A collaborative management method and system of Internet of Things and POS machine based on 5G chip

Abstract

The present invention discloses a method and system for collaborative management of the Internet of Things and POS machines based on a 5G chip, and relates to the technical field of collaborative management of the Internet of Things and POS machines, including establishing a connection between an Internet of Things device and a POS machine through a 5G network, collecting real-time data of the Internet of Things device and the POS machine and performing preprocessing; transmitting the collected Internet of Things device data to the POS machine through the 5G network for edge computing, and performing inventory demand forecasting and executing response measures according to the edge computing results. The present invention establishes an efficient connection through a 5G network, collects and preprocesses data of the Internet of Things device and the POS machine in real time, uses edge computing to forecast inventory demand, and realizes real-time processing and analysis of data of the Internet of Things device and the POS machine through encrypted transmission and visual report generation, reduces the burden on the central server, improves the timeliness of data processing and the accuracy of decision-making, as well as the efficiency and accuracy of inventory management, enhances data security, and meets market changes and customer needs.

Description

Technical Field

[0001] The present invention relates to the technical field of collaborative management of the Internet of Things and POS machines, and in particular to a collaborative management method and system of the Internet of Things and POS machines based on a 5G chip. Background Art

[0002] With the development of Internet of Things technology and 5G networks, the interconnection between smart devices has become more convenient and efficient. In the retail industry, POS machines, as core devices for sales data and transaction information, have gradually merged with Internet of Things devices to form the prototype of smart retail systems. The high speed, low latency and wide coverage of 5G technology make the data transmission between Internet of Things devices and POS machines more real-time and reliable, providing technical support for the realization of smart applications such as real-time inventory management, customer behavior analysis and environmental monitoring. However, the existing technical solutions still have many shortcomings in terms of data transmission security, real-time and processing efficiency. The traditional POS data processing system relies on wired networks or Wi-Fi networks, and the data transmission speed is slow and susceptible to interference, making it difficult to meet real-time requirements. The existing Internet of Things and POS systems mostly adopt a centralized data processing mode, and the inventory management of POS machines cannot be dynamically adjusted according to real-time data. The data processing and feedback speed is slow, which affects the real-time and accuracy of decision-making. The traditional encrypted transmission method also has obvious shortcomings in the face of increasingly complex network attacks. Summary of the invention

[0003] In view of the problems existing in the above-mentioned existing 5G chip-based IoT and POS machine collaborative management method and system, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is that the traditional POS data processing system relies on wired networks or Wi-Fi networks, the data transmission speed is slow and easily interfered with, and it is difficult to meet the real-time requirements. The existing Internet of Things and POS systems mostly adopt a centralized data processing mode, and the inventory management of the POS machines cannot be dynamically adjusted according to real-time data. The data processing and feedback speeds are slow, which affects the real-time and accuracy of decision-making. The traditional encrypted transmission method also has obvious shortcomings in the face of increasingly complex network attacks.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: a collaborative management method of the Internet of Things and POS machines based on a 5G chip, which includes establishing a connection between an Internet of Things device and a POS machine through a 5G network, collecting real-time data of the Internet of Things device and the POS machine and performing preprocessing; transmitting the collected Internet of Things device data to the POS machine through the 5G network for edge computing, predicting inventory demand based on the edge computing results and executing response measures; encrypting the data analysis results and transmitting them to a central server, generating a visual report for storage.

[0006] As a preferred solution of the collaborative management method of the Internet of Things and POS machines based on 5G chips described in the present invention, wherein: the establishment of a connection between the Internet of Things device and the POS machine through the 5G network refers to configuring network settings on the POS machine and the Internet of Things device, selecting and connecting to the 5G network, using a network diagnostic tool to detect the connection status of the device, installing a blockchain node on the POS machine after the connection is established, adding the blockchain node to the newly created blockchain network, and synchronizing the latest data through the blockchain network.

