Data analysis system based on artificial intelligence

Through the data analysis system based on artificial intelligence and the U-shaped conveyor belt detection agency, the inaccuracy problem of food selection and quality inspection of semi-finished dishes is solved, intelligent selection and quality inspection of food is realized, and sales adaptability and management efficiency are improved. , reduces preparation waste and cost.

CN118552006BActive Publication Date: 2024-11-05XUANMING INTELLIGENT TECHNOLOGY (XIAMEN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411008519.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-05
Estimated Expiration
2044-07-26

AI Technical Summary

Technical Problem

In the prior art, the selection and quality inspection of semi-finished dishes rely on manual operations, and the quantity and equipment are inaccurate, subjective and cannot be fully covered, resulting in uneven combinations, affecting sales, and high preparation waste and management costs.

Method used

Adopting a data analysis system based on artificial intelligence, the semi-finished dish statistics module, the demand semi-finished dish determination module, the demand semi-finished dish preparation quantity estimate module, the demand semi-finished dish matching transmission module, the food partition detection module, and the food combination judgment module and abnormal food ingredients return module to realize intelligent selection and quality inspection of ingredients, combined with the U-shaped conveyor belt inspection mechanism, ensure the accuracy and unity of food ingredients and the objective and accuracy of quality inspection.

Benefits of technology

It improves the sales adaptability of semi-finished dishes, reduces preparation waste, reduces management costs, ensures the accuracy of ingredients combinations and the comprehensiveness of quality inspection, and promotes the sales of semi-finished dishes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118552006B_ABST
    Figure CN118552006B_ABST
Patent Text Reader

Abstract

The present invention belongs to the technical field of food detection data analysis and management, and specifically discloses a data analysis system based on artificial intelligence, which is mainly used for data analysis and management of semi-finished dish detection. The present invention obtains the current season, and then uses supermarket supply records to determine the semi-finished dish demand corresponding to the current season, and estimates the preparation quantity corresponding to each semi-finished dish demand in combination with the current inventory quantity corresponding to various ingredients, thereby realizing on-demand preparation of semi-finished dishes based on consumer purchasing needs, and classifies and places ingredient clusters on various U-shaped conveyor belts for weight detection and quality detection by detection terminals, and then performs combination judgment on ingredient clusters according to the detection results, thereby realizing intelligent selection and quality detection of ingredients in semi-finished dishes, and effectively avoiding the phenomenon of uneven combinations of semi-finished dishes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of food detection data analysis and management, and in particular relates to semi-finished dish detection and management technology, specifically to a data analysis system based on artificial intelligence. Background Art

[0002] Busy modern lifestyles have prompted consumers to seek convenient cooking solutions. To meet this demand, semi-finished meals have come into being. Semi-finished meals supplied by supermarkets offer pre-processed and partially cooked options, providing convenience for busy families by saving a lot of preparation and cooking time.

[0003] Since semi-finished dishes are generally composed of a variety of ingredients in a certain proportion, they need to be selected in quantity during the preparation process. At the same time, the ingredients of semi-finished Chinese dishes need to be washed and cut multiple times. Each processing will introduce bacteria or contaminants, which can easily lead to quality problems. Therefore, semi-finished dishes need to be classified, selected and quality tested during the preparation process.

[0004] In the preparation process of semi-finished dishes, the classification and selection of ingredients and data analysis of quality inspection are particularly important. The classification and selection of ingredients and data analysis of quality inspection play an important role in ensuring the quality, safety, cost control and production efficiency of semi-finished dishes, and are an indispensable part of the modern food manufacturing industry.

[0005] However, in the prior art, the selection and quality inspection of combined ingredients in semi-finished dishes mostly rely on manual operations. First, when combined ingredients are manually selected, there is a problem of deviation between the amount of ingredients selected and the amount of ingredients distributed, which makes it impossible to accurately and uniformly allocate the amount of combined ingredients. Second, when combined ingredients are manually inspected for quality, they may be subjective and uncertain because they rely on manual visual inspection or manual sampling, and they cannot fully cover each ingredient. It can be seen that excessive reliance on manual labor for the preparation of semi-finished dishes can easily lead to uneven combinations of semi-finished dishes, which to a certain extent affects the preparation and presentation of semi-finished dishes and is not conducive to the sales of semi-finished dishes.

[0006] In addition, since the semi-finished dishes supplied in supermarkets are not single, the types and preparation quantities of semi-finished dishes need to be determined in advance during the preparation process. However, most of them are currently determined only based on the types and storage quantities of ingredients in stock, without taking consumers' purchasing needs into consideration. This can easily lead to the prepared semi-finished dishes being incompatible with consumers' purchasing needs, which, on the one hand, affects the sales revenue of semi-finished dishes, and on the other hand causes waste in the preparation of semi-finished dishes, which invisibly increases the management cost of the preparation of semi-finished dishes. Summary of the invention

[0007] In view of this, in order to solve the problems raised in the above background technology, a data analysis system based on artificial intelligence is proposed.

[0008] The purpose of the present invention can be achieved through the following technical solutions: an artificial intelligence-based data analysis system, which is mainly used for data analysis and management of semi-finished dish detection, including: a semi-finished dish statistics module, which is used to count the number of semi-finished dishes provided by the supermarket.

[0009] The module for determining the semi-finished dishes in demand is used to obtain the current season and determine the semi-finished dishes in demand corresponding to the current season from the supermarket sales records.

