t n Indicates the mth character in the point name, and n≥m;

[0081] Start from the first character s1 of S and compare it with the first character t1 of T. If s1=t1, continue to compare the next pair of characters; if s1≠t1, start from the second character s2 of S and compare it with t1, and so on;

[0082] If the i-th character s from S i Start comparing with the first character t1 of T (1≤i≤n-m+1), and then compare s in sequence i+j With t j+1 (j=0,1,…,m-1);

[0083] For all j = 0, 1, ..., m-1, there is s i+j =t j+1 , then it is determined that a substring that completely matches the place name string T is found starting from the i-th position in the point name string S, that is, the place corresponding to the collected place name is determined to be the place of the point;

[0084] If there exists some j (0≤j≤m-1) such that s i+j ≠t j+1 , then it is determined that starting from the i-th position of S, it is not possible to completely match T. At this time, it is necessary to continue to repeat the above comparison process starting from the i+1-th position of S;

[0085] After the above comparison process is performed on all possible starting positions i (1≤i≤n-m+1), if no complete match is found, it is determined that the place corresponding to the collection place name is not the place of the point.

[0086] Furthermore, the method for prioritizing the marked products in the product intelligent matching adjustment module 200 is as follows: giving the marked products a priority higher than the unmarked products, for example, i Assign priority p i (p i =1), unmarked product a m Assign priority p j (p j =0);

[0087] Because p i >p j , then the product a will be marked i Ranked in unmarked product am Before;

[0088] If two marked products appear, they are sorted randomly;

[0089] The product intelligent matching and adjustment module 200 uses a string matching method to determine the location of the point. When the user selects products for the point device, it can automatically filter out products that meet the requirements of the point location. The required products will be displayed to the user first, reducing the user's time and energy in screening through massive products, improving product selection efficiency, and providing users with a better product selection experience.

[0090] Example 2

[0091] In this embodiment, the user can independently input the required products and their corresponding product information through the connection channel established between the mobile device and the point architecture building module 100;

[0092] In actual operation, the product information uploaded by the user usually includes multiple product images. However, when the user selects the first image of the product from these product images, due to personal cognitive differences, vague product image selection standards, etc., inaccurate selection will occur, affecting the product display effect and the user's first impression of the product. Therefore, the product image commonality analysis module 300 senses that the user enters product information for point A through a mobile device, retrieves the product image set A in the product information and the product images corresponding to point A in the product information library, combines them into a product image set B, analyzes the commonalities of the product images in the product image set B, and retrieves the product images that meet the commonalities in the product image set A as the first image of the product;

[0093] Because the product images in product image set B are from the product information database and have been screened and processed to conform to the overall style of the platform and the general cognitive habits of users, selecting the first image by analyzing the commonalities of set B can better integrate the newly entered product images into the platform's product display system, conform to the visual habits and psychological expectations of users when browsing products, and make it easier for users to accept and recognize new products;

[0094] The product image commonality analysis module 300 analyzes the commonalities of the product images using a color histogram. The color histogram divides the color space of the product image into several intervals at equal intervals and counts the frequency of color occurrence in each interval to describe the color distribution of the product image.

[0095] If different product images have similar color composition and distribution, their color histograms will also be similar. By comparing the color histograms of different product images, we can find the commonalities among the product images in the product image set B.

[0096] The calculation formula is as follows:

[0097] If the color space of the product image is RGB color space, the value range of each color channel is 0-255, and each channel is divided into n intervals at equal intervals, then the entire color space is divided into n intervals. 3 For each pixel in the product image, determine the interval it belongs to based on its RGB value, and then add 1 to the interval count;

[0098] Let H(i,j,k) denote the counts of the i-th red channel bin, the j-th green channel bin, and the k-th blue channel bin in the color histogram, where i,j,k=0,1,…,n-1. For each pixel (r,g,b) in the product image, the corresponding bin index is calculated as:

[0099] in Represents the floor function, and then add 1 to the corresponding H(i,j,k), that is, H(i,j,k)=H(i,j,k)+1;

[0100] After traversing all pixels in the product image, H(i, j, k) represents the number of times the color combination appears in the product image. In order to facilitate the comparison of color histograms of different product images, the color histogram is normalized, that is, the count of each interval is divided by the total number of pixels N in the product image to obtain the normalized color histogram

[0101] The product image commonality analysis module 300 senses the color histogram of each product in product image set A, sequentially calculates the similarity between each color histogram and the color histogram of product image set B, and selects the product image with the highest similarity as the first product image to be entered;

[0102]

[0103] Where i is the similarity between the color histogram of each product image in product image set A and the color histograms of all product images in product image set B, P is the normalized color histogram of all product images in product image set B, and Q is the normalized color histogram of a single product image in product image set A.

[0104] Because different properties require different products in places, the product intelligent matching and adjustment module 200 can determine the location of the device based on the name of the location where the device is located. Therefore, in order to avoid the user being unable to select the first picture of the product when entering product information into the product information library, the product image commonality analysis module 300 calculates the commonality of the histogram of products required for the location, and then determines the first picture of the product entered by the user based on the commonality. For example, at the school location, the product image commonality analysis module 300 analyzes the commonality of the product images required by the school (possibly brighter colors, more lively styles, etc.), thereby selecting a more suitable first picture for the newly entered product, so that not only the product itself is more in line with the needs of users at this location, but the display effect of the product can also better attract users at this location, further improving the adaptability and sales potential of the product.

