An interconnected and synchronized supermarket management method and platform based on AI and IOT

Abstract

The present invention proposes an interconnected and synchronous supermarket management method and platform based on AI and IOT, which belongs to the field of artificial intelligence and supermarket management technology; the method includes: deploying IOT devices in supermarkets, and collecting relevant data in supermarkets in real time based on the deployed IOT devices; integrating the collected real-time data through the Internet of Things gateway, and preprocessing the integrated data, and transferring the preprocessed data to the cloud space; intelligently analyzing the data in the cloud space based on the AI ​​algorithm, and performing interconnected and synchronous management of various links based on the AI ​​analysis results. By deploying IOT devices in supermarkets, various data in supermarkets can be collected in real time, such as product information, customer flow, inventory status, etc., which improves the timeliness and accuracy of the data.

Description

Technical Field

[0001] The present invention proposes an interconnected and synchronized supermarket management method and platform based on AI and IOT, belonging to the technical field of artificial intelligence and supermarket management. Background Art

[0002] The traditional supermarket industry is undergoing a profound digital transformation. As consumer demands become increasingly diversified and personalized, customers' requirements for shopping experience have far exceeded the scope of simple commodity purchases. They expect to enjoy more convenient, intelligent, and personalized services in supermarkets. However, the traditional supermarket management model often relies on manual monitoring, manual data entry and analysis. This method is not only inefficient and difficult to cope with large-scale data processing needs, but also difficult to accurately capture customers' consumption habits, preference changes, and market trends, thus limiting supermarkets' ability to improve customer satisfaction, optimize operational efficiency, and achieve precision marketing. Summary of the invention

[0003] The present invention provides an interconnected and synchronized supermarket management method and platform based on AI and IOT, which is used to solve the problems mentioned in the above background technology:

[0004] The present invention proposes an interconnected and synchronized supermarket management method based on AI and IOT, the method comprising:

[0005] S1. Deploy IOT devices in supermarkets and collect relevant data in supermarkets in real time based on the deployed IOT devices;

[0006] S2. Integrate the collected real-time data through the IoT gateway, pre-process the integrated data, and transfer the pre-processed data to the cloud space;

[0007] S3, intelligently analyze the data in the cloud space based on AI algorithms, and conduct interconnection and synchronization management of each link based on the AI ​​analysis results;

[0008] S4. Provide personalized services to customers through multiple channels; collect user feedback and optimize supermarket management and services based on user feedback.

[0009] Furthermore, the S1 includes:

[0010] S11. Analyze the operation requirements of supermarkets and determine the required data types; select IoT devices based on the determined data types;

[0011] S12. Formulate a deployment strategy for IoT devices based on the supermarket layout diagram, and install the devices according to the deployment strategy;

[0012] S13. Calibrate each device individually, set the data collection frequency, and perform preliminary data processing on the device side, wherein the preliminary data processing includes data compression and outlier filtering.

[0013] Furthermore, the S2 includes:

[0014] S21. Collect data streams from different IoT devices through the IoT gateway and integrate the data through data integration tools to form a unified data stream;

[0015] S22, preprocessing the integrated data, wherein the preprocessing includes cleaning, removing duplicate, erroneous or invalid data; and performing integrity and consistency checks on the data through a data verification mechanism;

[0016] S23, extracting features from the preprocessed data, identifying features useful for subsequent analysis, and applying a feature selection algorithm to remove redundant or irrelevant features;

[0017] S24. Encrypt the pre-processed data using an encryption algorithm, and transmit the encrypted data to the cloud space through a security protocol.

[0018] Furthermore, the S3 includes:

[0019] S31. Use historical data and business rules to train the AI ​​model, and perform multiple iterations to optimize the trained AI model;

[0020] S32, allocating the data in the cloud space to the corresponding AI models for analysis according to different dimensions, and processing multiple analysis tasks in parallel for each analysis task;

[0021] S33. Conduct in-depth analysis of input data through AI models and output deep insight results; conduct predictive analysis based on historical data and current trends to provide suggestions for supermarket management;

[0022] S34. Formulate interconnection and synchronization management strategies for each link based on AI analysis results, automatically execute management strategies, and dynamically adjust strategies based on execution status.

[0023] Furthermore, the S4 includes:

[0024] S41. Develop personalized shopping recommendations and service push strategies based on customer behavior data and preference information, and push personalized content to customers through multiple channels;

[0025] S42. Collect user feedback through various feedback channels, organize and analyze the collected feedback, and identify customer concerns and improvement directions;

[0026] S43. Continuously optimize and iterate the service and operation strategies of supermarkets based on user feedback and AI analysis results.

[0027] The present invention proposes an interconnected and synchronized supermarket management platform based on AI and IOT, the platform comprising:

[0028] Data collection module: Deploy IoT devices in supermarkets, and collect relevant data in supermarkets in real time based on the deployed IoT devices;

[0029] Data processing module: Integrate the collected real-time data through the IoT gateway, pre-process the integrated data, and transmit the pre-processed data to the cloud space;

[0030] Data analysis module: intelligently analyzes the data in the cloud space based on AI algorithms, and performs interconnected and synchronous management of each link based on the AI ​​analysis results;

[0031] Service optimization module: Provide personalized services to customers through multiple channels; collect user feedback and optimize supermarket management and services based on user feedback.

[0032] Furthermore, the data acquisition module includes:

[0033] Type determination module: Analyze supermarket operation requirements and determine the required data types; select IoT devices based on the determined data types;

[0034] Equipment installation module: formulate deployment strategies for IoT devices based on the supermarket layout diagram, and install devices according to the deployment strategies;

[0035] Abnormal filtering module: calibrate each device individually, set the data collection frequency, and perform preliminary data processing on the device side, including data compression and abnormal value filtering.

[0036] Furthermore, the data processing module includes:

[0037] Data integration module: collects data streams from different IoT devices through the IoT gateway and integrates the data through data integration tools to form a unified data stream;

[0038] Data verification module: pre-processes the integrated data, including cleaning to remove duplicate, erroneous or invalid data; and verifies the integrity and consistency of the data through a data verification mechanism;

[0039] Irrelevant removal module: extracts features from preprocessed data, identifies features useful for subsequent analysis, and applies feature selection algorithms to remove redundant or irrelevant features;

[0040] Data encryption module: Encrypt the pre-processed data through encryption algorithm, and transmit the encrypted data to the cloud space through security protocol.

[0041] Furthermore, the data analysis module includes:

[0042] Iterative optimization module: Use historical data and business rules to train the AI ​​model, and perform multiple iterative optimizations on the trained AI model;

[0043] Parallel processing module: distributes the data in the cloud space to the corresponding AI models for analysis according to different dimensions, and processes multiple analysis tasks in parallel for each analysis task;

[0044] Predictive analysis module: In-depth analysis of input data through AI models to output deep insight results; predictive analysis based on historical data and current trends to provide suggestions for supermarket management;

[0045] Dynamic adjustment module: formulate interconnection and synchronization management strategies for each link based on AI analysis results, automatically execute management strategies, and dynamically adjust strategies based on execution status.

[0046] Furthermore, the service optimization module includes:

[0047] Content push module: Based on customer behavior data and preference information, formulate personalized shopping recommendations and service push strategies, and push personalized content to customers through multiple channels;

[0048] Feedback collection module: collect user feedback through various feedback channels, organize and analyze the collected feedback, and identify customer concerns and improvement directions;

[0049] Optimization and iteration module: Based on user feedback and AI analysis results, the service and operation strategies of supermarkets are continuously optimized and iterated.

[0050] Beneficial effects of the present invention: By deploying IOT devices in supermarkets, various data in supermarkets, such as product information, customer flow, inventory status, etc., can be collected in real time, thereby improving the timeliness and accuracy of the data. The data is integrated and preprocessed through the Internet of Things gateway to ensure the quality and consistency of the data, providing a reliable data foundation for subsequent AI analysis. Using AI algorithms to perform intelligent analysis of data can provide in-depth insights and predictions for supermarket management and help managers make more effective decisions. By automating the execution of management strategies and dynamically adjusting them according to the execution status, the efficiency and flexibility of supermarket management are improved. Based on customer behavior data and preference information, personalized shopping recommendations and services are provided to enhance the customer's shopping experience. By collecting user feedback and AI analysis results, the service and operation strategies of supermarkets are continuously optimized to enhance the competitiveness of supermarkets. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a step diagram of the method of the present invention;

[0052] Figure 2 This is a module diagram of the platform described in the present invention. DETAILED DESCRIPTION

[0053] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. The embodiments described are only a part of the embodiments of the present invention, rather than 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.

[0055] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0056] One embodiment of the present invention, as Figure 1 As shown, a method for interconnected and synchronized supermarket management based on AI and IOT, the method comprising:

[0057] S1. Deploy IoT devices in supermarkets and collect relevant data (environmental data, customer behavior data, and commodity flow data) in real time based on the deployed IoT devices (sensors, cameras, RFID tags, and smart door locks);

[0058] S2. Integrate the collected real-time data through the IoT gateway, pre-process the integrated data, and transfer the pre-processed data to the cloud space;

[0059] S3, intelligently analyze the data in the cloud space based on AI algorithms, and conduct interconnection and synchronization management of each link based on the AI ​​analysis results;

[0060] S4. Provide personalized services (shopping recommendations and discount information) to customers through multiple channels (smart displays and apps); collect user feedback and optimize supermarket management and services based on user feedback.

