Canteen management method and system based on big data, Internet of Things and image analysis

Abstract

The present invention relates to the technical field of canteen management, and discloses a canteen management method and system based on big data, the Internet of Things and image analysis, including: real-time collection, preprocessing and storage of data of various links in the dining area and meal preparation area of ​​the canteen, wherein the link data includes image information and non-image information; data analysis and image analysis are performed on the preprocessed link data to predict the number of diners, identity information of diners and dining dishes in the future, so as to automatically manage the canteen operation in combination with a preset management strategy, and at the same time, identify abnormal actions of staff and diners and possible failures of equipment, and perform alarms and / or parameter adjustments after identifying abnormal actions and / or failures. The method and system of the present invention can improve management efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of canteen management, and in particular to a canteen management method and system based on big data, the Internet of Things and image analysis. Background Art

[0002] In modern canteen management, traditional management methods mainly rely on manual operations and experience-based decision-making. Common practices include manually recording the number of people dining each day, purchasing ingredients based on historical data or estimates, and using manual inventory management. Although these methods meet the basic operational needs of canteens to a certain extent, as canteens expand in size and user needs diversify, these traditional methods are becoming increasingly insufficient.

[0003] At present, some canteens have begun to introduce information management systems, such as data collection through barcode scanning and electronic menus, but most of these systems operate independently, with limited data analysis capabilities, and it is difficult to provide comprehensive intelligent management support. At the same time, existing canteen management systems can usually only process limited data types and have difficulty integrating information from multiple sources, such as user dining preferences, real-time inventory changes, equipment operating status, etc., and have the following shortcomings:

[0004] Insufficient data collection: Traditional canteen management methods cannot comprehensively and real-time collect and analyze users’ dining behavior and preference data. Usually, the number of diners and food demand per day can only be estimated through historical data or manual surveys, which makes it difficult to flexibly respond to actual changes.

[0005] Inefficient management: Since management work relies on manual operations, it is inefficient and prone to statistical errors, which can lead to over-purchasing or under-supply of ingredients. In addition, manual decision-making is also difficult to respond quickly to emergencies, such as a sudden increase in the number of diners or equipment failure.

[0006] Lack of intelligent monitoring: In terms of food safety management, existing technologies mainly rely on manual inspections, which makes it difficult to monitor the kitchen operating environment and equipment operating status in real time. This may lead to health hazards during food storage and processing, increasing food safety risks. Summary of the invention

[0007] The present invention provides a canteen management method and system based on big data, the Internet of Things and image analysis, which realizes intelligent, automated and refined canteen management by integrating big data technology, Internet of Things equipment and image analysis technology. Through real-time data collection and analysis, the present invention can effectively improve canteen management efficiency, reduce food waste, enhance user dining experience, and enhance food safety monitoring capabilities.

[0008] The present invention is achieved through the following technical solutions:

[0009] A canteen management method based on big data, Internet of Things and image analysis, including:

[0010] Collect, pre-process and store data from all aspects of the dining area and food preparation area of ​​the cafeteria in real time, including image information and non-image information;

[0011] The pre-processed data of the above links are subjected to data analysis and image analysis to predict the number of diners, identity information of diners and information of dining dishes in the future, so as to automate the management of canteen operations in combination with preset management strategies. At the same time, abnormal actions of staff and diners and possible failures of equipment are identified, and alarms and / or parameter adjustments are made after abnormal actions and / or failures are identified.

[0012] As an optimization, the link data includes the temperature and humidity of the dining area and the food preparation area, the dining records of diners, the number of diners, the inventory of ingredients, the status of kitchen equipment and the staff's work actions. The dining records of diners include the diners' identity information, dining time and dining menu.

[0013] As an optimization, preprocessing the link data specifically involves using big data technology to clean up missing and erroneous link data.

[0014] As an optimization, the pre-processed data of the link is subjected to data analysis and image analysis to predict the number of diners, dining time, identity information of diners and dining dishes in the future, so as to automatically manage the canteen operation in combination with the preset management strategy. The specific process is as follows:

[0015] Establishing an identity behavior recognition model based on a deep learning algorithm, and training the identity behavior recognition model through an improved dung beetle algorithm, so as to input the image information into the identity behavior recognition model to recognize the identity information and behavior information of diners and staff;

[0016] The number of diners, dining time, identity information of diners and dining dishes in history are used for comprehensive analysis based on big data technology to predict the number of diners, dining time, identity information of diners and dining dishes in the future;

[0017] Calculate the amount of food used in the future period based on the predicted number of diners and food information in the future period;

[0018] Adjust the purchasing plan based on the inventory of ingredients and the usage of dishes in the future;

[0019] Determine whether the temperature and humidity in the dining area and food preparation area exceed the set thresholds.