[0010] The module for estimating the preparation quantity of semi-finished dishes on demand is used to obtain the single preparation supply days corresponding to each semi-finished dish on demand and the ingredient combination characteristics of each semi-finished dish on demand, and to estimate the preparation quantity corresponding to each semi-finished dish on demand based on the current inventory of various ingredients.

[0011] The on-demand semi-finished dishes matching and conveying module is used to set up a detection mechanism consisting of a combination area and several U-shaped conveyor belts, and transmit the on-demand semi-finished dishes corresponding to the current season, the ingredient combination characteristics and preparation quantities corresponding to each on-demand semi-finished dish to the matching end, and the operator sorts the on-demand semi-finished dishes corresponding to the current season, and selects ingredients according to the sorting to form ingredient clusters, and then classifies the ingredient clusters and places them on the incoming sections of each U-shaped conveyor belt, wherein the U-shaped conveyor belt consists of an incoming section, a transfer placement area and a return section.

[0012] The reference management library is used to store normal freshness-related indicators corresponding to various ingredients.

[0013] The food partition detection module is used to select the detection position on the incoming section of each U-shaped conveyor belt and install the detection terminal at the detection position to collect the appearance image and weight of the food cluster transmitted to the detection position, and analyze the selection compliance of the food cluster.

[0014] The ingredient combination judgment module is used to analyze the selection conformity of the corresponding ingredient clusters transmitted to the corresponding transfer placement areas of each U-shaped conveyor belt in real time according to the current matching requirements of the semi-finished dishes, and then judge whether they can be combined.

[0015] The food combination module is used to transfer the corresponding food clusters transmitted to the corresponding transfer placement areas of each U-shaped conveyor belt to the combination area when it is determined that they can be combined, so as to combine them into semi-finished dishes.

[0016] The abnormal food returning module is used to identify the abnormal food cluster and abnormal direction when it is judged that the combination cannot be performed, and to transfer the abnormal food cluster to the return section through the transfer placement area of ​​the conveyor belt where it is located, and at the same time convey the abnormal direction to the matching end.

[0017] In an alternative embodiment, the specific implementation process of determining the demand for semi-finished dishes corresponding to the current season from the supermarket sales records is as follows: select several historical years, and then extract the sales volume of each semi-finished dish in the current season in each historical year from the supermarket sales records, and classify the sales volume of the same semi-finished dish to form a number of sales volume data for each semi-finished dish in the current season.

[0018] Outliers are identified on several sales volume data of each semi-finished dish in the current season, and outliers are eliminated from several sales volume data. The sales volume data of each semi-finished dish in the current season that are retained are averaged to obtain the effective sales volume of each semi-finished dish in the current season.

[0019] Arrange the semi-finished dishes in descending order according to their effective sales volume, and select the semi-finished dishes ranked at or before the preferred sort number from the arrangement results according to the set preferred sort number as the semi-finished dishes in demand corresponding to the current season.

[0020] In an alternative embodiment, the ingredient combination characteristics include the types of combined ingredients and the corresponding allocation amounts of each combined ingredient.

[0021] In an alternative embodiment, the preparation quantity corresponding to each semi-finished dish on demand refers to the following estimation process: S1, counting the supply days corresponding to the current season, and combining the effective sales volume of each semi-finished dish on demand in the current season, the single preparation supply days corresponding to each semi-finished dish on demand and the characteristics of the ingredient combination using the expression Calculate the required inventory of each combination of ingredients for each required semi-finished dish , where Expressed as The effective sales volume of semi-finished dishes in the current season, Indicates the number of the semi-finished dish required. , Expressed as the number of supply days corresponding to the current season, Expressed as The number of days for a single preparation of the semi-finished dishes required, Expressed as Request for semi-finished dishes The corresponding distribution amount of combined ingredients, Indicates the combination ingredient number. , Indicates rounding up.

[0022] S2. Compare the types of combination ingredients corresponding to each required semi-finished dish, and then add up the required inventory of the same combination ingredients in each required semi-finished dish to obtain the total required inventory of each combination ingredient.

[0023] S3, based on the type of combined ingredients, filter out the current inventory corresponding to each combined ingredient from the current inventory corresponding to each ingredient, and compare it with the total required inventory of the corresponding combined ingredients, using the expression Count the inventory satisfaction of each combination of ingredients. Represents a natural constant.

[0024] S4. Compare the inventory satisfaction of each combination of ingredients with the set standard inventory satisfaction. If the inventory satisfaction of all combination ingredients is greater than or equal to the set standard inventory satisfaction, Get the preparation quantity corresponding to each required semi-finished dish, otherwise it will be weakened according to the set ratio Calculate the reduced inventory demand for each combination of ingredients for each required semi-finished dish , and then calculate the inventory satisfaction of each combination of ingredients in the weakened state based on S2 and S3, and then compare it with the set standard inventory satisfaction. If the inventory satisfaction of each combination of ingredients in the weakened state is greater than or equal to the set standard inventory satisfaction, then use the expression The preparation quantity corresponding to each required semi-finished dish is calculated, and if it is insufficient, it is weakened twice until the inventory satisfaction is greater than or equal to the set standard inventory satisfaction.

[0025] In an alternative embodiment, the transfer placement area of ​​the U-shaped conveyor belt is connected to the incoming section and the return section respectively, and the transfer placement area of ​​each U-shaped conveyor belt is connected to the combination area.