[0061] The working principle of the above technical solution is: IoT devices deployed in supermarkets (such as sensors, cameras, RFID tags, smart door locks, etc.) are the perception layer of the entire platform. These devices can collect various data in supermarkets in real time, including:

[0062] Environmental data: Through temperature and humidity sensors, air quality sensors and other equipment, the temperature, humidity, light intensity, air quality and other environmental parameters in the supermarket are monitored in real time to ensure the comfort and safety of the shopping environment.

[0063] Customer behavior data: Use cameras and facial recognition technology to capture and analyze customer behavior trajectories, dwell time, shopping preferences and other information in supermarkets to provide data support for personalized services.

[0064] Product flow data: Through RFID tags, smart shelves and other equipment, the movement trajectory, inventory changes, sales status, etc. of products can be tracked to help supermarkets achieve accurate inventory management and sales forecasts.

[0065] As the middleware of the platform, the IoT gateway is responsible for integrating data from different IoT devices to form a unified data flow. During the data integration process, the gateway will perform preliminary processing on the data, such as data cleaning (removing duplicate, erroneous or invalid data), data compression (reducing the burden of data transmission), etc. Subsequently, the pre-processed data will be securely transmitted to the cloud space for storage and analysis.

[0066] In the cloud, the collected data is deeply analyzed and mined based on AI algorithms. AI models can automatically identify patterns and trends in data, providing valuable insights for supermarket management. Based on AI analysis results, supermarkets can achieve interconnected and synchronous management of all links, including:

[0067] Environmental control: Based on the results of environmental data analysis, the temperature, humidity, lighting and other conditions in the supermarket are automatically adjusted to create a more comfortable shopping environment.

[0068] Inventory management: Based on the analysis of commodity flow data, sales trends are predicted, and replenishment processes are triggered in a timely manner to reduce out-of-stock and inventory backlogs.

[0069] Customer service: Utilize customer behavior data analysis to provide customers with personalized shopping recommendations and discount information, thereby improving customer satisfaction and purchase conversion rates.

[0070] Security management: Through smart door locks, cameras and other equipment, the security status of supermarkets can be monitored in real time, and abnormal behavior can be responded to immediately to ensure the safety of customers and employees.

[0071] Through multiple channels such as smart displays and apps, personalized shopping recommendations and discount information are presented to customers. These recommendations and discount information are customized based on the customer's shopping history and preferences, aiming to improve the customer's shopping experience and satisfaction. At the same time, supermarkets also collect user feedback through multiple channels, including customer opinions and suggestions on products, services, environment, etc. Based on user feedback, supermarkets can continuously optimize management and service processes, improve overall operational efficiency and customer satisfaction.

[0072] The effect of the above technical solution is: by deploying IoT devices, real-time data collection of the supermarket environment, customer behavior and commodity flow is realized. These data provide a comprehensive and timely information basis for supermarket management. Managers can make quick decisions based on real-time data, such as adjusting air conditioning temperature or lighting brightness according to environmental data, optimizing commodity display according to customer behavior data, and predicting inventory demand according to commodity flow data, thereby improving management efficiency and customer satisfaction.

[0073] As a data integration center, the IoT gateway processes and integrates data from different IoT devices in a unified manner, simplifying the data processing process. Through preprocessing and compression technology, the burden of data transmission is reduced and the efficiency of data transmission is improved. At the same time, the processed data is stored in the cloud space, which is convenient for subsequent intelligent analysis and large-scale application.

[0074] Based on AI algorithms, the data in the cloud space is deeply analyzed and mined to extract valuable information and patterns. The AI ​​analysis results provide scientific basis and decision support for supermarket management, such as predicting sales trends, identifying customer preferences, and discovering potential safety hazards. These analysis results can guide supermarkets to make more accurate and efficient management decisions.

[0075] Providing personalized shopping recommendations and discount information to customers through multiple channels such as smart displays and apps. This personalized service can enhance customers’ shopping experience and satisfaction, increase customer loyalty and repurchase rate. At the same time, personalized recommendations can also increase the sales conversion rate of goods and bring more profits to supermarkets.

[0076] Collect user feedback through multiple channels, including customers' opinions and suggestions on products, services, environment, etc. User feedback is an important source of continuous improvement for supermarkets. Based on user feedback, supermarkets can continuously optimize management and service processes, solve existing problems and deficiencies, and improve overall operational efficiency and customer satisfaction. This continuous optimization mechanism enables supermarkets to maintain their competitiveness and adapt to market changes.

[0077] In one embodiment of the present invention, the S1 includes:

[0078] S11. Analyze the operational needs of supermarkets and determine the required data types (such as environmental parameters, customer behavior, commodity flow, etc.); select the most appropriate IoT devices such as sensors, cameras, RFID tags, etc. based on the determined data types;

[0079] S12. Develop a deployment strategy for IoT devices based on the supermarket layout to ensure comprehensive data coverage without redundancy; install equipment according to the deployment strategy; ensure that the equipment is accurately positioned and securely fixed.

[0080] S13. Calibrate each device individually to ensure the accuracy of its measurement or monitoring data; set the data collection frequency, and perform preliminary data processing on the device side, the preliminary data processing including data compression and outlier filtering.

[0081] The working principle of the above technical solution is as follows: First, conduct an in-depth analysis of the operational needs of supermarkets to identify which data is critical to supermarket management. These data types may include environmental parameters (such as temperature, humidity, and light intensity), customer behavior (such as customer flow, dwell time, and shopping paths), and commodity flow (such as inventory changes and sales). According to the determined data type, select the most appropriate IoT device, such as sensors, cameras, RFID tags, etc. Each device has its specific application scenarios and advantages. Choosing the right device can ensure the accuracy and comprehensiveness of the data.

[0082] Develop a deployment strategy for IoT devices based on the layout of the supermarket. This includes determining the installation location, quantity, and coverage of the devices to ensure that data can fully cover all areas of the supermarket while avoiding data redundancy and waste. Install the devices according to the deployment strategy. During the installation process, it is necessary to ensure that the device is accurately positioned and firmly fixed so that data can be collected stably over the long term.

[0083] Each IoT device is calibrated individually to ensure the accuracy of its measurement or monitoring data. The calibration process may include adjusting the device's sensitivity, accuracy and other parameters to ensure the stability and reliability of the device under different environmental conditions. Preliminary data processing is performed on the device side, including data compression and outlier filtering. Data compression can reduce the burden of data transmission and the need for storage space; outlier filtering can remove erroneous data caused by device failure, environmental factors, etc., and improve the quality and reliability of data.

[0084] After device calibration and preliminary data processing are completed, the IoT device will start collecting data in real time according to the set data collection frequency. After the collected data is initially processed, it is integrated and transmitted through the IoT gateway. The IoT gateway is responsible for aggregating and formatting data from different devices, and then transmitting it to the cloud space through a secure network connection for storage and analysis.

[0085] The effect of the above technical solution is: through in-depth analysis of supermarket operation needs, the required data type can be accurately determined, so as to select the most appropriate IoT device for data collection. Ensure that the collected data is both comprehensive and targeted, and provide accurate data support for subsequent intelligent analysis and management.

[0086] Develop IoT device deployment strategies based on the supermarket layout to ensure comprehensive data coverage and no redundancy. Avoid blind and wasteful equipment installation, improve equipment efficiency and data collection accuracy. At the same time, a reasonable equipment layout also facilitates subsequent maintenance and management.

[0087] Each IoT device is calibrated individually to ensure the accuracy of its measurement or monitoring data, and a reasonable data collection frequency is set. This ensures the stability and reliability of data collection and reduces management decision errors caused by equipment failure or data errors. At the same time, a reasonable collection frequency can also reduce the burden of data transmission and processing while ensuring data timeliness.

[0088] Perform preliminary data processing on the device side, including data compression and outlier filtering. Data compression reduces the bandwidth requirements and storage space occupied by data transmission; outlier filtering improves the quality and reliability of data. These preliminary processing measures provide a cleaner and more effective data foundation for subsequent data analysis and application.

[0089] The entire S1 phase workflow is scientific and reasonable, ensuring the efficient deployment and stable operation of IoT devices. It lays a solid foundation for the intelligent management of supermarkets and improves the overall operational efficiency of supermarkets. Through real-time and accurate data collection and analysis, supermarket managers can understand the operating status in a timely manner and make more scientific and reasonable decisions, thereby improving customer satisfaction and the competitiveness of supermarkets.

[0090] In one embodiment of the present invention, the S13 includes:

[0091] S131. Use multiple calibration sources (such as standard equipment, known environmental parameters, etc.) to cross-calibrate IoT devices. According to the performance changes during the use of the device, trigger the automatic calibration program based on the preset threshold and dynamically adjust the calibration parameters to maintain the accuracy of the device.