[0020] As an optimization, the identity behavior recognition model takes the similarity between the actual identity information of diners and staff and the predicted identity information of diners and staff, and the sum of the similarities between the actual behavior information of diners and staff and the predicted behavior information of diners and staff as the objective function, and regards one set of combined values ​​of the parameters of the identity behavior recognition model as the location of a dung beetle. The specific process of training the identity behavior recognition model by the improved dung beetle algorithm is as follows:

[0021] S1. Initialize the position of the dung beetle population and the parameters of the dung beetle algorithm, wherein the parameters of the dung beetle algorithm include mapping parameters , number of iterations;

[0022] S2 calculates the fitness value of each dung beetle, where the fitness value is the similarity between the actual dining data and the predicted dining data;

[0023] S3, updating the dung beetles whose fitness values ​​are less than the first threshold through mode one, updating the dung beetles whose fitness values ​​are not less than the first threshold and less than the second threshold through mode two, and updating the dung beetles whose fitness values ​​are not less than the second threshold through mode three;

[0024] S4, updating the positions of dancing dung beetles, foraging dung beetles, stealing dung beetles, and breeding dung beetles, and determining whether each dung beetle is outside the boundary;

[0025] S5, recalculating the fitness value of the current dung beetle, and updating the dung beetle population according to the ranking of the fitness value, so as to obtain the optimal position, that is, the optimal solution;

[0026] S6. Determine whether the maximum number of iterations has been reached. If so, output the currently updated dung beetle population to obtain the optimal parameters in the identity behavior recognition model. Otherwise, repeat S3 to S6.

[0027] As an optimization, the initialization of the dung beetle population position is performed using the following formula:

[0028] Generate a chaotic sequence:

[0029] ;

[0030] in, For the i The state value of the generation, For the The status value of is a random number between 0 and 1. ; is the mapping parameter, ;

[0031] according to The population initialization position is obtained as:

[0032] ;

[0033] in, For the i The initial position of the dung beetle, is the lower bound of the search space for the optimization problem in the dung beetle algorithm, It is the upper bound of the search space for the optimization problem in the dung beetle algorithm.

[0034] As an optimization, in S3, the formula corresponding to the mode 1 is:

[0035] ;

[0036] The formula corresponding to the second mode is:

[0037] ;

[0038] The formula corresponding to the mode three is:

[0039] ;

[0040] Where N is the number of dung beetles, represents the fitness ranking of the i-th dung beetle after the t-th update. The fitness ranking is in descending order. is the current best position after the tth update, that is, the current local optimal position, represents the position of the i-th dung beetle after the t-th update, represents the position of the i-th dung beetle after the t+1th update, is the natural coefficient, which takes -1 or 1, where -1 means deviation from the original direction and 1 means no deviation. A random number, is the current lower bound, represents the current upper limit, m is a random floating point number, and , A random floating point number between -1 and 1.

[0041] The present invention also discloses a canteen management system based on big data, the Internet of Things and image analysis to implement the aforementioned canteen management method based on big data, the Internet of Things and image analysis, including:

[0042] Data collection module, used to collect data from all aspects of the dining area and food preparation area of ​​the cafeteria;

[0043] A data preprocessing module is used to clean the data of each link using big data technology to obtain accurate link data;

[0044] An image analysis module is used to analyze image information to obtain the status of kitchen equipment, the identity information of diners, the dining menu of diners, and the identity information and working actions of the staff, and at the same time, identify abnormal actions of the staff and diners;

[0045] A big data analysis module is used to analyze non-image information to obtain the temperature and humidity of the dining area and the meal preparation area, the status of the kitchen equipment, the number of diners, and the dining time, predict possible equipment failures through the kitchen equipment status, and predict the number of diners, dining time, identity information of diners, and dining dish information for a period of time in the future based on big data analysis technology, and let the predicted number of diners, dining time, identity information of diners, and dining dish information for a period of time in the future be the prediction result;

[0046] An intelligent decision control module is used to make adjustments based on the prediction results combined with a preset management strategy to obtain decision data;

[0047] A data storage module, used to store data of the intelligent decision control module and the data preprocessing module;

[0048] A communication module is used to send the acquired decision data to the management personnel, and at the same time, send the user data and menu information for a period of time in the future stored in the data storage module to the diners, wherein the user data includes the user's historical dining time, the amount of historical dining expenses, and the historical dining dishes information;

[0049] A user interface feedback module is used to receive feedback and opinions from diners. At the same time, the staff can enter the storage module through the user interface feedback module to view the stored data;

[0050] The alarm module is used to sound an alarm when the temperature and humidity in the dining area and food preparation area, the inventory of ingredients, the status of kitchen equipment, and the operating actions of staff and diners are abnormal.

[0051] As an optimization, the data acquisition module includes an intelligent meal vending machine, a temperature sensor, a humidity sensor, a camera, an RFID tag, and a management computer.

[0052] As an optimization, the image analysis module identifies abnormal movements of staff and diners from the image information based on a deep learning algorithm.

[0053] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0054] 1. Improve management efficiency: This invention uses the Internet of Things and big data technology to automate many aspects of canteen management (such as data collection, analysis, and decision-making), reducing manual intervention and improving overall management efficiency. For example, the system can automatically generate procurement plans and adjust menus in real time, reducing the complexity of manual operations; through real-time data collection and analysis, the system can quickly respond to changes in the number of diners, equipment failures and other emergencies, ensuring the stability and flexibility of canteen operations.

[0055] 2. Optimize resource utilization: Based on big data analysis, the system can more accurately predict food demand, reduce inventory backlogs and food waste, and improve resource utilization; by real-time monitoring of the operating status of kitchen equipment, the system can optimize equipment use, reduce energy consumption, and reduce operating costs.

[0056] 3. Enhanced food safety: The present invention uses image analysis technology to monitor the entire food processing process in real time to ensure that food processing meets safety standards and reduce food safety risks. The system can automatically detect abnormal behavior or equipment failure and immediately issue an early warning to prevent potential food safety issues.

[0057] 4. Improve user experience: By analyzing the user's dining behavior and preferences, the system can provide personalized dish recommendations and services to meet the diverse needs of users and improve user satisfaction. The system can speed up the dining settlement process through image recognition technology, reduce waiting time, and provide users with a more convenient dining experience.