[0026] In an alternative embodiment, the specific analysis process of the selection compliance of the ingredient cluster is: extracting freshness-related indicators from the appearance image of the ingredient cluster, and comparing them with the normal freshness-related indicators corresponding to the corresponding ingredients in the reference management library, and then importing the quality analysis algorithm corresponding to each freshness-related indicator in the quality analysis model to obtain the quality analysis value corresponding to each freshness-related indicator.

[0027] The quality analysis value corresponding to each freshness-related indicator is combined with the pre-configured weight factor corresponding to each freshness-related indicator into the formula , and calculate the quality coefficient of the food cluster.

[0028] Compare the weight of the ingredient cluster with the amount of the corresponding combination of ingredients in the current required semi-finished dish, and use the expression , calculate the weight coefficient of the food cluster, where Expressed as the preset allowable difference.

[0029] The quality coefficient of the ingredient cluster is compared with the qualified quality coefficients corresponding to the various pre-prepared ingredients. If the quality coefficient of the ingredient cluster is greater than or equal to the qualified quality coefficient corresponding to the corresponding ingredient, and the weight coefficient of the ingredient cluster is 1, the selection conformity of the ingredient cluster is set to 1, otherwise the selection conformity of the ingredient cluster is set to 0.

[0030] In an alternative embodiment, the method of judging whether a combination is possible is as follows: comparing the selection conformity of the corresponding food clusters on the transfer and placement areas corresponding to each U-shaped conveyor belt with the standard value; if the quality coefficients of the corresponding food clusters on the transfer and placement areas corresponding to each U-shaped conveyor belt are all standard values, then it is judged that a combination is possible; otherwise, it is judged that a combination is not possible.

[0031] In an alternative embodiment, the identification process of the abnormal ingredient cluster and the abnormal direction is as follows: the ingredient cluster whose selected compliance does not meet the standard value is recorded as an abnormal ingredient cluster, and the quality coefficient of the abnormal ingredient cluster is compared with the standard quality coefficient corresponding to the corresponding ingredient. If the quality coefficient of the abnormal ingredient cluster is less than the standard quality coefficient corresponding to the corresponding ingredient, the abnormal direction is that the quality does not meet the standard. At the same time, the weight coefficient of the abnormal ingredient cluster is obtained. If the weight coefficient of the abnormal ingredient cluster is 0, the abnormal direction is that the weight does not meet the standard.

[0032] In a replaceable embodiment, it also includes a detection mechanism start and stop intelligent control module, which is used to set a voice announcer on the combination area, and count the number of semi-finished dishes in the combination area in real time, and compare it with the preparation amount corresponding to the semi-finished dishes currently required for matching. If the number of semi-finished dishes in the combination area reaches the preparation amount corresponding to the semi-finished dishes currently required for matching, the detection mechanism is controlled to suspend operation, and a voice announcer is used to remind the user to pick up the dishes. Otherwise, the detection mechanism is controlled to continue running, and the number of remaining semi-finished dishes to be matched is counted and transmitted to the matching end.

[0033] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention obtains the current season and uses the historical sales volume of each semi-finished dish in the current season in the supermarket sales record to determine the demand semi-finished dishes corresponding to the current season, and at the same time combines the current inventory of various ingredients to estimate the preparation quantity corresponding to each demand semi-finished dish, thereby realizing on-demand preparation of semi-finished dishes based on consumer purchasing needs, making the prepared semi-finished dishes more reasonable and more adaptable. On the one hand, it increases the sales revenue of semi-finished dishes, and on the other hand, it greatly avoids the preparation waste of semi-finished dishes, which is conducive to reducing the preparation management cost of semi-finished dishes.

[0034] (2) The present invention sets up a detection mechanism consisting of a combination area and a plurality of U-shaped conveyor belts. During the preparation process of semi-finished dishes, operators select ingredients according to the required semi-finished dishes to form ingredient clusters, and classify the ingredient clusters and place them on each U-shaped conveyor belt for weight detection and quality detection by the detection terminal. Then, the ingredient clusters are combined and judged according to the detection results, thereby realizing intelligent selection and quality detection of ingredients in semi-finished dishes, greatly reducing human intervention, and making the quantity of combined ingredients accurate and uniform. At the same time, it also makes the quality detection of ingredients more objective and accurate and can achieve comprehensive coverage, effectively avoiding the phenomenon of inconsistent combinations of semi-finished dishes, and greatly promoting the sales of semi-finished dishes.

[0035] (3) When the present invention uses the detection mechanism to perform weight detection and quality detection on the food clusters on each U-shaped conveyor belt, when it is identified that the food clusters cannot be combined, the abnormal food clusters and abnormal directions are identified, and the abnormal food clusters are returned and conveyed using the return section of the U-shaped conveyor belt, so as to avoid the abnormal food clusters continuing to flow to the combination area and causing interference and pollution to the semi-finished dish combination. At the same time, the identification of abnormal directions also provides a clear direction for the processing of abnormal food clusters, which is conducive to speeding up the preparation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0037] Figure 1 It is a schematic diagram of the connection of each module of the system of the present invention.

[0038] Figure 2 It is a structural schematic diagram of the detection mechanism in the present invention.