[0092] S132. Set differentiated data collection frequencies for different types of IoT devices based on the specific needs of supermarket operations (such as peak period monitoring, specific event response, etc.);

[0093] S133. Analyze historical data through machine learning algorithms, predict data change trends based on the analysis results, and dynamically adjust the data collection frequency;

[0094] S134. Through compression algorithms, the collected data is compressed in real time on the device side to reduce the data transmission bandwidth requirements while retaining key information; combined with statistical methods and machine learning models, outliers caused by equipment failures, environmental interference, etc. are automatically identified and filtered out;

[0095] S135. Pre-label and classify the initially processed data according to data characteristics and supermarket operation requirements.

[0096] The working principle of the above technical solution is as follows: First, IoT devices are cross-calibrated using multiple calibration sources (such as standard equipment, known environmental parameters, etc.) to ensure the measurement consistency of the device under different conditions. Then, according to the changes in the performance of the device during use (such as accuracy drift, extended response time, etc.), the platform will automatically trigger the calibration procedure based on the preset threshold. This procedure will dynamically adjust the calibration parameters to compensate for the changes in device performance, thereby maintaining the measurement accuracy of the device.

[0097] Based on the specific needs of supermarket operations (such as monitoring customer flow during peak hours, tracking sales data during promotions for specific products, etc.), the platform sets differentiated data collection frequencies for different types of IoT devices. This differentiated setting is designed to ensure that sufficient data can be obtained to support decision-making during critical periods, while avoiding wasting resources during non-critical periods.

[0098] By analyzing historical data through machine learning algorithms, the platform can identify patterns and trends in data changes. Based on these analysis results, the platform can predict future data trends and dynamically adjust the frequency of data collection accordingly. For example, when it is predicted that the sales peak is approaching, the platform will automatically increase the frequency of data collection for related IoT devices to obtain more detailed data to support decision-making.

[0099] On the device side, the platform uses compression algorithms to compress the collected data in real time to reduce the demand for data transmission bandwidth. At the same time, key information is retained during the compression process to ensure that the data still has sufficient analytical value after compression. In addition, the platform also combines statistical methods and machine learning models to automatically identify and filter out outliers caused by equipment failures, environmental interference, etc. If these outliers are not eliminated, they may mislead subsequent data analysis.

[0100] According to the data characteristics and supermarket operation requirements, the platform pre-labels and classifies the data after preliminary processing. Pre-labeling refers to adding descriptive tags or metadata to the data for subsequent data analysis and processing. Classification is to divide the data into different categories or groups according to certain rules or standards. This process helps to improve the efficiency of data organization and facilitates subsequent data mining and decision support.

[0101] The effect of the above technical solution is: by using multiple calibration sources for cross-calibration and dynamically adjusting calibration parameters based on performance changes during device use, it can ensure that IoT devices always maintain high-precision measurement capabilities. This is crucial for scenarios that require accurate data in supermarket operations (such as inventory management, environmental monitoring, etc.).

[0102] Differentiated data collection frequencies are set for different types of IoT devices according to the specific needs of supermarket operations to avoid unnecessary waste of resources. During peak periods or specific events, increasing the frequency of data collection can ensure timely acquisition of key information to support decision-making.

[0103] Through machine learning algorithms, historical data is analyzed to predict data change trends and dynamically adjust the data collection frequency accordingly. This intelligent adjustment strategy can further optimize data collection efficiency while ensuring the timeliness of data.

[0104] The collected data is compressed in real time on the device side, significantly reducing the bandwidth required for data transmission. This not only reduces the cost of data transmission, but also improves the speed and stability of data transmission. At the same time, key information is retained during the compression process, ensuring that the analytical value of the data is not affected.

[0105] Combine statistical methods and machine learning models to automatically identify and filter out outliers caused by equipment failure, environmental interference, etc. This step is crucial to improving data quality because outliers may mislead data analysis results. By eliminating outliers, the accuracy and reliability of subsequent analysis results can be ensured.

[0106] Pre-label and classify the preliminarily processed data according to the data characteristics and supermarket operation requirements. This step facilitates subsequent data management and analysis, making the data more orderly, easy to retrieve and understand. At the same time, pre-labeling and classification also help improve the efficiency and accuracy of data analysis.

[0107] In one embodiment of the present invention, the S133 includes:

[0108] Use time series analysis algorithms to decompose historical data collected by IoT devices and identify seasonal, trend, and random components;

[0109] Extract key features from the decomposed data, such as periodic fluctuations, trend growth rates, abnormal fluctuation points, etc., and construct a feature vector set based on the extracted key features;

[0110] Select a machine learning model, such as a time series prediction model (ARIMA, LSTM, etc.), a regression model, or a classification model, based on data characteristics and target requirements (such as predictive data collection requirements); train the selected machine learning model using the extracted feature vector set as training data;

[0111] Input the latest IoT device data into the trained machine learning model for real-time prediction. Based on the model prediction results and the specific needs of supermarket operations (such as peak period monitoring, specific event response, etc.), calculate and predict the most appropriate data collection frequency.

[0112] According to the prediction results and preset thresholds (such as data change rate, prediction error range, etc.), set the conditions for triggering the adjustment of data collection frequency; when the adjustment conditions are met, the data collection frequency adjustment program is automatically triggered to dynamically adjust the collection frequency of IoT devices.

[0113] Collect the data collection results after adjustment, evaluate the adjustment effect, and continuously optimize the machine learning model based on the feedback results.

[0114] The working principle of the above technical solution is as follows: First, the historical data collected by IoT devices is decomposed using time series analysis algorithms (such as STL decomposition, seasonal adjustment, etc.) to identify seasonal, trend and random components in the data. This step helps to deeply understand the inherent structure and change patterns of the data. Key features are extracted from the decomposed data, such as periodic fluctuations, trend growth rates, abnormal fluctuation points, etc. These features can reflect the main change patterns and potential laws of the data, and provide important information for the subsequent training of machine learning models.

[0115] Select an appropriate machine learning model based on data characteristics and target requirements (such as forecasting data collection requirements). In the field of time series forecasting, commonly used models include time series forecasting models such as ARIMA and LSTM, as well as regression models or classification models (if the problem can be converted into a regression or classification problem). Use the extracted feature vector set as training data to train the selected machine learning model. During the training process, the model will learn the laws and patterns in the data and try to build a model that can accurately predict future data changes.

[0116] Input the latest IoT device data into the trained machine learning model for real-time prediction. The prediction results will reflect the changing trend of the data and the level that may be reached in the future. Based on the model prediction results, combined with the specific needs of supermarket operations (such as peak period monitoring, specific event response, etc.), calculate and predict the most appropriate data collection frequency. This step is to ensure that sufficient data can be obtained to support decision-making during critical periods, while avoiding wasting resources during non-critical periods.

[0117] According to the prediction results and preset thresholds (such as data change rate, prediction error range, etc.), set the conditions for triggering the adjustment of data collection frequency. When the adjustment conditions are met, the adjustment program of data collection frequency is automatically triggered to dynamically adjust the collection frequency of IoT devices.

[0118] Collect the data collection results after adjustment and evaluate whether the adjustment effect has achieved the expected goal. Evaluation indicators may include data collection efficiency, data quality, and support for supermarket operation decisions. Continuously optimize the machine learning model based on the feedback results. This includes adjusting model parameters, improving feature extraction methods, introducing new data sources, etc. to improve the prediction accuracy and generalization ability of the model.

[0119] The effect of the above technical solution is: through the combination of time series analysis algorithm and machine learning model, the future change trend of IoT device data can be accurately predicted. This makes the setting of data collection frequency more scientific and reasonable, avoiding the waste of resources caused by blind collection. The data collection frequency is dynamically adjusted according to the prediction results and the specific needs of supermarket operations. During peak periods or specific events, by increasing the frequency of data collection, key information can be obtained in a timely manner to provide support for decision-making; during non-critical periods, the collection frequency can be appropriately reduced to save resources.

[0120] Extracting key features from the decomposed data and constructing feature vector sets provide rich and valuable information for the training of machine learning models. This helps the model to more accurately capture the laws and patterns of data changes, thereby improving the accuracy of predictions. In the feature extraction process, outliers caused by equipment failures, environmental interference, etc. can be automatically identified and processed, thereby avoiding the interference of outliers on the prediction results and ensuring the quality of the data.

[0121] Set the conditions that trigger the adjustment of data collection frequency and automatically trigger the adjustment procedure. This automated process greatly reduces the burden of manual intervention and improves the level of intelligent management. Collect and evaluate the data collection effect after adjustment. Continuously optimize the machine learning model based on the evaluation results to enable it to better adapt to changes in data and changes in supermarket operation needs.

[0122] Through accurate prediction and dynamic adjustment of data collection frequency, timely and accurate data support is provided for supermarket operations. These data can help management make more scientific and reasonable decisions, thereby improving operational efficiency and service quality. By optimizing data collection frequency and reducing unnecessary resource waste, the operating costs of supermarkets can be reduced to a certain extent. At the same time, high-quality data can also bring more business opportunities and revenue sources to enterprises.

[0123] In one embodiment of the present invention, the S2 includes:

[0124] S21. Collect data streams from different IoT devices through the IoT gateway and integrate the data through data integration tools (such as Kafka, Flume, etc.) to form a unified data stream;

[0125] S22, preprocessing the integrated data, wherein the preprocessing includes cleaning, removing duplicate, erroneous or invalid data; and performing integrity and consistency checks on the data through a data verification mechanism;

[0126] S23, extract features from the preprocessed data, identify features useful for subsequent analysis, apply feature selection algorithms, remove redundant or irrelevant features, and reduce data dimensions;

[0127] S24. Encrypt the pre-processed data using an encryption algorithm, and transmit the encrypted data to the cloud space through a security protocol (such as HTTPS, TLS).