[0105] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart operation platform for vending machines, characterized by: It includes a point structure building module (100), a commodity intelligent matching adjustment module (200) and a commodity image commonality analysis module (300), wherein: The commodity intelligent matching adjustment module (200) collects the names of places with different attributes and the names of required commodities, associates the place attributes with the required commodity names, and combines the commodity names into a set of required commodity names; matches the point names in the set of required commodity names corresponding to the point place names; marks the corresponding commodities, and prioritizes the marked commodities when selecting commodities; The commodity image commonality analysis module (300) senses the point name A and enters the commodity information, calls out the commodity image set A in the commodity information, senses the commodity images of the point name A in the commodity information library, combines them into a commodity image set B, analyzes the commonality of the commodity images in the commodity image set B, calls out the commodity images that meet the commonality in the commodity image set A, and uses them as the first image of the entered commodity.

2. The vending machine intelligent operation platform according to claim 1, characterized in that: The point structure building module (100) is used to establish a connection channel with a user's mobile device and to create points using a point creation technology; The point structure construction module (100) stores a plurality of account passwords with different usage permissions. When a user establishes a connection channel with the point structure construction module (100), the mobile device sends the login account password to the point structure construction module (100) via a communication network. The point structure construction module (100) senses the login account password sent by the user's mobile device and compares the login account password with the account password stored in the point structure construction module (100) one by one. If the login account password is exactly the same as the account password stored in the point structure construction module (100), the point structure construction module (100) sends a connection success response to the mobile device via the communication network to establish a connection channel. If the login account password is not exactly the same as the account password stored in the point structure construction module (100), the point structure construction module (100) sends a connection failure response to the mobile device via the communication network and refuses to establish a connection channel.

3. The vending machine intelligent operation platform according to claim 2, characterized in that: The point creation technology in the point architecture building module (100) sets point creation rules: when creating a point, a line can be created only after an area is created, and a point can be created only after a line is created; Each point has a corresponding point name, which is the attribute of the location of the equipment.

4. The vending machine intelligent operation platform according to claim 3, characterized in that: After the point structure construction module (100) successfully establishes a connection channel with the mobile device, the point structure construction module (100) senses the creation area, line and point request and the device SN code sent by the mobile device through the connection channel; After receiving the request to create an area, a line, and a point, the point architecture building module (100) compares the request information to see if it is complete. If the request information is incomplete, the point architecture building module (100) outputs a complete signal to the mobile device.

5. The vending machine intelligent operation platform according to claim 4 is characterized by: The expression for comparing whether the request information is complete by the point structure building module (100) is as follows: Perceive that the creation of regions, lines and points in the point structure building module (100) requires m1, m2, and m3 mandatory information fields, F rj (j=1,2,…,m1) indicates the required information field of the region, F lk (k=1,2,…,m2) represents the required information field of the line, F ps (s=1,2,…,m3) indicates the required information field of the point; Define the field validity judgment function Valid(F) as follows: The regional information integrity judgment formula is: The formula for judging line information integrity is: The formula for judging the integrity of point information is: When Validity r =1, Validity l =1 and Validity p =1, the request information is complete, and the point structure building module (100) creates the point.

6. The vending machine intelligent operation platform according to claim 1, characterized in that: The names of places with different attributes and the names of required products in the product intelligent matching adjustment module (200) are collected by a web crawler. The web crawler simulates the behavior of a browser and automatically captures information on a web page according to the rules. The specific steps are as follows: Step 1: Use various commercial websites and industry forums as the starting URL of the web crawler. The web crawler sends HTTP requests to the servers of various commercial websites and industry forums to request the web page content corresponding to the specified URL; Step 2: After receiving the request, the servers of various commercial websites and industry forums find the corresponding web page according to the requested URL and send the web page content in HTML format as a response to the crawler program; Step 3: After receiving the web page content, the crawler program uses the HTML parser to parse the page and extract the names of places with different attributes and the names of the required products.

7. The vending machine intelligent operation platform according to claim 6, characterized in that: The string matching method in the commodity intelligent matching adjustment module (200) is as follows: starting from the first character of the point name string, the first character of the place name string is compared; If they are equal, continue to compare the next character, and so on, until all characters of the place name string are matched successfully. If all characters of the place name string can find corresponding equal characters in the point name string in sequence, then the place where the place name is collected is determined to be the place where the point is located.

8. The vending machine intelligent operation platform according to claim 7, characterized in that: The commodity intelligent matching adjustment module (200) senses the desired commodity names in different locations, matches the desired commodity names with the commodity names in the commodity information database again through a string matching method, retrieves the matching commodity names in the commodity information database, marks the compared commodities, and prioritizes the marked commodities when selecting commodities; The priority sorting method is as follows: give the marked product a priority higher than the unmarked product. i Assign priority p i (p i =1), unmarked product a m Assign priority p j (p j =0); If two marked products appear, they will be sorted randomly.

9. The vending machine intelligent operation platform according to claim 5, characterized in that: The commodity image commonality analysis module (300) analyzes the commonality of commodity images by color histogram analysis. The color histogram divides the color space of the commodity image into a number of intervals at equal intervals, and counts the frequency of color occurrence in each interval to describe the color distribution of the commodity image. If different product images are similar in color composition and distribution, the color histograms of the product images will also be similar. By comparing the color histograms of different product images, the commonalities of the product images in the product image set B can be found.

10. The vending machine intelligent operation platform according to claim 9, characterized in that: The commodity image commonality analysis module (300) senses the color histogram of each commodity in the commodity image set A, sequentially calculates the similarity between each color histogram and the color histogram of the commodity image set B, and retrieves the commodity image with the highest similarity as the first image to be entered into the commodity.

Images