[0058] 5. Reduce operating costs: The system of the present invention can optimize inventory management, reduce food waste and over-purchasing, and significantly reduce operating costs; by real-time monitoring of equipment status, the system can perform predictive maintenance to avoid repair costs and downtime losses caused by sudden equipment failures.

[0059] 6. Data-driven decision support: The comprehensive data analysis reports and intelligent decision-making suggestions provided by the system of the present invention provide canteen managers with a scientific basis for decision-making, helping them optimize management strategies and improve operational efficiency and economic benefits.

[0060] 7. Scalability and sustainability: The system architecture of the present invention is flexible and easy to expand and upgrade. Whether adding new functional modules or connecting new IoT devices, the system can easily adapt to meet the management of canteens of different sizes and needs; by continuously optimizing resource utilization and improving management efficiency, this patent can generate significant economic and environmental benefits in long-term operation and promote the sustainable development of canteens.

[0061] 8. Improve the management level of the canteen: The present invention integrates various aspects of canteen management into an intelligent platform, realizes systematization and unification of management, and improves the overall management level; through the data visualization function of the system, managers can clearly understand the operation status of the canteen, which is convenient for management and adjustment of strategies.

[0062] 9. Promote innovation and modernization: This invention makes full use of cutting-edge technologies such as big data, the Internet of Things and image analysis, promotes the innovation and modernization of traditional canteen management models, and improves the technical level of the entire industry.

[0063] To sum up, this patent not only effectively improves the canteen's operational efficiency and service quality by introducing intelligent and automated management methods, but also brings significant positive effects in resource optimization, food safety assurance, and user experience improvement, providing a comprehensive and advanced solution for modern canteen management. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0065] Figure 1 This is a module connection diagram of a canteen management system based on big data, the Internet of Things and image analysis described in the present invention. DETAILED DESCRIPTION

[0066] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0067] The intelligent canteen management method based on big data, the Internet of Things and image analysis proposes a comprehensive system architecture and operation process, aiming to achieve real-time monitoring, data analysis and intelligent decision-making of all aspects of the canteen. By introducing big data, the Internet of Things and image analysis technology, the problems of low efficiency, incomplete data, waste of resources and poor food safety monitoring in traditional canteen management methods are solved, and the goals of improving data collection and analysis capabilities, improving management efficiency, optimizing resource utilization, enhancing food safety monitoring and improving user experience are achieved, and the canteen management is made intelligent, refined and safe, which comprehensively improves the canteen's operational efficiency and service quality.

[0068] The following is a detailed description of the technical solution:

[0069] This embodiment 1 provides a canteen management method based on big data, the Internet of Things, and image analysis, including:

[0070] T1. Real-time collection, pre-processing and storage of data from all aspects of the dining area and food preparation area of ​​the cafeteria, including image information and non-image information;

[0071] First, the link data is collected in real time.

[0072] The system (the composition of the system will be described in detail later) first collects various types of data through IoT devices and sensors.

[0073] In some embodiments, the link data includes the temperature and humidity of the dining area and the food preparation area, the dining records of diners, the number of diners, the inventory of ingredients, the status of kitchen equipment, and the work actions of the staff. The dining records of diners include the identity information of the diners, dining time, and dining menu.

[0074] Here, the temperature and humidity of the dining area and the food preparation area, and the food inventory are non-image information. The temperature and humidity can be obtained based on the temperature and humidity sensors, while the food inventory can be input into the management computer by the management personnel.

[0075] The dining records of the diners, the number of diners, and the work actions of the staff are image information.

[0076] It should be noted that the status of kitchen equipment may include both image information and non-image information. For example, the external phenomenon of the kitchen equipment captured by a camera is image information, and the usage data of the kitchen equipment obtained by the internal sensor of the kitchen equipment is non-image information.

[0077] Secondly, the link data is preprocessed.

[0078] In some embodiments, preprocessing link data specifically involves using big data technology to clean missing and erroneous link data, such as denoising, filtering, and normalization, to ensure data quality and consistency.

[0079] T2. Perform data analysis and image analysis on the pre-processed link data to predict the number of diners, dining time, diners' identity information and dining menu information in the future, so as to automate the canteen operation in combination with the preset management strategy. At the same time, identify abnormal actions of staff and diners and possible failures of equipment, and issue alarms and / or adjust parameters after identifying abnormal actions and / or failures.

[0080] Through the Internet of Things and image analysis technology, the system can monitor key aspects of canteen operations in real time, such as food processing, equipment status, user flow, etc. Once an abnormality is detected (such as excessive temperature, equipment failure, abnormal user behavior, etc.), the system will immediately issue an early warning and automatically take preset measures, such as notifying managers or directly controlling equipment shutdown.

[0081] The system will regularly analyze historical data and real-time data to generate various reports on canteen operations, such as user preference analysis, dining peak forecast, food consumption forecast, etc. The intelligent decision-making module automatically adjusts the operation strategy based on these analysis results, such as dynamically adjusting the menu, optimizing procurement and inventory management, and customizing user services.

[0082] The system collects user feedback and opinions through the user interface, incorporates these feedback into data analysis, and continuously optimizes management strategies and service quality. At the same time, the system can provide personalized services such as recommending dishes and optimizing dining times based on user historical data and preferences.

[0083] Next, the implementation process of the above functions is introduced in detail.