[0128] The working principle of the above technical solution is as follows: the IoT gateway acts as a data distribution center, responsible for receiving data streams from different IoT devices (such as sensors, cameras, RFID tags, etc.). These devices send data through their own communication protocols, and the IoT gateway acts as an intermediate bridge to receive data in various formats in a unified manner. Through data integration tools (such as Kafka, Flume, etc.), the IoT gateway integrates data streams from different devices to form a unified data stream. These tools have the characteristics of high throughput and low latency, which can ensure efficient transmission and processing of data.

[0129] Perform preliminary processing on the integrated data to remove duplicate, erroneous or invalid data. This includes identifying and deleting duplicate records, correcting erroneous data values, and eliminating incomplete or non-compliant data. Through the data verification mechanism, the cleaned data is checked for integrity and consistency. This helps ensure the accuracy and reliability of the data and provides a high-quality data foundation for subsequent analysis.

[0130] Identify features useful for subsequent analysis from preprocessed data. These features may be numerical (such as environmental parameters such as temperature and humidity) or non-numerical (such as customer behavior patterns, commodity flow trends, etc.). Apply feature selection algorithms to remove redundant or irrelevant features to reduce data dimensions and reduce the complexity of subsequent analysis. This process helps improve analysis efficiency and accuracy.

[0131] Encrypt the data after preprocessing and feature extraction. This helps protect the confidentiality and privacy of the data and prevents unauthorized access or tampering during transmission. The encrypted data is transmitted from the IoT gateway to the cloud space through security protocols (such as HTTPS and TLS). These protocols provide security mechanisms such as encryption, authentication, and integrity verification during data transmission to ensure the security and reliability of data during transmission.

[0132] The effect of the above technical solution is: through the collaborative work of IoT gateways and data integration tools (such as Kafka, Flume, etc.), data streams from different IoT devices can be efficiently collected and integrated to form a unified data stream. This efficient data integration method helps reduce the complexity and delay of data processing and provides timely and accurate data support for subsequent data analysis and management.

[0133] The integrated data is preprocessed, including cleaning, removing duplicate, erroneous or invalid data, and checking integrity and consistency through a data verification mechanism. This process significantly improves the quality of the data and reduces analysis errors and management decision-making errors caused by data problems. High-quality data lays a solid foundation for subsequent intelligent analysis and management.

[0134] Perform feature extraction on the preprocessed data to identify features that are useful for subsequent analysis, and use feature selection algorithms to remove redundant or irrelevant features to reduce data dimensions. The optimized data dimensions not only reduce the complexity and amount of calculations required for data processing, but also improve the efficiency and accuracy of data analysis. At the same time, reducing data dimensions also helps discover potential patterns and rules in the data, providing deeper insights for supermarket management.

[0135] The pre-processed data is encrypted through encryption algorithms and transmitted to the cloud space through security protocols (such as HTTPS and TLS). This measure effectively ensures the security and privacy of data during transmission and storage, and prevents the risk of data leakage and illegal access. At the same time, encrypted transmission also enhances the integrity and reliability of data, ensuring the accuracy and consistency of data.

[0136] After processing in the S2 stage, the data becomes more standardized, accurate, and easy to analyze. This provides high-quality input for subsequent AI algorithm analysis, making the analysis results more accurate and reliable. Supermarket managers can make more accurate decisions and management based on these analysis results, thereby improving the operational efficiency and customer satisfaction of supermarkets.

[0137] In one embodiment of the present invention, S3 includes:

[0138] S31. Use historical data and business rules to train the AI ​​model, and perform multiple iterations to optimize the trained AI model;

[0139] S32, allocating the data in the cloud space to the corresponding AI models for analysis according to different dimensions (such as time, space, and data type), and processing multiple analysis tasks in parallel for each analysis task;

[0140] S33. Conduct in-depth analysis of input data through AI models and output deep insight results (such as customer shopping preferences, inventory change trends, safety hazard warnings, etc.); conduct predictive analysis based on historical data and current trends to provide suggestions for supermarket management;

[0141] S34. Formulate interconnected and synchronized management strategies for each link (such as inventory replenishment plans, environmental control plans, safety monitoring strategies, etc.) based on AI analysis results, automatically execute management strategies, and dynamically adjust strategies based on execution status.

[0142] The working principle of the above technical solution is: use historical data and business rules as training sets to conduct preliminary training of AI models. These historical data may include customer shopping records, inventory changes, environmental parameters, etc., while business rules reflect the specific requirements and logic of supermarket operations. The preliminarily trained AI model is optimized for multiple iterations. This includes adjusting model parameters according to the performance of the model, improving the algorithm structure or introducing new features, etc., to improve the accuracy and generalization ability of the model.

[0143] The data in the cloud space is divided according to different dimensions (such as time, space, and data type) and assigned to the corresponding AI models for analysis. This division helps to achieve parallel processing of data and improve the efficiency and speed of data analysis. Each analysis task is processed in parallel, that is, multiple analysis tasks are started at the same time, and multi-core processors or distributed computing resources are used to accelerate the processing process. This can significantly shorten the time of data analysis and improve the real-time nature of supermarket management.

[0144] Through the trained AI model, the input data is deeply analyzed to discover the patterns and trends behind the data. These analyses may involve customer shopping preferences, inventory change trends, safety hazard warnings, and other aspects. Based on historical data and current trends, predictive analysis is performed to provide forward-looking suggestions for supermarket management. These suggestions may include future inventory demand forecasts, customer flow forecasts, safety hazard warnings, etc., which help supermarket managers prepare and respond in advance.

[0145] Develop interconnected and synchronized management strategies for each link based on the results of AI analysis. These strategies may include inventory replenishment plans, environmental control plans, safety monitoring strategies, etc., aimed at optimizing the operational efficiency and customer experience of supermarkets; these management strategies are automatically executed to reduce manual intervention and errors. At the same time, the strategies are dynamically adjusted and optimized based on feedback from the execution to ensure the continued effectiveness and adaptability of supermarket management.

[0146] The effect of the above technical solution is that through the in-depth learning and iterative optimization of historical data and business rules by AI models, the laws and trends in supermarket operations can be captured more accurately. This enables supermarket managers to make management decisions based on more scientific and accurate data analysis results, thereby improving the effectiveness and success rate of decision-making.

[0147] Processing multiple analysis tasks in parallel can significantly shorten the time for data processing and analysis and improve operational efficiency. Supermarket managers can obtain the required information and insights more quickly, thus responding to market changes and customer needs more promptly and optimizing resource allocation and business processes.

[0148] AI models can deeply explore the patterns and trends behind the data, output deep insights, and conduct predictive analysis. This provides supermarket managers with a more comprehensive and in-depth perspective to examine supermarket operations, helping them identify potential problems and opportunities and develop more accurate management strategies.

[0149] Management strategies developed based on AI analysis results can be automatically executed and dynamically adjusted based on the execution situation. This reduces manual intervention and errors, and improves management efficiency and accuracy. At the same time, dynamic adjustment strategies can ensure that supermarket management always keeps pace with actual conditions and adapts to market changes and changes in customer needs.

[0150] By gaining a deeper understanding of customers’ shopping preferences and behavior patterns, supermarkets can provide more accurate and personalized services and recommendations.

[0151] This helps improve customer satisfaction and loyalty, increase repeat customers and word-of-mouth communication, and bring more business opportunities and profits to supermarkets.

[0152] Through accurate inventory management and forecasting analysis, supermarkets can reduce inventory backlogs and waste, and reduce inventory costs. At the same time, by optimizing environmental control and safety monitoring strategies, supermarkets can reduce energy consumption and the risk of safety accidents, and further reduce operating costs.

[0153] In one embodiment of the present invention, the S32 includes:

[0154] Clarify the dimensions of data division; label each piece of data with a corresponding dimension label; build an efficient data indexing platform based on dimension labels;

[0155] Use AI algorithms to analyze the historical processing capacity and current load of each AI model, and assign data to the most suitable model for analysis; set priorities for analysis tasks based on the importance and urgency of the data;

[0156] Dynamically monitor the processing capacity and task queue length of each AI model and automatically adjust the data allocation strategy;

[0157] Break down large analytical tasks into multiple small tasks that can be processed in parallel, each of which corresponds to a subset of the data or a specific dimension;

[0158] Allocate computing resources to parallel tasks based on the resource status of the cloud platform (such as CPU, memory, storage, etc.); start parallel analysis tasks and monitor the task execution status in real time, including progress, resource consumption, exception handling, etc.

[0159] Collect the results of each parallel analysis task and perform preliminary data integration; use data verification algorithms to check whether the integrated data is consistent;

[0160] According to the verification results, the abnormal or inconsistent data is corrected; the performance indicators such as the execution efficiency and resource utilization of this parallel analysis task are evaluated; the evaluation results are fed back to the data allocation and parallel analysis strategies to continuously optimize the data allocation and task scheduling strategies.