[0039] Figure numerals: 1 - incoming section, 2 - return section, 3 - transfer placement area, 4 - combination area, A - conveying direction of the incoming section, B - conveying direction of the return section. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0041] See also Figure 1As shown, the present invention proposes a data analysis system based on artificial intelligence, which is mainly used for data analysis and management of semi-finished dish detection, including a semi-finished dish statistics module, a semi-finished dish demand determination module, a semi-finished dish preparation quantity estimation module, a semi-finished dish matching and transmission module, a reference management library, a food partition detection module, a food combination judgment module, a food combination module, an abnormal food return module and a detection mechanism start-stop intelligent control module, wherein the semi-finished dish statistics module is connected to the semi-finished dish demand determination module, the semi-finished dish demand determination module is connected to the semi-finished dish preparation quantity estimation module, the semi-finished dish demand determination module and the semi-finished dish preparation quantity estimation module are both connected to the semi-finished dish matching and transmission module, the semi-finished dish matching and transmission module is connected to the food partition detection module, the food partition detection module is connected to the food combination judgment module, the food combination judgment module is respectively connected to the food combination module and the abnormal food return module, the food combination module and the semi-finished dish preparation quantity estimation module are both connected to the detection mechanism start-stop intelligent control module, and the reference management library is connected to the food partition detection module.

[0042] The semi-finished dish statistics module is used to count the number of semi-finished dishes provided by the supermarket.

[0043] It should be noted that the semi-finished dishes mentioned in the present invention refer to the types of semi-finished dishes.

[0044] The module for determining the semi-finished dishes in demand is used to obtain the current season, and based on this, determine the semi-finished dishes in demand corresponding to the current season from the supermarket sales records. The specific implementation process is as follows: select several historical years, and then extract the sales volume of each semi-finished dish in the current season in each historical year from the supermarket sales records, and classify the sales volume of the same semi-finished dish to form a number of sales volume data of each semi-finished dish in the current season.

[0045] It should be added that in order to ensure that the extracted sales data of semi-finished dishes is not too small and to provide reliable data support for the analysis of effective sales, the number of historical years selected should not be too small, and can be no less than 5. At the same time, the historical years should be selected nearby based on the current year. For example, if the current year is 2022, the selected historical years can be 2021, 2020, 2019, 2018, and 2017.

[0046] Outliers are identified on several sales volume data of each semi-finished dish in the current season, and outliers are eliminated from several sales volume data. The sales volume data of each semi-finished dish in the current season that are retained are averaged to obtain the effective sales volume of each semi-finished dish in the current season.

[0047] Applied to the above embodiment, the reason for identifying outliers in the sales volume data is that due to external factors such as data entry errors and statistical deviations, the extracted sales volume data may deviate from the normal state. In order to ensure the quality and accuracy of the extracted sales volume data, it is necessary to identify outliers.

[0048] Further identification of outliers in sales data can be done in the following way: cluster the sales data of each semi-finished dish in the current season using a clustering algorithm, and then count the sales data of each semi-finished dish in each cluster from the division results, and compare them with each other to obtain the maximum and minimum sales data, thereby using the expression The difference in sales data corresponding to each semi-finished dish is calculated. The greater the difference between the maximum sales data and the minimum sales data, the greater the probability of an outlier.

[0049] The difference in the sales data corresponding to each semi-finished dish is compared with the set allowable difference. If the difference in the sales data corresponding to a semi-finished dish is greater than the set allowable difference, the cluster corresponding to the minimum sales data is selected, and the sales data in the cluster is taken as an outlier. Otherwise, the distance between each sales data in each cluster corresponding to the semi-finished dish and the center of the cluster to which it belongs is obtained as the Euclidean distance of each sales data in the cluster, and compared with the set distance threshold, and then the sales data with a Euclidean distance greater than the distance threshold are selected from each cluster of the semi-finished dish as an outlier.

[0050] The principle of using clustering algorithm to identify outliers in the present invention is that normal data points will form larger clusters, while abnormal data points usually do not form very large clusters, resulting in the existence of small clusters or isolated clusters. At the same time, normal data points will be concentrated in the center of the cluster, while abnormal data points usually appear to be far away from the center of the cluster.

[0051] Arrange the semi-finished dishes in descending order according to their effective sales volume, and select the semi-finished dishes ranked at or before the preferred sort number from the arrangement results according to the set preferred sort number as the semi-finished dishes in demand corresponding to the current season.

[0052] As an example of the above scheme, the arrangement order of the semi-finished dishes is 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, and the set preferred sort number is 4, then the selected semi-finished dishes are 1, 2, 3, 4.

[0053] The module for estimating the preparation quantity of semi-finished dishes on demand is used to obtain the single preparation supply days corresponding to each semi-finished dish on demand and the ingredient combination characteristics of each semi-finished dish on demand, wherein the ingredient combination characteristics include the types of combined ingredients and the corresponding allocations of each combined ingredient, and estimate the preparation quantity corresponding to each semi-finished dish on demand in combination with the current inventory of various ingredients. The specific estimation process is: S1. Counting the supply days corresponding to the current season. The specific statistical method is to obtain the distribution months corresponding to the current season and the days corresponding to each distribution month, and then accumulate the days corresponding to each distribution month to obtain the supply days corresponding to the current season. For example, the distribution months corresponding to the current season are March-May. At this time, the days from March to May are accumulated to obtain the supply days corresponding to the current season, and the expression is used in combination with the effective sales volume of each semi-finished dish on demand in the current season, the single preparation supply days corresponding to each semi-finished dish on demand and the ingredient combination characteristics. Calculate the required inventory of each combination of ingredients for each required semi-finished dish , where Expressed as The effective sales volume of semi-finished dishes in the current season, Indicates the number of the semi-finished dish required. , Expressed as the number of supply days corresponding to the current season, Expressed as The number of days for a single preparation of the semi-finished dishes required, Expressed as Request for semi-finished dishes The corresponding distribution amount of combined ingredients, Indicates the combination ingredient number. , Indicates rounding up.