[0084] In some embodiments, the pre-processed link data is subjected to data analysis and image analysis to predict the number of diners, dining time, identity information of diners, and dining dishes in the future, so as to automatically manage the canteen operation in combination with the preset management strategy. The specific process is as follows:

[0085] Establishing an identity behavior recognition model based on a deep learning algorithm, and training the identity behavior recognition model through an improved dung beetle algorithm, so as to input the image information into the identity behavior recognition model to recognize the identity information and behavior information of diners and staff;

[0086] The number of diners, dining time, identity information of diners and dining dishes in history are used for comprehensive analysis based on big data technology to predict the number of diners, dining time, identity information of diners and dining dishes in the future;

[0087] Calculate the amount of food used in the future period based on the predicted number of diners, dining time, and dining food information in the future period;

[0088] Adjust the purchasing plan based on the inventory of ingredients and the usage of dishes in the future;

[0089] Determine whether the temperature and humidity in the dining area and food preparation area exceed the set thresholds.

[0090] In some embodiments, the identity behavior recognition model takes the similarity between the actual identity information of diners and staff and the predicted identity information of diners and staff, and the sum of the similarities between the actual behavior information of diners and staff and the predicted behavior information of diners and staff as the objective function, and regards a group of combined values ​​of each parameter of the identity behavior recognition model as the location of a dung beetle. The specific process of training the identity behavior recognition model by using the improved dung beetle algorithm is as follows:

[0091] S1. Initialize the position of the dung beetle population and the parameters of the dung beetle algorithm. The parameters of the dung beetle algorithm include mapping parameters. , number of iterations;

[0092] In some embodiments, initializing the position of the dung beetle population is specifically performed by the following formula:

[0093] Generate a chaotic sequence:

[0094] ;

[0095] in, For the i The state value of the generation, For the The status value of is a random number between 0 and 1. ; is the mapping parameter, ;

[0096] according to The population initialization position is obtained as:

[0097] ;

[0098] in, For the i The initial position of the dung beetle, is the lower bound of the search space for the optimization problem in the dung beetle algorithm, It is the upper bound of the search space for the optimization problem in the dung beetle algorithm.

[0099] Using the above-mentioned wonton mapping can generate a highly diverse initial population and improve the diversity of the initial solutions.

[0100] S2 calculates the fitness value of each dung beetle, and the fitness value is the similarity between the actual dining data and the predicted dining data;

[0101] S3, updating the dung beetles whose fitness values ​​are less than the first threshold through mode one, updating the dung beetles whose fitness values ​​are not less than the first threshold and less than the second threshold through mode two, and updating the dung beetles whose fitness values ​​are not less than the second threshold through mode three;

[0102] In some embodiments, the formula corresponding to mode 1 is:

[0103] ;

[0104] The formula corresponding to mode 2 is:

[0105] ;

[0106] The formula corresponding to mode three is:

[0107] ;

[0108] Where N is the number of dung beetles, represents the fitness ranking of the i-th dung beetle after the t-th update. The fitness ranking is in descending order. is the current best position after the tth update, that is, the current local optimal position, represents the position of the i-th dung beetle after the t-th update, represents the position of the i-th dung beetle after the t+1th update, is the natural coefficient, which takes -1 or 1, where -1 means deviation from the original direction and 1 means no deviation. A random number, is the current lower bound, represents the current upper limit, m is a random floating point number, and , A random floating point number between -1 and 1.

[0109] By setting the first threshold and the second threshold, different situations have different update formulas, which not only fully utilizes the current position information, but also adds random interference information, which is conducive to the dung beetle position jumping out of the local optimum, so that the dung beetle individuals can perform global search and optimization within a given area, thereby improving the optimization ability of the dung beetle algorithm. In this embodiment, the first threshold is 0.312 and the second threshold is 0.698.

[0110] S4, updating the positions of dancing dung beetles, foraging dung beetles, stealing dung beetles, and breeding dung beetles, and determining whether each dung beetle is outside the boundary;

[0111] The update formulas for dancing dung beetles, foraging dung beetles, stealing dung beetles, and breeding dung beetles are the update formulas for the existing dung beetle algorithm, which will not be repeated here.

[0112] S5, recalculating the fitness value of the current dung beetle, and updating the dung beetle population according to the ranking of the fitness value, so as to obtain the optimal position, that is, the optimal solution;

[0113] S6. Determine whether the maximum number of iterations has been reached. If so, output the currently updated dung beetle population to obtain the optimal parameters in the identity behavior recognition model. Otherwise, repeat S3 to S6.

[0114] Next, the system used in the present invention is specifically introduced. Example

[0115] A canteen management system based on big data, Internet of Things and image analysis, such as Figure 1 As shown, a canteen management method based on big data, Internet of Things and image analysis to implement the above-mentioned embodiment includes:

[0116] Data acquisition and preprocessing, data analysis and intelligent decision-making, image analysis and monitoring, user interface and feedback.

[0117] Data collection and preprocessing:

[0118] Data collection module, used to collect data from all aspects of the dining area and food preparation area of ​​the cafeteria;

[0119] In some embodiments, the data acquisition module includes a smart meal vending machine, a temperature sensor, a humidity sensor, a camera, and an RFID tag.

[0120] Based on the Internet of Things technology, smart devices installed in various links of the canteen (such as smart meal vending machines, temperature sensors, cameras, RFID tags, etc.) collect data in real time. These data include but are not limited to: user dining records, food inventory, kitchen equipment status, temperature and humidity during food processing, etc.

[0121] That is, the first is the deployment of IoT devices: IoT devices are deployed in various key links of the canteen, including temperature sensors, humidity sensors, cameras, RFID tags, etc. These devices collect user dining data, food inventory, kitchen environment data (such as temperature, humidity), equipment operation status, etc. in real time.