[0161] The working principle of the above technical solution is as follows: First, according to business needs and data characteristics, the dimensions of data division, such as time, space, data type, etc., are clarified. These dimensions are helpful for subsequent data management and analysis. Each piece of data is labeled with a corresponding dimension, and an efficient data indexing platform is built based on these labels. This can speed up data query and access speed and improve data retrieval efficiency.

[0162] Use AI algorithms to analyze the historical processing capacity and current load of each AI model to evaluate the performance status of each model. Set priorities for analysis tasks based on the importance and urgency of the data. Ensure that critical data is processed first to meet the timeliness of business needs.

[0163] Dynamically monitor the processing capacity and task queue length of each AI model and automatically adjust the data allocation strategy. When a model is overloaded, some tasks will be allocated to other idle or high-capacity models to achieve load balancing.

[0164] Break down large analysis tasks into multiple small tasks that can be processed in parallel. Each small task corresponds to a subset of the data or a specific dimension so that they can be executed in parallel. Allocate sufficient computing resources to parallel tasks based on the resource status of the cloud platform (such as CPU, memory, storage, etc.). After starting the parallel analysis task, monitor the task execution status in real time, including progress, resource consumption, exception handling, etc.

[0165] Collect the results of each parallel analysis task and perform preliminary data integration. This step combines the scattered analysis results into a unified data set. Use data verification algorithms to check whether the integrated data is consistent. Ensuring the accuracy of analysis results is an important part of the data processing process. Correct abnormal or inconsistent data to optimize the final analysis results. At the same time, evaluate the performance indicators of this parallel analysis task, such as execution efficiency and resource utilization.

[0166] Feedback the evaluation results to the data allocation and parallel analysis strategies. Based on the feedback results, the data allocation and task scheduling strategies are continuously optimized to improve the overall analysis efficiency. This closed-loop feedback mechanism helps to continuously improve the data processing process and enhance the intelligence level of the platform.

[0167] The effect of the above technical solution is: by clarifying the dimensions of data division and labeling the data with dimension tags, combined with an efficient data indexing platform, the speed of data query and access is significantly accelerated. This makes it possible to quickly locate the required information in massive data, greatly improving the efficiency of data processing.

[0168] AI algorithms are used to analyze the historical processing capacity and current load of each AI model to achieve intelligent data allocation. This allocation method ensures that data is allocated to the most suitable model for processing, improving the accuracy and efficiency of analysis. The processing capacity and task queue length of each AI model are dynamically monitored, and the data allocation strategy is automatically adjusted to achieve load balancing. This avoids the situation where a single model is overloaded while other models are idle, and improves resource utilization.

[0169] Break down large analysis tasks into multiple small tasks that can be processed in parallel, making full use of the computing resources of the cloud platform. Each small task is executed independently without interfering with each other, which significantly shortens the overall analysis time. Allocate sufficient computing resources to parallel tasks based on the resource status of the cloud platform to ensure the smooth execution of tasks. Monitor the execution status of tasks in real time, discover and solve potential problems in a timely manner, and improve the stability and reliability of parallel processing.

[0170] The results of each parallel analysis task are collected and preliminarily integrated, and the data verification algorithm is used to check whether the integrated data is consistent. This step ensures the accuracy and reliability of the analysis results and avoids decision-making errors caused by inconsistent data. Abnormal or inconsistent data is corrected to optimize the final analysis results. This further improves the accuracy and availability of the data and provides more reliable data support for business decisions.

[0171] The performance indicators of this parallel analysis task, such as execution efficiency and resource utilization, are evaluated, and the evaluation results are fed back to the data allocation and parallel analysis strategy. This closed-loop feedback mechanism helps to continuously optimize the data allocation and task scheduling strategies and improve the overall analysis efficiency. Based on the feedback results, the data processing process and analysis strategy are continuously optimized, so that the platform can continuously adapt to changes in business needs and the growth of data volume. This ability to continuously improve makes the S32 technical solution more flexible and scalable.

[0172] In one embodiment of the present invention, the S4 includes:

[0173] S41. Develop personalized shopping recommendations and service push strategies based on customer behavior data and preference information, and push personalized content to customers through multiple channels such as smart displays and apps;

[0174] S42. Collect user feedback through various feedback channels (such as online surveys, in-app feedback, social media, etc.), organize and analyze the collected feedback, and identify customer concerns and improvement directions;

[0175] S43. Continuously optimize and iterate the service and operation strategies of supermarkets based on user feedback and AI analysis results.

[0176] The working principle of the above technical solution is as follows: First, the platform collects customers' shopping behavior data (such as purchase history, browsing history, search keywords, etc.) and preference information (such as product preferences, price sensitivity, shopping time, etc.). Based on these data, AI algorithms are used to analyze customers' shopping habits and preferences, and formulate personalized shopping recommendations and service push strategies. Personalized shopping recommendations, discount information, service reminders and other content are pushed to customers through multiple channels such as smart displays, APPs, text messages, and emails.

[0177] Supermarkets provide a variety of feedback channels, such as online surveys, in-app feedback functions, and social media interactions, so that customers can provide feedback anytime and anywhere. The platform automatically collects these feedbacks, including customers' evaluations and suggestions on products, services, environment, etc. The collected feedback is sorted and analyzed using AI technologies such as natural language processing and sentiment analysis to identify customer concerns, satisfaction levels, and improvement directions.

[0178] Combine AI analysis results and user feedback to evaluate the current supermarket service and operation strategies and identify existing problems and deficiencies. Based on the evaluation results, formulate targeted optimization plans to adjust and improve service and operation strategies. This may include product structure adjustment, service process optimization, environmental improvement and other aspects. Iterate and implement the optimized strategy in the actual operation of the supermarket and continuously monitor its effect. Through continuous trial and error and iteration, gradually improve the service quality and operation efficiency of the supermarket.

[0179] The technical solution of the S4 stage forms a closed-loop management platform. Starting from the improvement of personalized customer experience, by collecting and analyzing user feedback, the service and operation strategies of supermarkets are continuously optimized, and these optimized strategies are applied to actual operations, and feedback is continuously collected for a new round of optimization. This closed-loop process is cyclical, promoting the continuous improvement and development of supermarkets.

[0180] The effect of the above technical solution is: the personalized shopping recommendation strategy based on customer behavior data and preference information can accurately push the goods and services that customers are interested in, and improve the pertinence and satisfaction of shopping. Pushing personalized content to customers through multiple channels such as smart displays and apps increases the convenience and timeliness of customers' access to information and improves the overall shopping experience of customers. Personalized content push not only meets the shopping needs of customers, but also enhances the interaction between customers and supermarkets, and improves customer loyalty and stickiness.

[0181] Through personalized recommendations and service push, supermarkets can more accurately target customer groups, improve the effectiveness and conversion rate of marketing activities, and reduce marketing costs. Based on user feedback and AI analysis results, supermarkets can timely adjust product structure, optimize inventory management, improve service processes, etc., to achieve optimal allocation and efficient use of resources. The timely collection and analysis of user feedback provides valuable data support for supermarket management, helping them make decisions more quickly and respond to market changes.

[0182] By sorting and analyzing user feedback, supermarkets can clearly identify customer concerns and improvement directions, providing clear goals for improving service quality. Based on user feedback and AI analysis results, supermarkets can continuously optimize and iterate service and operation strategies to form a virtuous cycle of continuous improvement. By continuously improving service quality and operational efficiency, supermarkets can stand out in the fierce market competition and enhance their competitiveness and market position.

[0183] The establishment of multiple feedback channels gives customers the opportunity to participate in the improvement process of supermarkets, enhancing their sense of participation and belonging. Personalized shopping experience, timely feedback response and continuous service quality improvement all help to improve customer loyalty and make customers more willing to choose the supermarket for long-term shopping.

[0184] One embodiment of the present invention, as Figure 2 As shown, an interconnected and synchronized supermarket management platform based on AI and IOT, the platform includes:

[0185] Data collection module: Deploy IoT devices in supermarkets and collect relevant data (environmental data, customer behavior data, and commodity flow data) in real time based on the deployed IoT devices (sensors, cameras, RFID tags, and smart door locks);

[0186] Data processing module: Integrate the collected real-time data through the IoT gateway, pre-process the integrated data, and transmit the pre-processed data to the cloud space;

[0187] Data analysis module: intelligently analyzes the data in the cloud space based on AI algorithms, and performs interconnected and synchronous management of each link based on the AI ​​analysis results;

[0188] Service optimization module: Provide personalized services (shopping recommendations and discount information) to customers through multiple channels (smart displays and apps); collect user feedback and optimize supermarket management and services based on user feedback.

[0189] The working principle of the above technical solution is: IoT devices deployed in supermarkets (such as sensors, cameras, RFID tags, smart door locks, etc.) are the perception layer of the entire platform. These devices can collect various data in supermarkets in real time, including:

[0190] Environmental data: Through temperature and humidity sensors, air quality sensors and other equipment, the temperature, humidity, light intensity, air quality and other environmental parameters in the supermarket are monitored in real time to ensure the comfort and safety of the shopping environment.

[0191] Customer behavior data: Use cameras and facial recognition technology to capture and analyze customer behavior trajectories, dwell time, shopping preferences and other information in supermarkets to provide data support for personalized services.

[0192] Product flow data: Through RFID tags, smart shelves and other equipment, the movement trajectory, inventory changes, sales status, etc. of products can be tracked to help supermarkets achieve accurate inventory management and sales forecasts.