[0054] It should be understood that the single preparation and supply days corresponding to each semi-finished dish on demand mentioned above are set based on the shelf life of each semi-finished dish on demand. Since semi-finished dishes are usually made with relatively fresh ingredients, the shelf life of these ingredients will decrease to a certain extent after processing and treatment. Therefore, the shelf life of semi-finished dishes is shorter, which means that not a lot of semi-finished dishes will be prepared at one time.

[0055] S2. Compare the types of combination ingredients corresponding to each required semi-finished dish, and then add up the required inventory of the same combination ingredients in each required semi-finished dish to obtain the total required inventory of each combination ingredient.

[0056] S3, based on the type of combined ingredients, filter out the current inventory corresponding to each combined ingredient from the current inventory corresponding to each ingredient, and compare it with the total required inventory of the corresponding combined ingredients, using the expression Count the inventory satisfaction of each combination of ingredients. Represents a natural constant, where the higher the current inventory is than the total required inventory, the greater the inventory satisfaction.

[0057] S4. Compare the inventory satisfaction of each combination of ingredients with the set standard inventory satisfaction. If the inventory satisfaction of all combination ingredients is greater than or equal to the set standard inventory satisfaction, Get the preparation quantity corresponding to each required semi-finished dish. Otherwise, it means that the total required inventory of each combination of ingredients is too large, so the reduction ratio is set according to the set ratio. Calculate the reduced inventory demand for each combination of ingredients for each required semi-finished dish , and then calculate the inventory satisfaction of each combination of ingredients in the weakened state based on S2 and S3, and then compare it with the set standard inventory satisfaction. If the inventory satisfaction of each combination of ingredients in the weakened state is greater than or equal to the set standard inventory satisfaction, then use the expression The preparation quantity corresponding to each required semi-finished dish is calculated, and if it is insufficient, it is weakened twice until the inventory satisfaction is greater than or equal to the set standard inventory satisfaction.

[0058] What you need to know is that the weakening ratio range is , an exemplary weakening ratio is .

[0059] The present invention obtains the current season and determines the demand semi-finished dishes corresponding to the current season by using the historical sales volume of each semi-finished dish in the current season in the supermarket sales record, and at the same time estimates the preparation quantity corresponding to each demand semi-finished dish in combination with the current inventory quantity corresponding to various ingredients, thereby realizing on-demand preparation of semi-finished dishes based on consumer purchasing needs, making the prepared semi-finished dishes more reasonable and more adaptable. On the one hand, it increases the sales revenue of semi-finished dishes, and on the other hand, it greatly avoids the preparation waste of semi-finished dishes, which is conducive to reducing the preparation management cost of semi-finished dishes.

[0060] The on-demand semi-finished dish pairing and conveying module is used to set up a detection mechanism consisting of a combination area (4) and a plurality of U-shaped conveyor belts, and transmits the on-demand semi-finished dishes corresponding to the current season, the food combination characteristics and the preparation quantity corresponding to each on-demand semi-finished dish to the pairing end, and the operator sorts the on-demand semi-finished dishes corresponding to the current season, selects food ingredients according to the sorting, forms food ingredient clusters, and then classifies and places the food ingredient clusters on the incoming sections 1 of each U-shaped conveyor belt, wherein the U-shaped conveyor belt consists of an incoming section 1, a transfer placement area 3 and a return section 2, and the transfer placement area 3 of the U-shaped conveyor belt is respectively connected to the incoming section 1 and the return section 2, and the transfer placement area 3 of each U-shaped conveyor belt is connected to the combination area 4, and at the same time, each U-shaped conveyor belt is arranged in parallel, and the conveying directions of the incoming section 1 and the return section 2 are opposite.

[0061] The structure of the above detection mechanism can be found in Figure 2 shown.

[0062] It should be noted that the above-mentioned ingredient cluster refers to the state formed after the ingredients are selected according to the allocated amounts.

[0063] It should be understood that when the required semi-finished dishes are conveyed, each U-shaped conveyor belt corresponds to the combined ingredients in the required semi-finished dishes one by one, and the corresponding relationship changes with the change of the required semi-finished dishes.

[0064] The reference management library is used to store normal freshness-related indicators corresponding to various food ingredients.

[0065] The food partition detection module is used to select a detection position on the incoming section 1 of each U-shaped conveyor belt, and install a detection terminal at the detection position, wherein the detection terminal includes a high-definition camera and a weight sensor, which performs appearance image acquisition and weight detection on the food cluster transmitted to the detection position, and analyzes the selection compliance of the food cluster.

[0066] The above-mentioned medium and high definition cameras are used to collect appearance images of the food cluster, and the weight sensor is used to detect the weight of the food cluster.

[0067] It should be added that the detection position selected on the input section of the U-shaped conveyor belt should not move with the conveying of the conveyor belt, and can be set on the support frame of the conveyor belt corresponding to the input section.

[0068] Preferably, the specific analysis process of the selection compliance of the ingredient cluster is: extracting freshness-related indicators from the appearance image of the ingredient cluster, specifically, the freshness-related indicators include but are not limited to color, surface texture, surface defects, etc., and comparing them with the normal freshness-related indicators corresponding to the corresponding ingredients in the reference management library, and then importing the quality analysis algorithm corresponding to each freshness-related indicator in the quality analysis model to obtain the quality analysis value corresponding to each freshness-related indicator.