[0007] As a preferred solution of the collaborative management method of the Internet of Things and POS machines based on 5G chips described in the present invention, wherein: the real-time data of the Internet of Things devices and POS machines are collected and pre-processed, which refers to collecting inventory data through sensors, monitoring customer behavior through smart cameras and RFID tags, recording customer stay time and number of interactions, and recording the transaction amount, commodity type and sales time after the transaction is completed from the POS machine in real time, and marking the transaction ID with a unique identifier, storing the transaction data in a local database, cleaning the collected data, and removing outliers using the 3 times standard deviation method;

[0008] De-duplicate the data based on its timestamp and unique identifier, and standardize the cleaned data;

[0009] Align inventory data, customer behavior data, and transaction data based on timestamps, aggregate the aligned data, and generate comprehensive data records.

[0010] As a preferred solution of the collaborative management method of the Internet of Things and POS machines based on 5G chips described in the present invention, wherein: the transmitting the collected Internet of Things device data to the POS machine through the 5G network for edge computing refers to transmitting the collected Internet of Things device data to the POS machine through the 5G network at a predetermined time interval. After receiving the data from the Internet of Things device, the POS machine stores the data in a local database and calculates the real-time inventory consumption based on the received inventory data and transaction data:

[0011] P=C+VM,

[0012] Among them, P is inventory consumption, C is the initial inventory quantity, V is the purchase quantity, and M is the sales quantity;

[0013] Calculate the customer behavior index based on customer behavior data:

[0014] ,

[0015] Among them, D is the customer behavior index, F is the customer's stay time in the store, I is the number of interactions between customers and products, and B is the total number of visits.

[0016] As a preferred solution of the collaborative management method of the Internet of Things and POS machines based on 5G chips described in the present invention, the inventory demand forecasting and response measures based on edge computing results include:

[0017] After data cleaning and standardization of the calculated inventory consumption data and customer behavior index, the standardized data is divided into different channels to process the inventory consumption data and customer behavior index respectively, and the separated inventory consumption data and customer behavior index are divided into multiple time series segments in chronological order using the sliding window technology;

[0018] Construct a bidirectional LSTM model as an inventory demand forecasting model and design the inventory demand forecasting model architecture to handle the time-dependent characteristics of inventory consumption and customer behavior data respectively;

[0019] Construct forward LSTM and backward LSTM, concatenate the outputs of forward and backward LSTM at each time step, generate the comprehensive hidden state of the time step and generate preliminary prediction values ​​through the fully connected layer;

[0020] The segmented time series data is input into the inventory demand forecasting model for model training. The mean square error is used as the loss function to evaluate the model prediction error. The weight and bias of the model are updated through the back propagation algorithm to obtain the inventory demand forecasting model that has been initially trained.

[0021] After the initial training of the model is completed, the parameters that need to be optimized are determined, the search space of the particle swarm is defined and initialized using the CRLPSO algorithm, and the inertia weight, acceleration constant, search range, and objective function are set;

[0022] Through the iteration of the CRLPSO algorithm, the weight and learning rate of the model are gradually adjusted. After multiple iterations, when the model error converges to the global optimal solution, the iteration is stopped and the optimized weight and learning rate are confirmed;

[0023] An attention layer is added to the inventory demand forecasting model. The attention mechanism is used to identify and process the time step that has the greatest impact on the forecast. The time step with the greatest impact is taken as the key time step and given a higher weight. After all time steps are weighted, the implicit state of the key time step is generated and an adaptive discrete wavelet transform is performed to generate enhanced samples:

[0024] ,

[0025] In the formula, It is the decomposition result of time series signal at multiple time scales. are the wavelet coefficients, is the mother wavelet function, s(t) is the adaptive scaling function of the time step, is a weighting function, which represents the importance weight of time step k at time t. is the time derivative of the customer behavior index, D is the customer behavior index, is the standard deviation of the customer behavior index D, which is used to measure the volatility of customer behavior changes. is the smoothing parameter, a and b are the upper and lower limits of the integral, indicating the time interval of the calculation, and dt represents a small time increment;

[0026] Generate enhanced samples with different characteristics based on high-frequency and low-frequency reconstructed signals and add noise to different scale components to generate new samples. Store the generated enhanced sample data as a new training data set and re-input it into the inventory demand forecasting model for model training. Continue to use the mean square error as the loss function, iterate and optimize the model parameters based on the Adam optimizer until the loss is minimized and then stop the iteration. Input the iterated model parameters into the inventory demand forecasting model to obtain the inventory demand forecasting model.