[0193] As the middleware of the platform, the IoT gateway is responsible for integrating data from different IoT devices to form a unified data flow. During the data integration process, the gateway will perform preliminary processing on the data, such as data cleaning (removing duplicate, erroneous or invalid data), data compression (reducing the burden of data transmission), etc. Subsequently, the pre-processed data will be securely transmitted to the cloud space for storage and analysis.

[0194] In the cloud, the collected data is deeply analyzed and mined based on AI algorithms. AI models can automatically identify patterns and trends in data, providing valuable insights for supermarket management. Based on AI analysis results, supermarkets can achieve interconnected and synchronous management of all links, including:

[0195] Environmental control: Based on the results of environmental data analysis, the temperature, humidity, lighting and other conditions in the supermarket are automatically adjusted to create a more comfortable shopping environment.

[0196] Inventory management: Based on the analysis of commodity flow data, sales trends are predicted, and replenishment processes are triggered in a timely manner to reduce out-of-stock and inventory backlogs.

[0197] Customer service: Utilize customer behavior data analysis to provide customers with personalized shopping recommendations and discount information, thereby improving customer satisfaction and purchase conversion rates.

[0198] Security management: Through smart door locks, cameras and other equipment, the security status of supermarkets can be monitored in real time, and abnormal behavior can be responded to immediately to ensure the safety of customers and employees.

[0199] Through multiple channels such as smart displays and apps, personalized shopping recommendations and discount information are presented to customers. These recommendations and discount information are customized based on the customer's shopping history and preferences, aiming to improve the customer's shopping experience and satisfaction. At the same time, supermarkets also collect user feedback through multiple channels, including customer opinions and suggestions on products, services, environment, etc. Based on user feedback, supermarkets can continuously optimize management and service processes, improve overall operational efficiency and customer satisfaction.

[0200] The effect of the above technical solution is: by deploying IoT devices, real-time data collection of the supermarket environment, customer behavior and commodity flow is realized. These data provide a comprehensive and timely information basis for supermarket management. Managers can make quick decisions based on real-time data, such as adjusting air conditioning temperature or lighting brightness according to environmental data, optimizing commodity display according to customer behavior data, and predicting inventory demand according to commodity flow data, thereby improving management efficiency and customer satisfaction.

[0201] As a data integration center, the IoT gateway processes and integrates data from different IoT devices in a unified manner, simplifying the data processing process. Through preprocessing and compression technology, the burden of data transmission is reduced and the efficiency of data transmission is improved. At the same time, the processed data is stored in the cloud space, which is convenient for subsequent intelligent analysis and large-scale application.

[0202] Based on AI algorithms, the data in the cloud space is deeply analyzed and mined to extract valuable information and patterns. The AI ​​analysis results provide scientific basis and decision support for supermarket management, such as predicting sales trends, identifying customer preferences, and discovering potential safety hazards. These analysis results can guide supermarkets to make more accurate and efficient management decisions.

[0203] Providing personalized shopping recommendations and discount information to customers through multiple channels such as smart displays and apps. This personalized service can enhance customers’ shopping experience and satisfaction, increase customer loyalty and repurchase rate. At the same time, personalized recommendations can also increase the sales conversion rate of goods and bring more profits to supermarkets.

[0204] Collect user feedback through multiple channels, including customers' opinions and suggestions on products, services, environment, etc. User feedback is an important source of continuous improvement for supermarkets. Based on user feedback, supermarkets can continuously optimize management and service processes, solve existing problems and deficiencies, and improve overall operational efficiency and customer satisfaction. This continuous optimization mechanism enables supermarkets to maintain their competitiveness and adapt to market changes.

[0205] In one embodiment of the present invention, the data acquisition module includes:

[0206] Type determination module: Analyzes supermarket operation needs and determines the required data types (such as environmental parameters, customer behavior, commodity flow, etc.); selects the most appropriate IoT devices such as sensors, cameras, RFID tags, etc. based on the determined data types;

[0207] Equipment installation module: formulate deployment strategies for IoT devices based on the supermarket layout to ensure comprehensive data coverage and no redundancy; install equipment according to the deployment strategy; ensure that the equipment is accurately positioned and firmly fixed.

[0208] Abnormal filtering module: calibrate each device individually to ensure the accuracy of its measurement or monitoring data; set the data collection frequency, and perform preliminary data processing on the device side, which includes data compression and abnormal value filtering.

[0209] The working principle of the above technical solution is as follows: First, conduct an in-depth analysis of the operational needs of supermarkets to identify which data is critical to supermarket management. These data types may include environmental parameters (such as temperature, humidity, and light intensity), customer behavior (such as customer flow, dwell time, and shopping paths), and commodity flow (such as inventory changes and sales). According to the determined data type, select the most appropriate IoT device, such as sensors, cameras, RFID tags, etc. Each device has its specific application scenarios and advantages. Choosing the right device can ensure the accuracy and comprehensiveness of the data.

[0210] Develop a deployment strategy for IoT devices based on the layout of the supermarket. This includes determining the installation location, quantity, and coverage of the devices to ensure that data can fully cover all areas of the supermarket while avoiding data redundancy and waste. Install the devices according to the deployment strategy. During the installation process, it is necessary to ensure that the device is accurately positioned and firmly fixed so that data can be collected stably over the long term.

[0211] Each IoT device is calibrated individually to ensure the accuracy of its measurement or monitoring data. The calibration process may include adjusting the device's sensitivity, accuracy and other parameters to ensure the stability and reliability of the device under different environmental conditions. Preliminary data processing is performed on the device side, including data compression and outlier filtering. Data compression can reduce the burden of data transmission and the need for storage space; outlier filtering can remove erroneous data caused by device failure, environmental factors, etc., and improve the quality and reliability of data.

[0212] After device calibration and preliminary data processing are completed, the IoT device will start collecting data in real time according to the set data collection frequency. After the collected data is initially processed, it is integrated and transmitted through the IoT gateway. The IoT gateway is responsible for aggregating and formatting data from different devices, and then transmitting it to the cloud space through a secure network connection for storage and analysis.

[0213] The effect of the above technical solution is: through in-depth analysis of supermarket operation needs, the required data type can be accurately determined, so as to select the most appropriate IoT device for data collection. Ensure that the collected data is both comprehensive and targeted, and provide accurate data support for subsequent intelligent analysis and management.

[0214] Develop IoT device deployment strategies based on the supermarket layout to ensure comprehensive data coverage and no redundancy. Avoid blind and wasteful equipment installation, improve equipment efficiency and data collection accuracy. At the same time, a reasonable equipment layout also facilitates subsequent maintenance and management.

[0215] Each IoT device is calibrated individually to ensure the accuracy of its measurement or monitoring data, and a reasonable data collection frequency is set. This ensures the stability and reliability of data collection and reduces management decision errors caused by equipment failure or data errors. At the same time, a reasonable collection frequency can also reduce the burden of data transmission and processing while ensuring data timeliness.

[0216] Perform preliminary data processing on the device side, including data compression and outlier filtering. Data compression reduces the bandwidth requirements and storage space occupied by data transmission; outlier filtering improves the quality and reliability of data. These preliminary processing measures provide a cleaner and more effective data foundation for subsequent data analysis and application.

[0217] The entire S1 phase workflow is scientific and reasonable, ensuring the efficient deployment and stable operation of IoT devices. It lays a solid foundation for the intelligent management of supermarkets and improves the overall operational efficiency of supermarkets. Through real-time and accurate data collection and analysis, supermarket managers can understand the operating status in a timely manner and make more scientific and reasonable decisions, thereby improving customer satisfaction and the competitiveness of supermarkets.

[0218] In one embodiment of the present invention, the data processing module includes:

[0219] Data integration module: collects data streams from different IoT devices through the IoT gateway, and integrates the data through data integration tools (such as Kafka, Flume, etc.) to form a unified data stream;

[0220] Data verification module: pre-processes the integrated data, including cleaning to remove duplicate, erroneous or invalid data; and verifies the integrity and consistency of the data through a data verification mechanism;

[0221] Irrelevant removal module: extract features from preprocessed data, identify features useful for subsequent analysis, apply feature selection algorithms to remove redundant or irrelevant features, and reduce data dimensions;

[0222] Data encryption module: Encrypts the pre-processed data through encryption algorithms, and transmits the encrypted data to the cloud space through security protocols (such as HTTPS, TLS).

[0223] The working principle of the above technical solution is as follows: the IoT gateway acts as a data distribution center, responsible for receiving data streams from different IoT devices (such as sensors, cameras, RFID tags, etc.). These devices send data through their own communication protocols, and the IoT gateway acts as an intermediate bridge to receive data in various formats in a unified manner. Through data integration tools (such as Kafka, Flume, etc.), the IoT gateway integrates data streams from different devices to form a unified data stream. These tools have the characteristics of high throughput and low latency, which can ensure efficient transmission and processing of data.

[0224] Perform preliminary processing on the integrated data to remove duplicate, erroneous or invalid data. This includes identifying and deleting duplicate records, correcting erroneous data values, and eliminating incomplete or non-compliant data. Through the data verification mechanism, the cleaned data is checked for integrity and consistency. This helps ensure the accuracy and reliability of the data and provides a high-quality data foundation for subsequent analysis.