[0069] In particular, the present invention selects color, surface texture, and surface defects as freshness-related indicators of food clusters because food usually has different colors when fresh and stale, and the freshness of food can be intuitively displayed through color contrast. At the same time, fresh food usually has specific texture and surface characteristics. For example, vegetables, fruits, or meat may have a smooth, uniform surface when fresh, while rotten or stale food may have a softened, wrinkled, or sticky texture. In addition, food surface defects are often signs of food decay or deterioration. Therefore, it can be seen that the freshness of food clusters can be clearly determined by analyzing color, surface texture, and surface defects.

[0070] As an example of the above scheme, when the freshness-related indicator is color, the quality analysis algorithm corresponding to color is implemented by converting the color of the food cluster and the normal color of the corresponding food into chromaticity for comparison, and using the algorithm expression Get the quality analysis value corresponding to the color.

[0071] When the freshness-related indicator is surface texture, the quality analysis algorithm corresponding to the surface texture is implemented by overlapping and comparing the surface texture profile of the food cluster with the normal surface texture profile of the corresponding food, and using the algorithm expression Get the quality analysis value corresponding to the surface texture.

[0072] When the freshness-related indicator is a surface defect, the quality analysis algorithm corresponding to the surface defect is implemented by extracting the surface defect type and surface defect area of ​​the food cluster from the appearance image, wherein the surface defect type may specifically be scratches, breakages, spots, etc.

[0073] The surface defect types of the food cluster are matched with the defect influencing factors corresponding to various preset surface defect types, thereby matching the defect influencing factors of the food cluster.

[0074] Import the defect influencing factors of the food cluster into the algorithm expression in combination with the surface defect area , and obtain the quality analysis value corresponding to the surface defect.

[0075] The quality analysis value corresponding to each freshness-related indicator is combined with the pre-configured weight factor corresponding to each freshness-related indicator into the formula , and calculate the quality coefficient of the food cluster.

[0076] Compare the weight of the ingredient cluster with the amount of the corresponding combination of ingredients in the current required semi-finished dish, and use the expression , calculate the weight coefficient of the food cluster, where Expressed as the preset allowable difference.

[0077] The quality coefficient of the ingredient cluster is compared with the qualified quality coefficients corresponding to the various pre-prepared ingredients. If the quality coefficient of the ingredient cluster is greater than or equal to the qualified quality coefficient corresponding to the corresponding ingredient, and the weight coefficient of the ingredient cluster is 1, the selection conformity of the ingredient cluster is set to 1, otherwise the selection conformity of the ingredient cluster is set to 0.

[0078] The food combination judgment module is used to analyze the selection compliance of the corresponding food clusters transmitted to the transfer placement area 3 corresponding to each U-shaped conveyor belt in real time according to the current matching requirements of the semi-finished dishes, and then judge whether they can be combined. The specific judgment method is: compare the selection compliance of the corresponding food clusters on the transfer placement area 3 corresponding to each U-shaped conveyor belt with the standard value, where the standard value is 1. If the quality coefficients of the corresponding food clusters on the transfer placement area 3 corresponding to each U-shaped conveyor belt are all standard values, it is judged that they can be combined, otherwise it is judged that they cannot be combined.

[0079] The food combination module is used to transfer the corresponding food clusters transmitted to the transfer placement area 3 corresponding to each U-shaped conveyor belt to the combination area when it is determined that the food clusters can be combined, so as to be combined into semi-finished dishes.

[0080] The present invention sets a detection mechanism consisting of a combination area and a plurality of U-shaped conveyor belts. In the process of preparing semi-finished dishes, an operator selects ingredients according to the required semi-finished dishes to form ingredient clusters, and classifies the ingredient clusters and places them on each U-shaped conveyor belt for weight detection and quality detection by a detection terminal. Then, a combination judgment is made on the ingredient clusters according to the detection results, thereby realizing intelligent selection and quality detection of ingredients in semi-finished dishes, greatly reducing human intervention, and making the quantity allocation of combined ingredients accurate and unified. At the same time, the quality detection of ingredients is more objective and accurate and can achieve comprehensive coverage, effectively avoiding the phenomenon of inconsistent combinations of semi-finished dishes, and greatly promoting the sales of semi-finished dishes.

[0081] The abnormal food returning module is used to identify abnormal food clusters and abnormal directions when it is determined that they cannot be combined, and to transfer the abnormal food clusters to the return section 2 via the transfer placement area 3 of the conveyor belt where they are located, and at the same time convey the abnormal directions to the matching end.

[0082] Preferably, the identification process of abnormal food clusters and abnormal directions is as follows: the food cluster whose degree of conformity does not meet the standard value is recorded as an abnormal food cluster, and the quality coefficient of the abnormal food cluster is compared with the standard quality coefficient corresponding to the corresponding food. If the quality coefficient of the abnormal food cluster is less than the standard quality coefficient corresponding to the corresponding food, the abnormal direction is that the quality does not meet the standard. At the same time, the weight coefficient of the abnormal food cluster is obtained. If the weight coefficient of the abnormal food cluster is 0, the abnormal direction is that the weight does not meet the standard.