[0027] Input the inventory consumption and customer behavior index obtained by edge computing into the inventory demand forecasting model to obtain the inventory demand in the future period;

[0028] Generate an inventory demand forecast report based on the forecast results, including the future inventory demand quantity and time for each commodity, and automatically generate replenishment orders based on the generated forecast report and place replenishment orders.

[0029] As a preferred solution of the collaborative management method of the Internet of Things and POS machines based on 5G chips described in the present invention, the encrypted transmission of the data analysis results to the central server refers to adding noise to all data processed by edge computing to generate obfuscated data:

[0030] ,

[0031] Among them, R is the obfuscated data, Q is the original data, and w is the noise, which refers to the randomly generated disturbance value. is the confusion coefficient;

[0032] Divide the obfuscated data into blocks and mark each block with a serial number. Each block of data is 128 bits long. Use the AES-256 algorithm to encrypt each block of data to generate ciphertext.

[0033] Use the TLS protocol to establish a secure connection, and transmit the encrypted obfuscated data to the central server through the 5G network. After receiving the encrypted data, the central server uses the same key to decrypt the received data, sorts and synthesizes the decrypted block data according to the serial number order marked in the database, restores the complete obfuscated data, removes the noise in the synthesized obfuscated data, and restores it to the original data.

[0034] As a preferred solution of the collaborative management method of the Internet of Things and POS machines based on 5G chips described in the present invention, the generating of visual reports for storage refers to generating an inventory level chart based on the inventory data in the received data and the predicted inventory demand, generating a customer behavior analysis chart based on the customer behavior index and the interaction data, integrating the produced charts into the report, storing the generated reports in a central server and implementing secure access control on the stored data, backing up the stored data in the cloud and regularly checking the integrity of the stored data and the cloud backup data, and using blockchain technology to record all data changes.

[0035] Another object of the present invention is to provide a 5G chip-based Internet of Things and POS machine collaborative management system, which includes:

[0036] A data collection module is used to collect data from IoT devices and POS machines and pre-process the data after the connection between the IoT devices and POS machines is established through the 5G network;

[0037] Edge computing module, used to transmit IoT device data to POS machines through 5G networks for edge computing, calculating inventory consumption and customer behavior index;

[0038] Demand forecasting module, used to forecast inventory demand in real time based on edge computing results and replenish stocks based on the forecast results;

[0039] Data encryption module, used to encrypt and obfuscate data analysis results before transmitting them to the central server;

[0040] The data storage module is used to generate visual reports on the decrypted data and store them, and then perform access control and cloud backup on the data.

[0041] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of a method for collaborative management of the Internet of Things and POS machines based on a 5G chip when executing the computer program.

[0042] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for collaborative management of the Internet of Things and POS machines based on a 5G chip.

[0043] The beneficial effects of the present invention are as follows: the present invention establishes an efficient connection through the 5G network, collects and preprocesses the data of IoT devices and POS machines in real time, uses edge computing to predict inventory demand, and realizes real-time processing and analysis of IoT devices and POS machine data through encrypted transmission and visual report generation, thereby reducing the burden on the central server, improving the timeliness of data processing and the accuracy of decision-making, as well as the efficiency and accuracy of inventory management, enhancing data security, and meeting market changes and customer needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0045] Figure 1 The figure is a flowchart of the collaborative management method of the Internet of Things and POS machines based on 5G chips.