[0225] Identify features useful for subsequent analysis from preprocessed data. These features may be numerical (such as environmental parameters such as temperature and humidity) or non-numerical (such as customer behavior patterns, commodity flow trends, etc.). Apply feature selection algorithms to remove redundant or irrelevant features to reduce data dimensions and reduce the complexity of subsequent analysis. This process helps improve analysis efficiency and accuracy.

[0226] Encrypt the data after preprocessing and feature extraction. This helps protect the confidentiality and privacy of the data and prevents unauthorized access or tampering during transmission. The encrypted data is transmitted from the IoT gateway to the cloud space through security protocols (such as HTTPS and TLS). These protocols provide security mechanisms such as encryption, authentication, and integrity verification during data transmission to ensure the security and reliability of data during transmission.

[0227] The effect of the above technical solution is: through the collaborative work of IoT gateways and data integration tools (such as Kafka, Flume, etc.), data streams from different IoT devices can be efficiently collected and integrated to form a unified data stream. This efficient data integration method helps reduce the complexity and delay of data processing and provides timely and accurate data support for subsequent data analysis and management.

[0228] The integrated data is preprocessed, including cleaning, removing duplicate, erroneous or invalid data, and checking integrity and consistency through a data verification mechanism. This process significantly improves the quality of the data and reduces analysis errors and management decision-making errors caused by data problems. High-quality data lays a solid foundation for subsequent intelligent analysis and management.

[0229] Perform feature extraction on the preprocessed data to identify features that are useful for subsequent analysis, and use feature selection algorithms to remove redundant or irrelevant features to reduce data dimensions. The optimized data dimensions not only reduce the complexity and amount of calculations required for data processing, but also improve the efficiency and accuracy of data analysis. At the same time, reducing data dimensions also helps discover potential patterns and rules in the data, providing deeper insights for supermarket management.

[0230] The pre-processed data is encrypted through encryption algorithms and transmitted to the cloud space through security protocols (such as HTTPS and TLS). This measure effectively ensures the security and privacy of data during transmission and storage, and prevents the risk of data leakage and illegal access. At the same time, encrypted transmission also enhances the integrity and reliability of data, ensuring the accuracy and consistency of data.

[0231] After this stage of processing, the data becomes more standardized, accurate and easy to analyze. This provides high-quality input for subsequent AI algorithm analysis, making the analysis results more accurate and reliable. Supermarket managers can make more accurate decisions and management based on these analysis results, thereby improving the operational efficiency and customer satisfaction of supermarkets.

[0232] In one embodiment of the present invention, the data analysis module includes:

[0233] Iterative optimization module: Use historical data and business rules to train the AI ​​model, and perform multiple iterative optimizations on the trained AI model;

[0234] Parallel processing module: distributes the data in the cloud space to the corresponding AI models for analysis according to different dimensions (such as time, space, and data type), and processes multiple analysis tasks in parallel for each analysis task;

[0235] Predictive analysis module: Use AI models to conduct in-depth analysis of input data and output in-depth insights (such as customer shopping preferences, inventory change trends, safety hazard warnings, etc.); perform predictive analysis based on historical data and current trends to provide suggestions for supermarket management;

[0236] Dynamic adjustment module: formulate interconnected and synchronous management strategies for each link (such as inventory replenishment plans, environmental control plans, safety monitoring strategies, etc.) based on AI analysis results, automatically execute management strategies, and dynamically adjust strategies based on execution status.

[0237] The working principle of the above technical solution is: use historical data and business rules as training sets to conduct preliminary training of AI models. These historical data may include customer shopping records, inventory changes, environmental parameters, etc., while business rules reflect the specific requirements and logic of supermarket operations. The preliminarily trained AI model is optimized for multiple iterations. This includes adjusting model parameters according to the performance of the model, improving the algorithm structure or introducing new features, etc., to improve the accuracy and generalization ability of the model.

[0238] The data in the cloud space is divided according to different dimensions (such as time, space, and data type) and assigned to the corresponding AI models for analysis. This division helps to achieve parallel processing of data and improve the efficiency and speed of data analysis. Each analysis task is processed in parallel, that is, multiple analysis tasks are started at the same time, and multi-core processors or distributed computing resources are used to accelerate the processing process. This can significantly shorten the time of data analysis and improve the real-time nature of supermarket management.

[0239] Through the trained AI model, the input data is deeply analyzed to discover the patterns and trends behind the data. These analyses may involve customer shopping preferences, inventory change trends, safety hazard warnings, and other aspects. Based on historical data and current trends, predictive analysis is performed to provide forward-looking suggestions for supermarket management. These suggestions may include future inventory demand forecasts, customer flow forecasts, safety hazard warnings, etc., which help supermarket managers prepare and respond in advance.

[0240] Develop interconnected and synchronized management strategies for each link based on the results of AI analysis. These strategies may include inventory replenishment plans, environmental control plans, safety monitoring strategies, etc., aimed at optimizing the operational efficiency and customer experience of supermarkets; these management strategies are automatically executed to reduce manual intervention and errors. At the same time, the strategies are dynamically adjusted and optimized based on feedback from the execution to ensure the continued effectiveness and adaptability of supermarket management.

[0241] The effect of the above technical solution is that through the in-depth learning and iterative optimization of historical data and business rules by AI models, the laws and trends in supermarket operations can be captured more accurately. This enables supermarket managers to make management decisions based on more scientific and accurate data analysis results, thereby improving the effectiveness and success rate of decision-making.

[0242] Processing multiple analysis tasks in parallel can significantly shorten the time for data processing and analysis and improve operational efficiency. Supermarket managers can obtain the required information and insights more quickly, thus responding to market changes and customer needs more promptly and optimizing resource allocation and business processes.

[0243] AI models can deeply explore the patterns and trends behind the data, output deep insights, and conduct predictive analysis. This provides supermarket managers with a more comprehensive and in-depth perspective to examine supermarket operations, helping them identify potential problems and opportunities and develop more accurate management strategies.

[0244] Management strategies developed based on AI analysis results can be automatically executed and dynamically adjusted based on the execution situation. This reduces manual intervention and errors, and improves management efficiency and accuracy. At the same time, dynamic adjustment strategies can ensure that supermarket management always keeps pace with actual conditions and adapts to market changes and changes in customer needs.

[0245] By gaining a deeper understanding of customers’ shopping preferences and behavior patterns, supermarkets can provide more accurate and personalized services and recommendations.

[0246] This helps improve customer satisfaction and loyalty, increase repeat customers and word-of-mouth communication, and bring more business opportunities and profits to supermarkets.

[0247] Through accurate inventory management and forecasting analysis, supermarkets can reduce inventory backlogs and waste, and reduce inventory costs. At the same time, by optimizing environmental control and safety monitoring strategies, supermarkets can reduce energy consumption and the risk of safety accidents, and further reduce operating costs.

[0248] In one embodiment of the present invention, the service optimization module includes:

[0249] Content push module: Based on customer behavior data and preference information, formulate personalized shopping recommendations and service push strategies, and push personalized content to customers through multiple channels such as smart displays and apps;

[0250] Feedback collection module: collect user feedback through various feedback channels (such as online surveys, in-app feedback, social media, etc.), organize and analyze the collected feedback, and identify customer concerns and improvement directions;

[0251] Optimization and iteration module: Based on user feedback and AI analysis results, the service and operation strategies of supermarkets are continuously optimized and iterated.

[0252] The working principle of the above technical solution is as follows: First, the platform collects customers' shopping behavior data (such as purchase history, browsing history, search keywords, etc.) and preference information (such as product preferences, price sensitivity, shopping time, etc.). Based on these data, AI algorithms are used to analyze customers' shopping habits and preferences, and formulate personalized shopping recommendations and service push strategies. Personalized shopping recommendations, discount information, service reminders and other content are pushed to customers through multiple channels such as smart displays, APPs, text messages, and emails.

[0253] Supermarkets provide a variety of feedback channels, such as online surveys, in-app feedback functions, and social media interactions, so that customers can provide feedback anytime and anywhere. The platform automatically collects these feedbacks, including customers' evaluations and suggestions on products, services, environment, etc. The collected feedback is sorted and analyzed using AI technologies such as natural language processing and sentiment analysis to identify customer concerns, satisfaction levels, and improvement directions.

[0254] Combine AI analysis results and user feedback to evaluate the current supermarket service and operation strategies and identify existing problems and deficiencies. Based on the evaluation results, formulate targeted optimization plans to adjust and improve service and operation strategies. This may include product structure adjustment, service process optimization, environmental improvement and other aspects. Iterate and implement the optimized strategy in the actual operation of the supermarket and continuously monitor its effect. Through continuous trial and error and iteration, gradually improve the service quality and operation efficiency of the supermarket.

[0255] The technical solution at this stage forms a closed-loop management platform. Starting from the improvement of customer personalized experience, by collecting and analyzing user feedback, the service and operation strategies of supermarkets are continuously optimized, and these optimized strategies are applied to actual operations, and feedback is continuously collected for a new round of optimization. This closed-loop process is cyclical, promoting the continuous improvement and development of supermarkets.