[0122] IoT and data integration: IoT devices are used to collect data from the physical world and integrate multi-dimensional data (such as temperature, humidity, images, etc.) into a unified database. The system forms a comprehensive understanding of canteen operations through correlation analysis of these data.

[0123] A data storage module, used to store data of the intelligent decision control module and the data preprocessing module;

[0124] The communication module is used to send the acquired decision data to the management personnel, and at the same time, send the user data and menu information for a period of time in the future stored in the data storage module to the diners, wherein the user data includes the user's historical dining time, the amount of historical dining expenses, the historical dining time, and the historical dining dishes information;

[0125] This is data transmission and storage: All collected data is transmitted to the central data processing system via wireless network. In order to ensure the stability and security of data transmission, encryption protocol and redundancy mechanism are adopted. Data storage adopts distributed database technology to support large-scale data storage and fast retrieval.

[0126] The data preprocessing module is used to clean the data of each link using big data technology to obtain accurate link data; and use big data technology to clean, store, analyze and model the data.

[0127] This is data preprocessing: the raw data will go through preprocessing steps before storage, such as data cleaning (removing noise and invalid data), data formatting, data normalization, etc., to ensure data quality and consistency.

[0128] Data analysis and intelligent decision-making:

[0129] A big data analysis module is used to analyze non-image information to obtain the temperature and humidity of the dining area and the food preparation area, the status of the kitchen equipment, and the number of diners, predict possible equipment failures through the status of the kitchen equipment, and predict the number of diners, dining time, identity information of the diners, and dining dishes in the future based on big data analysis technology, and let the predicted number of diners, dining time, identity information of the diners, and dining dishes in the future be the prediction result;

[0130] This module uses big data technology combined with image data to conduct a comprehensive analysis of the massive amount of data collected. The analysis includes user dining behavior analysis, food consumption prediction, inventory optimization, equipment status evaluation, environmental monitoring data analysis, etc. This module supports real-time data processing and historical data analysis, helping managers make data-based decisions, such as optimizing menus, adjusting procurement plans, and predicting the number of diners.

[0131] Big data processing frameworks such as Hadoop and Spark are used to process massive amounts of data in a distributed manner to mine hidden patterns and trends in the data. These analysis results are used to support intelligent decision-making in the system.

[0132] User behavior analysis: By analyzing users' dining times, food choices, and amounts spent, the system can identify users' dietary preferences and optimize menu configuration and service strategies.

[0133] Food consumption forecast: Based on historical data and real-time data, the system can predict food consumption in the future, helping managers optimize procurement plans and reduce waste.

[0134] The number of diners on that day, their identity information, the total meal information on that day, and the number of diners in history N days before that day, their identity information, and the total meal information are input into the trained identity behavior recognition model to predict the number of diners, their identity information, and meal information for a period of time in the future.

[0135] Equipment status assessment: By analyzing equipment operation data, the system can predict possible equipment failures in advance, perform preventive maintenance, and avoid operational interruptions caused by sudden equipment failures.

[0136] The alarm module is used to sound an alarm when the temperature and humidity in the dining area and food preparation area, the inventory of ingredients, the status of kitchen equipment, and the operating actions of staff and diners are abnormal.

[0137] Combined with the results of big data analysis, the system can automatically generate the canteen's operating strategy. The main functions include:

[0138] Menu optimization: Based on user preference analysis and food consumption prediction, the system automatically adjusts the menu to provide the most popular dishes that meet health needs.

[0139] Inventory management: Based on consumption forecasts and inventory data, the system automatically generates a purchase plan and monitors inventory in real time to avoid out-of-stock or expiration issues.

[0140] Equipment control: Based on the equipment status assessment, the system can automatically adjust the equipment's operating parameters, or automatically shut down or alarm when necessary.

[0141] Image Analysis and Monitoring:

[0142] An image analysis module is used to analyze image information to obtain the status of kitchen equipment, the identity information of diners, the dining menu of diners, and the identity information and working actions of the staff, and at the same time, identify abnormal actions of the staff and diners;

[0143] The image analysis module identifies abnormal movements of staff and diners from image information based on deep learning algorithms.

[0144] In some embodiments, an identity behavior recognition model is established based on a deep learning algorithm, and the identity behavior recognition model is trained by an improved dung beetle algorithm, so that the image information is input into the identity behavior recognition model to recognize the identity information and behavior information of diners and staff;

[0145] In some embodiments, the identity behavior recognition model takes the similarity between the actual identity information of diners and staff and the predicted identity information of diners and staff, and the sum of the similarities between the actual behavior information of diners and staff and the predicted behavior information of diners and staff as the objective function, and regards a group of combined values ​​of each parameter of the identity behavior recognition model as the location of a dung beetle. The specific process of training the identity recognition model by the improved dung beetle algorithm is as follows:

[0146] S1. Initialize the position of the dung beetle population and the parameters of the dung beetle algorithm. The parameters of the dung beetle algorithm include mapping parameters. , number of iterations;

[0147] In some embodiments, initializing the position of the dung beetle population is specifically performed by the following formula:

[0148] Generate a chaotic sequence:

[0149] ;

[0150] in, For the i The state value of the generation, For the The status value of is a random number between 0 and 1. ; is the mapping parameter, ;

[0151] according to The population initialization position is obtained as:

[0152] ;

[0153] in, For the i The initial position of the dung beetle, is the lower bound of the search space for the optimization problem in the dung beetle algorithm, It is the upper bound of the search space for the optimization problem in the dung beetle algorithm.