[0083] It is important to know that since the returned abnormal food clusters need to be processed by the operator, the return section must be returned to the matching end. At the same time, when it is judged that the combination is impossible, the U-shaped conveyor belt where the normal food clusters are located is suspended, so that the normal food clusters will not be transferred to the combination area until the normal food clusters are re-transmitted from the transfer placement area of ​​the conveyor belt where the abnormal food clusters are located. The suspended conveyor belt will then be resumed, thereby transferring the normal food clusters on each U-shaped conveyor belt to the combination area.

[0084] When the present invention utilizes the detection mechanism to perform weight detection and quality detection on the food clusters on each U-shaped conveyor belt, when it is identified that they cannot be combined, the abnormal food clusters and abnormal directions are identified, and the abnormal food clusters are returned and conveyed using the return section of the U-shaped conveyor belt, so as to avoid the abnormal food clusters continuing to flow to the combination area and causing interference and pollution to the semi-finished dish combination. At the same time, the identification of abnormal directions also provides a clear direction for the processing of abnormal food clusters, which is conducive to speeding up the preparation process.

[0085] The detection mechanism start-stop intelligent control module is used to set a voice announcer on the combination area, and count the number of semi-finished dishes in the combination area in real time, and compare it with the preparation amount corresponding to the semi-finished dishes currently being matched. If the number of semi-finished dishes in the combination area reaches the preparation amount corresponding to the semi-finished dishes currently being matched, the detection mechanism is controlled to suspend operation, and a voice announcer is used to remind people to pick up dishes. Otherwise, the detection mechanism is controlled to continue running, and at the same time, the preparation amount corresponding to the semi-finished dishes currently being matched is subtracted from the number of semi-finished dishes in the combination area, and the number of remaining semi-finished dishes to be matched is counted and transmitted to the matching end.

[0086] Due to the return of abnormal ingredient clusters, the operator on the matching end has to select an amount of ingredient clusters that exceeds the preparation amount of the required semi-finished dishes, but the operator on the matching end does not know how much it exceeds. By real-time counting of the number of remaining semi-finished dishes to be matched and communicating it, the operator on the matching end can promptly know the subsequent amount of ingredient clusters to be selected, avoiding food waste caused by blind selection.

[0087] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. The data analysis system based on artificial intelligence is characterized by: The system is mainly used for data analysis and management of semi-finished dish detection, including: Semi-finished dishes statistics module, used to count the number of semi-finished dishes provided by the supermarket; The module for determining the semi-finished dishes required is used to obtain the current season and determine the semi-finished dishes required corresponding to the current season from the supermarket sales records; The module for estimating the preparation quantity of semi-finished dishes on demand is used to obtain the single preparation supply days corresponding to each semi-finished dish on demand and the ingredient combination characteristics of each semi-finished dish on demand, and estimate the preparation quantity corresponding to each semi-finished dish on demand based on the current inventory of various ingredients; The on-demand semi-finished dish pairing and conveying module is used to set up a detection mechanism consisting of a combination area (4) and a plurality of U-shaped conveyor belts, and transmit the on-demand semi-finished dishes corresponding to the current season, the ingredient combination characteristics and the preparation quantity corresponding to each on-demand semi-finished dish to the pairing end, and the operator sorts the on-demand semi-finished dishes corresponding to the current season, selects ingredients according to the sorting, forms ingredient clusters, and then classifies and places the ingredient clusters on the incoming sections (1) of each U-shaped conveyor belt, wherein the U-shaped conveyor belt consists of an incoming section (1), a transfer placement area (3) and a return section (2); Reference management library, used to store normal freshness related indicators corresponding to various ingredients; The food partition detection module is used to select a detection position on the incoming section (1) of each U-shaped conveyor belt, and install a detection terminal at the detection position, which performs appearance image acquisition and weight detection on the food clusters transmitted to the detection position, and analyzes the selection compliance of the food clusters; The ingredient combination judgment module is used to analyze the selection conformity of the corresponding ingredient clusters transmitted to the corresponding transfer placement area (3) of each U-shaped conveyor belt in real time according to the current matching demand of the semi-finished dish, and then judge whether they can be combined; An ingredient combination module, for transferring the corresponding ingredient clusters conveyed to the transfer placement areas (3) corresponding to the U-shaped conveyor belts to the combination area (4) when it is determined that the ingredients can be combined, so as to combine them into a semi-finished dish; An abnormal food returning module is used to identify an abnormal food cluster and abnormal direction when it is determined that the food cannot be combined, and to convey the abnormal food cluster to the return section (2) via the transfer placement area (3) of the conveyor belt where the abnormal food cluster is located, and to convey the abnormal direction to the matching end; The preparation quantity corresponding to each required semi-finished dish is shown in the following estimation process: S1. Count the supply days corresponding to the current season, and combine the effective sales volume of each semi-finished dish in the current season, the single preparation supply days corresponding to each semi-finished dish, and the characteristics of the ingredients combination to use the expression Calculate the required inventory of each combination of ingredients for each required semi-finished dish , where Expressed as The effective sales volume of semi-finished dishes in the current season, Indicates the number of the semi-finished dish required. , Expressed as the number of supply days corresponding to the current season, Expressed as The number of days for a single preparation of the semi-finished dishes required, Expressed as Request for semi-finished dishes The corresponding distribution amount of combined ingredients, Indicates the combination ingredient number. , Indicates rounding up; S2. Compare the types of combination ingredients corresponding to each required semi-finished dish, and then add up the required inventory of the same combination ingredients in each required semi-finished dish to obtain the total required inventory of each combination ingredient; S3, based on the type of combined ingredients, filter out the current inventory corresponding to each combined ingredient from the current inventory corresponding to each ingredient, and compare it with the total required inventory of the corresponding combined ingredients, using the expression Count the inventory satisfaction of each combination of ingredients. represents a natural constant; S4. Compare the inventory satisfaction of each combination of ingredients with the set standard inventory satisfaction. If the inventory satisfaction of all combination ingredients is greater than or equal to the set standard inventory satisfaction, Get the preparation quantity corresponding to each required semi-finished dish, otherwise it will be weakened according to the set ratio Calculate the reduced inventory demand for each combination of ingredients for each required semi-finished dish , and then calculate the inventory satisfaction of each combination of ingredients in the weakened state based on S2 and S3, and then compare it with the set standard inventory satisfaction. If the inventory satisfaction of each combination of ingredients in the weakened state is greater than or equal to the set standard inventory satisfaction, then use the expression The preparation quantity corresponding to each required semi-finished dish is calculated, and if it is insufficient, it is weakened twice until the inventory satisfaction is greater than or equal to the set standard inventory satisfaction.