[0256] The effect of the above technical solution is: the personalized shopping recommendation strategy based on customer behavior data and preference information can accurately push the goods and services that customers are interested in, and improve the pertinence and satisfaction of shopping. Pushing personalized content to customers through multiple channels such as smart displays and apps increases the convenience and timeliness of customers' access to information and improves the overall shopping experience of customers. Personalized content push not only meets the shopping needs of customers, but also enhances the interaction between customers and supermarkets, and improves customer loyalty and stickiness.

[0257] Through personalized recommendations and service push, supermarkets can more accurately target customer groups, improve the effectiveness and conversion rate of marketing activities, and reduce marketing costs. Based on user feedback and AI analysis results, supermarkets can timely adjust product structure, optimize inventory management, improve service processes, etc., to achieve optimal allocation and efficient use of resources. The timely collection and analysis of user feedback provides valuable data support for supermarket management, helping them make decisions more quickly and respond to market changes.

[0258] By sorting and analyzing user feedback, supermarkets can clearly identify customer concerns and improvement directions, providing clear goals for improving service quality. Based on user feedback and AI analysis results, supermarkets can continuously optimize and iterate service and operation strategies to form a virtuous cycle of continuous improvement. By continuously improving service quality and operational efficiency, supermarkets can stand out in the fierce market competition and enhance their competitiveness and market position.

[0259] The establishment of multiple feedback channels gives customers the opportunity to participate in the improvement process of supermarkets, enhancing their sense of participation and belonging. Personalized shopping experience, timely feedback response and continuous service quality improvement all help to improve customer loyalty and make customers more willing to choose the supermarket for long-term shopping.

[0260] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for interconnected and synchronized supermarket management based on AI and IOT, characterized in that: The method comprises: S1. Deploy IOT devices in supermarkets and collect relevant data in supermarkets in real time based on the deployed IOT devices; S2. Integrate the collected real-time data through the IoT gateway, pre-process the integrated data, and transfer the pre-processed data to the cloud space; S3, intelligently analyze the data in the cloud space based on AI algorithms, and conduct interconnection and synchronization management of each link based on the AI ​​analysis results; S4. Provide personalized services to customers through multiple channels; collect user feedback and optimize supermarket management and services based on user feedback; The S3 includes: S31. Use historical data and business rules to train the AI ​​model, and perform multiple iterations to optimize the trained AI model; S32, allocating the data in the cloud space to the corresponding AI models for analysis according to different dimensions, and processing multiple analysis tasks in parallel for each analysis task; S33. Conduct in-depth analysis of input data through AI models and output deep insight results; conduct predictive analysis based on historical data and current trends to provide suggestions for supermarket management; S34. Formulate interconnection and synchronization management strategies for each link based on AI analysis results, automatically execute management strategies, and dynamically adjust strategies based on execution status; The S32 includes: Clarify the dimensions of data division; label each piece of data with a corresponding dimension label; and build a data index platform based on the dimension label; Use AI algorithms to analyze the historical processing capacity and current load of each AI model, and assign data to the model for analysis; set priorities for analysis tasks based on the importance and urgency of the data; Dynamically monitor the processing capacity and task queue length of each AI model and automatically adjust the data allocation strategy; Break down large analytical tasks into multiple small tasks that can be processed in parallel, each of which corresponds to a subset of the data or a specific dimension; Allocate computing resources to parallel tasks based on the resource status of the cloud platform; start parallel analysis tasks and monitor the task execution status in real time; Collect the results of each parallel analysis task and perform preliminary data integration; use data verification algorithms to check whether the integrated data is consistent; Based on the verification results, correct abnormal or inconsistent data; evaluate the performance indicators of this parallel analysis task; feed the evaluation results back to the data allocation and parallel analysis strategy to continuously optimize the data allocation and task scheduling strategy.

2. According to claim 1, the interconnected synchronous supermarket management method based on AI and IOT is characterized in that: Said S1 comprises: S11. Analyze the operation requirements of supermarkets and determine the required data types; select IoT devices based on the determined data types; S12. Formulate a deployment strategy for IoT devices based on the supermarket layout diagram, and install the devices according to the deployment strategy; S13. Calibrate each device individually, set the data collection frequency, and perform preliminary data processing on the device side, wherein the preliminary data processing includes data compression and outlier filtering.

3. According to claim 1, the interconnected synchronous supermarket management method based on AI and IOT Internet of Things is characterized in that: The S2 comprises: S21. Collect data streams from different IoT devices through the IoT gateway and integrate the data through data integration tools to form a unified data stream; S22, preprocessing the integrated data, wherein the preprocessing includes cleaning, removing duplicate, erroneous or invalid data; and performing integrity and consistency checks on the data through a data verification mechanism; S23, extracting features from the preprocessed data, identifying features useful for subsequent analysis, and applying a feature selection algorithm to remove redundant or irrelevant features; S24. Encrypt the pre-processed data using an encryption algorithm, and transmit the encrypted data to the cloud space through a security protocol.

4. According to claim 1, the interconnected synchronous supermarket management method based on AI and IOT is characterized in that: The S4 comprises: S41. Develop personalized shopping recommendations and service push strategies based on customer behavior data and preference information, and push personalized content to customers through multiple channels; S42. Collect user feedback through various feedback channels, organize and analyze the collected feedback, and identify customer concerns and improvement directions; S43. Continuously optimize and iterate the service and operation strategies of supermarkets based on user feedback and AI analysis results.

5. An interconnected and synchronized supermarket management platform based on AI and IOT, characterized in that: The platform includes: Data collection module: Deploy IoT devices in supermarkets and collect relevant data in supermarkets in real time based on the deployed IoT devices; Data processing module: Integrate the collected real-time data through the IoT gateway, pre-process the integrated data, and transmit the pre-processed data to the cloud space; Data analysis module: intelligently analyzes the data in the cloud space based on AI algorithms, and performs interconnected and synchronous management of each link based on the AI ​​analysis results; Service optimization module: Provide personalized services to customers through multiple channels; collect user feedback and optimize supermarket management and services based on user feedback; The data analysis module comprises: Iterative optimization module: Use historical data and business rules to train the AI ​​model, and perform multiple iterative optimizations on the trained AI model; Parallel processing module: distributes the data in the cloud space to the corresponding AI models for analysis according to different dimensions, and processes multiple analysis tasks in parallel for each analysis task; Predictive analysis module: In-depth analysis of input data through AI models to output deep insight results; predictive analysis based on historical data and current trends to provide suggestions for supermarket management; Dynamic adjustment module: formulate interconnection and synchronization management strategies for each link based on AI analysis results, automatically execute management strategies, and dynamically adjust strategies based on execution status; The processing steps of the parallel processing module include: Clarify the dimensions of data division; label each piece of data with a corresponding dimension label; and build a data index platform based on the dimension label; Use AI algorithms to analyze the historical processing capacity and current load of each AI model, and assign data to the model for analysis; set priorities for analysis tasks based on the importance and urgency of the data; Dynamically monitor the processing capacity and task queue length of each AI model and automatically adjust the data allocation strategy; Break down large analytical tasks into multiple small tasks that can be processed in parallel, each of which corresponds to a subset of the data or a specific dimension; Allocate computing resources to parallel tasks based on the resource status of the cloud platform; start parallel analysis tasks and monitor the task execution status in real time; Collect the results of each parallel analysis task and perform preliminary data integration; use data verification algorithms to check whether the integrated data is consistent; Based on the verification results, correct abnormal or inconsistent data; evaluate the performance indicators of this parallel analysis task; feed the evaluation results back to the data allocation and parallel analysis strategy to continuously optimize the data allocation and task scheduling strategy.

6. According to claim 5, the interconnected and synchronized supermarket management platform based on AI and IOT is characterized in that: The data acquisition module comprises: Type determination module: Analyze supermarket operation requirements and determine the required data types; select IoT devices based on the determined data types; Equipment installation module: formulate deployment strategies for IoT devices based on the supermarket layout diagram, and install devices according to the deployment strategies; Abnormal filtering module: calibrate each device individually, set the data collection frequency, and perform preliminary data processing on the device side, including data compression and abnormal value filtering.

7. The interconnected and synchronized supermarket management platform based on AI and IOT according to claim 5 is characterized in that: The data processing module comprises: Data integration module: collects data streams from different IoT devices through the IoT gateway and integrates the data through data integration tools to form a unified data stream; Data verification module: pre-processes the integrated data, including cleaning to remove duplicate, erroneous or invalid data; and verifies the integrity and consistency of the data through a data verification mechanism; Irrelevant removal module: extracts features from preprocessed data, identifies features useful for subsequent analysis, and applies feature selection algorithms to remove redundant or irrelevant features; Data encryption module: Encrypt the pre-processed data through encryption algorithm, and transmit the encrypted data to the cloud space through security protocol.

8. According to claim 5, the interconnected and synchronized supermarket management platform based on AI and IOT is characterized in that: The service optimization module comprises: Content push module: Based on customer behavior data and preference information, formulate personalized shopping recommendations and service push strategies, and push personalized content to customers through multiple channels; Feedback collection module: collect user feedback through various feedback channels, organize and analyze the collected feedback, and identify customer concerns and improvement directions; Optimization and iteration module: Based on user feedback and AI analysis results, the service and operation strategies of supermarkets are continuously optimized and iterated.

Images