[0154] Using the above-mentioned wonton mapping can generate a highly diverse initial population and improve the diversity of the initial solutions.

[0155] S2 calculates the fitness value of each dung beetle, and the fitness value is the similarity between the actual dining data and the predicted dining data;

[0156] S3, updating the dung beetles whose fitness values ​​are less than the first threshold through mode one, updating the dung beetles whose fitness values ​​are not less than the first threshold and less than the second threshold through mode two, and updating the dung beetles whose fitness values ​​are not less than the second threshold through mode three;

[0157] In some embodiments, the formula corresponding to mode 1 is:

[0158] ;

[0159] The formula corresponding to mode 2 is:

[0160] ;

[0161] The formula corresponding to mode three is:

[0162] ;

[0163] Where N is the number of dung beetles, represents the fitness ranking of the i-th dung beetle after the t-th update. The fitness ranking is in descending order. is the current best position after the tth update, that is, the current local optimal position, represents the position of the i-th dung beetle after the t-th update, represents the position of the i-th dung beetle after the t+1th update, is the natural coefficient, which takes -1 or 1, where -1 means deviation from the original direction and 1 means no deviation. A random number, is the current lower bound, represents the current upper limit, m is a random floating point number, and , A random floating point number between -1 and 1.

[0164] By setting the first threshold and the second threshold, different situations have different update formulas, which not only fully utilizes the current position information, but also adds random interference information, which is conducive to the dung beetle position jumping out of the local optimum, so that the dung beetle individuals can perform global search and optimization within a given area, thereby improving the optimization ability of the dung beetle algorithm. In this embodiment, the first threshold is 0.312 and the second threshold is 0.698.

[0165] S4, updating the positions of dancing dung beetles, foraging dung beetles, stealing dung beetles, and breeding dung beetles, and determining whether each dung beetle is outside the boundary;

[0166] The update formulas for dancing dung beetles, foraging dung beetles, stealing dung beetles, and breeding dung beetles are the update formulas for the existing dung beetle algorithm, which will not be repeated here.

[0167] S5, recalculating the fitness value of the current dung beetle, and updating the dung beetle population according to the ranking of the fitness value, so as to obtain the optimal position, that is, the optimal solution;

[0168] S6. Determine whether the maximum number of iterations has been reached. If so, output the currently updated dung beetle population to obtain the optimal parameters in the identity behavior recognition model. Otherwise, repeat S3 to S6.

[0169] Image acquisition and processing: Through high-definition cameras deployed in the cafeteria, the system collects image data of the dining area and kitchen in real time. These image data are used for user identification, queue counting, dining behavior analysis, kitchen operation monitoring, etc.

[0170] This module is used for user identification, queue counting, dining behavior analysis, kitchen operation monitoring, etc. Using deep learning algorithms, this module can identify abnormal behaviors, such as food processing that is not carried out in accordance with regulations and abnormal equipment operation.

[0171] User identification: Through image analysis and facial recognition technology, the system can automatically identify the user's identity and realize functions such as automatic settlement and personalized service push. At the same time, identity recognition technology can prevent unauthorized persons from entering the cafeteria or kitchen area.

[0172] Behavior analysis and monitoring: Using deep learning algorithms, the system can analyze users’ dining behaviors, such as meal times, waiting time, dining habits, etc., to help optimize canteen layout and service processes.

[0173] Food safety monitoring: In the kitchen area, the system monitors the food processing process in real time through image analysis to detect any violations of operating procedures, such as irregular hygiene operations, equipment not used as required, etc. Once an abnormality is detected, the system will immediately alarm and notify relevant management personnel.

[0174] Intelligent decision-making control module, used to make adjustments based on the prediction results and preset management strategies to obtain decision data;

[0175] This module combines data analysis results with preset management strategies to automatically manage canteen operations. For example, it can adjust the supply of dishes based on the real-time number of diners, automatically generate procurement plans based on inventory conditions, or issue alarms and take corresponding measures based on kitchen environment monitoring results.

[0176] The system feeds back the results of analysis and monitoring to the control module to realize automatic control of the equipment, such as adjusting the temperature, starting or stopping the equipment, issuing alarms, etc. This closed-loop control system ensures the operational efficiency and safety of the canteen.

[0177] User Interface Feedback:

[0178] The user interface feedback module is used to receive feedback and opinions from diners. At the same time, the staff can enter the storage module through the user interface feedback module to view the stored data; it provides an interactive interface for canteen managers and users, including management background, mobile applications, etc. Managers can view data in real time, set management strategies, receive system suggestions, etc. through this interface; users can view personal dining records, obtain personalized recommendations, and participate in canteen management feedback through mobile applications.

[0179] Management backend: Provides a comprehensive management platform for canteen managers, displaying real-time data, analysis reports and system suggestions. Managers can view various operating indicators of the canteen through the backend, adjust management strategies, view alarm records, and perform system maintenance.

[0180] User mobile application: Users can view their dining records, get personalized dish recommendations, participate in feedback surveys, reserve seats, etc. through the mobile application. The system continuously optimizes service quality based on user feedback.

[0181] Therefore, in brief, the implementation process of the present invention is as follows:

[0182] 1. Data collection process

[0183] The IoT devices (data acquisition modules) in the cafeteria collect data in real time and upload it to the server.

[0184] The data preprocessing module cleans, formats and stores the data.