[0046] Figure 2 This is a structural diagram of the IoT and POS machine collaborative management system based on 5G chips. DETAILED DESCRIPTION

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0050] Example 1, reference Figure 1 , which is the first embodiment of the present invention, and provides a collaborative management method of the Internet of Things and POS machines based on a 5G chip. The collaborative management method of the Internet of Things and POS machines based on a 5G chip includes:

[0051] S1. Establish a connection between IoT devices and POS machines through the 5G network, collect real-time data from IoT devices and POS machines, and perform pre-processing;

[0052] Specifically, establishing a connection between IoT devices and POS machines through a 5G network means configuring network settings on the POS machines and IoT devices, selecting and connecting to the 5G network, using network diagnostic tools to detect the connection status of the devices, installing blockchain nodes on the POS machines after the connection is established, adding the blockchain nodes to the newly created blockchain network, and synchronizing the latest data through the blockchain network.

[0053] The high speed and low latency characteristics of the 5G network are utilized to ensure real-time connection and data transmission between IoT devices and POS machines. Through blockchain technology, the integrity and security of data during transmission are guaranteed. The introduction of the blockchain network provides data with non-tamperable and traceable guarantees, which not only improves the speed and reliability of data transmission, but also enhances data security and prevents data from being tampered with or lost during transmission. This lays a solid foundation for subsequent data collection, processing and analysis, and ensures the stable operation of the entire system.

[0054] Furthermore, collecting and preprocessing real-time data from IoT devices and POS machines refers to collecting inventory data through sensors, monitoring customer behavior through smart cameras and RFID tags, recording customer stay time and number of interactions, and recording the transaction amount, product type, and sales time after the transaction is completed from the POS machine in real time, and marking the transaction ID with a unique identifier, storing the transaction data in a local database, cleaning the collected data, and removing outliers using the 3-times standard deviation method;

[0055] De-duplicate the data based on its timestamp and unique identifier, and standardize the cleaned data;

[0056] Align inventory data, customer behavior data, and transaction data based on timestamps, aggregate the aligned data, and generate comprehensive data records.

[0057] Through a variety of sensors and technical means, we have achieved comprehensive collection of inventory, customer behavior and transaction data, and ensured the accuracy and consistency of the data through preprocessing steps. Data cleaning removes outliers, and standardization and data alignment ensure the availability of data in subsequent analysis. By systematically collecting and preprocessing data, the quality of the data and the accuracy of the analysis are improved, which not only helps to improve the accuracy of inventory management, but also better understand customer behavior, optimize sales strategies, and improve customer satisfaction. Comprehensive real-time data collection provides a reliable data foundation for subsequent edge computing and predictive analysis.

[0058] S2: The collected IoT device data is transmitted to the POS machine through the 5G network for edge computing, and inventory demand is predicted and response measures are executed based on the edge computing results;

[0059] Specifically, the edge computing of the POS terminal through the 5G network refers to the transmission of the collected IoT device data to the POS terminal through the 5G network at a predetermined time interval. After receiving the data from the IoT device, the POS terminal stores the data in the local database and calculates the real-time inventory consumption based on the received inventory data and transaction data:

[0060] P=C+VM,

[0061] Among them, P is inventory consumption, C is the initial inventory quantity, V is the purchase quantity, and M is the sales quantity;

[0062] Calculate the customer behavior index based on customer behavior data:

[0063] ,

[0064] Among them, D is the customer behavior index, F is the customer's stay time in the store, I is the number of interactions between customers and products, and B is the total number of visits.

[0065] Data is transmitted to the POS machine through the 5G network for edge computing, which makes full use of the computing power of the POS machine, reduces the burden on the central server, and improves the efficiency and timeliness of data processing. Calculating inventory consumption and customer behavior index can provide real-time decision support for inventory management and sales strategies, effectively shorten the delay of data processing, and improve the response speed and processing efficiency of the system. By calculating inventory consumption in real time, inventory can be adjusted more timely to avoid out-of-stock or excessive inventory. By calculating the customer behavior index, customer needs can be understood more accurately, product display and sales strategies can be optimized, and customer experience and sales performance can be improved.