2. The artificial intelligence-based data analysis system according to claim 1, characterized in that: The specific implementation process of determining the semi-finished dishes required for the current season from the supermarket sales records is as follows: Select several historical years, and then extract the sales volume of each semi-finished dish in the current season in each historical year from the supermarket sales records, and classify the sales volume of the same semi-finished dish to form several sales volume data of each semi-finished dish in the current season; Identify outliers on a number of sales volume data of each semi-finished dish in the current season, remove outliers from the sales volume data, calculate the mean of the remaining sales volume data of each semi-finished dish in the current season, and obtain the effective sales volume of each semi-finished dish in the current season; Arrange the semi-finished dishes in descending order according to their effective sales volume, and select the semi-finished dishes ranked at or before the preferred sort number from the arrangement results according to the set preferred sort number as the semi-finished dishes in demand corresponding to the current season.

3. The artificial intelligence-based data analysis system according to claim 1, characterized in that: The food combination characteristics include the types of combined food ingredients and the corresponding distribution amounts of each combined food ingredient.

4. The artificial intelligence-based data analysis system according to claim 1, characterized in that: The transfer placement area (3) of the U-shaped conveyor belt is respectively connected to the incoming section (1) and the return section (2), and the transfer placement area (3) of each U-shaped conveyor belt is connected to the combination area (4).

5. The artificial intelligence-based data analysis system according to claim 1, characterized in that: The specific analysis process of the selection compliance of the food cluster is as follows: Extract freshness-related indicators from the appearance image of the food cluster, and compare them with the normal freshness-related indicators corresponding to the corresponding food in the reference management library, and then import the quality analysis algorithm corresponding to each freshness-related indicator in the quality analysis model to obtain the quality analysis value corresponding to each freshness-related indicator; The quality analysis value corresponding to each freshness-related indicator is combined with the pre-configured weight factor corresponding to each freshness-related indicator into the formula , calculate the quality coefficient of the food cluster; Compare the weight of the ingredient cluster with the amount of the corresponding combination of ingredients in the current required semi-finished dish, and use the expression , calculate the weight coefficient of the food cluster, where Expressed as a preset allowable difference; The quality coefficient of the ingredient cluster is compared with the qualified quality coefficients corresponding to the various pre-prepared ingredients. If the quality coefficient of the ingredient cluster is greater than or equal to the qualified quality coefficient corresponding to the corresponding ingredient, and the weight coefficient of the ingredient cluster is 1, the selection conformity of the ingredient cluster is set to 1, otherwise the selection conformity of the ingredient cluster is set to 0.

6. The artificial intelligence-based data analysis system according to claim 1, characterized in that: The implementation method of determining whether a combination is possible is as follows: The selection conformity of the corresponding food clusters on the transfer placement area (3) corresponding to each U-shaped conveyor belt is compared with the standard value. If the quality coefficients of the corresponding food clusters on the transfer placement area (3) corresponding to each U-shaped conveyor belt are all standard values, it is determined that they can be combined, otherwise it is determined that they cannot be combined.

7. The artificial intelligence-based data analysis system according to claim 6, characterized in that: The identification process of the abnormal food cluster and abnormal direction is as follows: The food cluster whose selected compliance does not meet the standard value is recorded as an abnormal food cluster, and the quality coefficient of the abnormal food cluster is compared with the standard quality coefficient of the corresponding food. If the quality coefficient of the abnormal food cluster is less than the standard quality coefficient of the corresponding food, the abnormality points to the quality not meeting the standard. At the same time, the weight coefficient of the abnormal food cluster is obtained. If the weight coefficient of the abnormal food cluster is 0, the abnormality points to the weight not meeting the standard.

8. The artificial intelligence-based data analysis system according to claim 1, characterized in that: It also includes a detection mechanism start-stop intelligent control module, which is used to set a voice announcer on the combination area (4), and count the number of semi-finished dishes in the combination area (4) in real time, and compare it with the preparation amount corresponding to the semi-finished dishes currently required for matching. If the number of semi-finished dishes in the combination area (4) reaches the preparation amount corresponding to the semi-finished dishes currently required for matching, the detection mechanism is controlled to suspend operation and use the voice announcer to remind people to pick up dishes. Otherwise, the detection mechanism is controlled to continue operating and count the number of semi-finished dishes remaining to be matched, and transmit it to the matching end.