[0185] 2. Data analysis process

[0186] The big data analysis module performs real-time and batch analysis on data to generate user behavior analysis reports, food consumption forecasts, equipment status assessments, etc.

[0187] 3. Decision-making and execution process

[0188] The intelligent decision-making control module automatically generates and executes operation strategies based on the analysis results.

[0189] The system monitors the execution effect in real time and adjusts the strategy based on feedback.

[0190] 4. Image processing and monitoring process

[0191] The camera (data acquisition module) collects images in real time and transmits them to the image analysis module.

[0192] The image analysis module performs user identification, behavior analysis, and food safety monitoring.

[0193] 5. User feedback and optimization process

[0194] Users provide feedback through the mobile application (user interface feedback module), and the system collects and analyzes this feedback.

[0195] Based on the feedback results, the system adjusts its operating strategy and service quality.

[0196] In order to ensure the efficiency and reliability of the system, the following is the implementation method of this patent:

[0197] System architecture optimization: Distributed architecture is adopted to ensure the scalability and high availability of the system. The data processing part adopts cloud computing technology to support parallel processing of large-scale data.

[0198] Hardware configuration: Select high-performance IoT devices and high-definition cameras to ensure the accuracy of data collection and the clarity of image processing. The server is equipped with high-performance CPU / GPU and large-capacity storage to meet the needs of real-time data processing.

[0199] Algorithm optimization: In image processing and big data analysis, deep learning and machine learning algorithms are used to improve the recognition accuracy and analysis efficiency of the system. Regular model updates and algorithm optimization are performed to ensure that the system is always in the best condition.

[0200] User experience optimization: The user interface is simple and intuitive, which is convenient for users and managers to operate. The mobile application supports personalized push and real-time interaction, which enhances user experience and stickiness.

[0201] Security: The system is equipped with strict data encryption and authority management mechanisms to ensure the security of user data and system operations. The canteen equipment has emergency handling functions and can automatically switch to manual operation mode in the event of system failure to ensure the continuity of operations.

[0202] Through these technical contents and implementation methods, this patent can greatly improve the intelligence level of canteen management and achieve more efficient, safer and more user-friendly operations.

[0203] The present invention is based on real-time data transmission technology between IoT devices and between devices and a central system to ensure fast and reliable data transmission, and to process the collected data in real time (such as cleaning and formatting) to ensure its accuracy and availability.

[0204] The present invention uses big data technology to comprehensively analyze historical data and real-time data in canteen management, mines user dining behaviors, food consumption patterns, equipment operation rules, etc., and thus generates algorithms and methods for intelligent decision-making for optimized management.

[0205] The present invention uses cameras deployed in the cafeteria to collect real-time images of the dining area and kitchen, and uses deep learning algorithms to perform image analysis, including methods such as user identity recognition, queue count, dining behavior analysis, and kitchen operation monitoring.

[0206] In summary, the present invention can achieve the following benefits:

[0207] Intelligent and automated management: The system uses the Internet of Things and big data technologies to achieve comprehensive intelligent management of the canteen, reduce manual operations and improve management efficiency.

[0208] Real-time and accuracy: Real-time data collection and analysis enable the system to respond quickly to changes and emergencies, ensuring the stability and safety of canteen operations.

[0209] Personalization and improvement of user experience: By analyzing user data, the system can provide services that are more in line with user needs and improve user satisfaction.

[0210] Full-process monitoring and food safety: Image analysis is combined with the Internet of Things to achieve full-process monitoring of canteen operations, ensure food safety, and reduce human errors.

[0211] The application scenarios of the present invention may include the following scenarios:

[0212] Large canteens in universities, enterprises, hospitals, etc.: Suitable for places that need to handle a large number of diners and have high requirements for food safety and operational efficiency.

[0213] Chain restaurant management: Help chain canteens or catering companies optimize overall operations through unified management and data analysis.

[0214] Personalized dining services: Provide users with personalized services based on personal data to enhance dining experience.

[0215] In summary, the present invention aims to completely transform the traditional canteen management model through technical means and create a more intelligent, sophisticated and efficient canteen management system.

[0216] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A canteen management method based on big data, Internet of Things and image analysis, characterized in that: include: Collect, pre-process and store data from all aspects of the dining area and food preparation area of ​​the cafeteria in real time, including image information and non-image information; Perform data analysis and image analysis on the pre-processed data of the link to predict the number of diners, dining time, identity information of diners and dining dishes in the future, so as to automate the management of canteen operations in combination with preset management strategies. At the same time, identify abnormal actions of staff and diners and possible failures of equipment, and issue alarms and / or adjust parameters after identifying abnormal actions and / or failures; The pre-processed data is subjected to data analysis and image analysis to predict the number of diners, dining time, identity information of diners and dining dishes in the future, so as to automatically manage the canteen operation in combination with the preset management strategy. The specific process is as follows: Establishing an identity behavior recognition model based on a deep learning algorithm, and training the identity behavior recognition model through an improved dung beetle algorithm, so as to input the image information into the identity behavior recognition model to recognize the identity information and behavior information of diners and staff; The number of diners, dining time, identity information of diners and dining dishes in history are used for comprehensive analysis based on big data technology to predict the number of diners, dining time, identity information of diners and dining dishes in the future; Calculate the amount of food used in the future period based on the predicted number of diners and food information in the future period; Adjust the purchasing plan based on the inventory of ingredients and the usage of dishes in the future; Determine whether the temperature and humidity in the dining area and food preparation area exceed the set threshold; The identity behavior recognition model takes the similarity between the actual identity information of diners and staff and the predicted identity information of diners and staff, and the similarity between the actual behavior information of diners and staff and the predicted behavior information of diners and staff as the objective function, and regards a group of combined values ​​of each parameter of the identity behavior recognition model as the location of a dung beetle. The specific process of training the identity behavior recognition model by the improved dung beetle algorithm is as follows: S1, initializing the position of the dung beetle population and the parameters of the dung beetle algorithm, wherein the parameters of the dung beetle algorithm include a mapping parameter α and the number of iterations; S2 calculates the fitness value of each dung beetle; S3, updating the dung beetles whose fitness values ​​are less than the first threshold through mode one, updating the dung beetles whose fitness values ​​are not less than the first threshold and less than the second threshold through mode two, and updating the dung beetles whose fitness values ​​are not less than the second threshold through mode three; In S3, the formula corresponding to the mode 1 is: x i (t+1)=x * (t)-ωx i (t); The formula corresponding to the second mode is: The formula corresponding to the mode three is: Where N is the number of dung beetles, represents the fitness ranking of the i-th dung beetle after the t-th update. The fitness ranking is in descending order. * (t) is the current best position after the tth update, that is, the current local optimal position, x i (t) represents the position of the i-th dung beetle after the t-th update, x i (t+1) represents the position of the i-th dung beetle after the t+1th update, α is a natural coefficient, which takes -1 or 1, -1 means deviation from the original direction, 1 means no deviation, b∈(0,1) is a random number, Lb * is the current lower bound, Ub * Represents the current upper limit, ω is a random floating point number, and ω∈[0.4,0.9], β is a random floating point number between -1 and 1; S4, updating the positions of dancing dung beetles, foraging dung beetles, stealing dung beetles, and breeding dung beetles, and determining whether each dung beetle is outside the boundary; S5, recalculating the fitness value of the current dung beetle, and updating the dung beetle population according to the ranking of the fitness value, so as to obtain the optimal position, that is, the optimal solution; S6. Determine whether the maximum number of iterations has been reached. If so, output the currently updated dung beetle population to obtain the optimal parameters in the identity behavior recognition model. Otherwise, repeat S3 to S6.

2. A canteen management method based on big data, Internet of Things and image analysis according to claim 1, characterized in that: The link data includes the temperature and humidity of the dining area and the food preparation area, the dining records of diners, the number of diners, the inventory of ingredients, the status of kitchen equipment and the work actions of the staff. The dining records of diners include the identity information of the diners, dining time and dining menu.

3. A canteen management method based on big data, Internet of Things and image analysis according to claim 1, characterized in that: Preprocessing the link data specifically involves using big data technology to clean up missing and erroneous link data.

4. The canteen management method based on big data, Internet of Things and image analysis according to claim 1 is characterized in that: Initializing the position of the dung beetle population is done using the following formula: Generate a chaotic sequence: Among them, S i is the state value of the i-th generation, S i+1 is the state value of the i+1th node, S i is a random number between 0 and 1, S0 = 0.342; α is the mapping parameter, α∈(0,1); According to S i The population initialization position is obtained as: X i =Lb+(Ub-Lb)S i ; Among them, X i is the initial position of the i-th dung beetle, Lb is the lower bound of the search space for the optimization problem in the dung beetle algorithm, and Ub is the upper bound of the search space for the optimization problem in the dung beetle algorithm.

5. A canteen management system based on big data, the Internet of Things and image analysis, to implement a canteen management method based on big data, the Internet of Things and image analysis as described in any one of claims 1 to 4, characterized in that: include: Data collection module, used to collect data from all aspects of the dining area and food preparation area of ​​the cafeteria; A data preprocessing module is used to clean the data of each link using big data technology to obtain accurate link data; An image analysis module is used to analyze image information to obtain the status of kitchen equipment, the identity information of diners, the dining menu of diners, and the identity information and working actions of the staff, and at the same time, identify abnormal actions of the staff and diners; A big data analysis module is used to analyze non-image information to obtain the temperature and humidity of the dining area and the meal preparation area, the status of the kitchen equipment, the number of diners, and the dining time, predict possible equipment failures through the kitchen equipment status, and predict the number of diners, dining time, identity information of diners, and dining dish information for a period of time in the future based on big data analysis technology, and let the predicted number of diners, dining time, identity information of diners, and dining dish information for a period of time in the future be the prediction result; An intelligent decision control module is used to make adjustments based on the prediction results combined with a preset management strategy to obtain decision data; A data storage module, used to store data of the intelligent decision control module and the data preprocessing module; A communication module is used to send the acquired decision data to the management personnel, and at the same time, send the user data and menu information for a period of time in the future stored in the data storage module to the diners, wherein the user data includes the user's historical dining time, the amount of historical dining expenses, and the historical dining dishes information; A user interface feedback module is used to receive feedback and opinions from diners. At the same time, the staff can enter the storage module through the user interface feedback module to view the stored data; The alarm module is used to sound an alarm when the temperature and humidity in the dining area and food preparation area, the inventory of ingredients, the status of kitchen equipment, and the operating actions of staff and diners are abnormal.

6. A canteen management system based on big data, Internet of Things and image analysis according to claim 5, characterized in that: The data acquisition module includes an intelligent meal vending machine, a temperature sensor, a humidity sensor, a camera, an RFID tag, and a management computer.

7. A canteen management system based on big data, Internet of Things and image analysis according to claim 6, characterized in that: The image analysis module identifies abnormal movements of staff and diners from the image information based on a deep learning algorithm.

Images