[0066] Furthermore, inventory demand forecasting and response measures based on edge computing results include:

[0067] After data cleaning and standardization of the calculated inventory consumption data and customer behavior index, the standardized data is divided into different channels to process the inventory consumption data and customer behavior index respectively, and the separated inventory consumption data and customer behavior index are divided into multiple time series segments in chronological order using the sliding window technology;

[0068] Build a bidirectional LSTM model as an inventory demand forecasting model and design the inventory demand forecasting model architecture to handle the time-dependent characteristics of inventory consumption and customer behavior data respectively;

[0069] Construct forward LSTM and backward LSTM, concatenate the outputs of forward and backward LSTM at each time step, generate the comprehensive hidden state of the time step and generate preliminary prediction values ​​through the fully connected layer;

[0070] The segmented time series data is input into the inventory demand forecasting model for model training. The mean square error is used as the loss function to evaluate the model prediction error. The weight and bias of the model are updated through the back propagation algorithm to obtain the inventory demand forecasting model that has been initially trained.

[0071] After the initial training of the model is completed, the parameters that need to be optimized are determined, the search space of the particle swarm is defined and initialized using the CRLPSO algorithm, and the inertia weight, acceleration constant, search range, and objective function are set;

[0072] Through the iteration of the CRLPSO algorithm, the weight and learning rate of the model are gradually adjusted. After multiple iterations, when the model error converges to the global optimal solution, the iteration is stopped and the optimized weight and learning rate are confirmed;

[0073] An attention layer is added to the inventory demand forecasting model. The attention mechanism is used to identify and process the time steps that have the greatest impact on the forecast (promotional periods, holidays), and the time steps with the greatest impact are taken as key time steps and given higher weights. After all time steps are weighted, the implicit state of the key time step (the time step vector after attention weighting) is generated and an adaptive discrete wavelet transform is performed to generate enhanced samples:

[0074] ,

[0075] These implicit states represent the model’s understanding of the importance of each time step, rather than directly generating predictions. The weighted implicit states of the time steps will be used for subsequent retraining, helping the model to focus more on important time steps during the retraining process, further improving the accuracy of the model’s future predictions.

[0076] The original discrete wavelet transform formula can capture the different scale characteristics of the signal, but its application in inventory demand forecasting is limited because it cannot dynamically handle the nonlinear changes of customer behavior or inventory demand, and is not adaptable to changes in specific key time steps. In order to capture this dynamic change, an adaptive time scale function s(t) is introduced, which is dynamically adjusted according to time t:

[0077] ,

[0078] In actual inventory forecasting or customer behavior analysis, the importance of different time steps is different, so a weighted function is introduced , by dynamically assigning different weights to each time step, it can accurately capture and reflect the signal changes in a specific period of time:

[0079] ,

[0080] Changes in customer behavior or inventory demand are often nonlinear, but the traditional DWT formula does not take this nonlinear factor into account. Therefore, a nonlinear adjustment term is added:

[0081] ,

[0082] By adding this nonlinear adjustment term, the formula can capture the changes in the customer behavior index and reflect the impact of behavioral changes on inventory demand forecasts. The model can not only handle the dynamic changes in time steps, but also adjust according to the fluctuations in customer behavior;

[0083] In the existing DWT formula, the characteristics of the local time step are processed. In actual inventory forecasting or customer behavior analysis, the cumulative impact over a period of time needs to be considered. Therefore, an integral operation is added to capture the cumulative time dependency from time a to b, and the final improved discrete wavelet transform formula is obtained:

[0084] ,

[0085] In the formula, It is the decomposition result of time series signal at multiple time scales. is the wavelet coefficient, which is used to adjust the amplitude of the wavelet basis function and reflects the contribution of time step k to the signal. is the mother wavelet function, is the difference between time t and time step k after dynamic adjustment of the time scale s(t). s(t) is the adaptive scale function of the time step, reflecting the scale changes at different time steps. is a weighting function, which represents the importance weight of time step k at time t. is the time derivative of the customer behavior index, which indicates the rate of change of the customer behavior index over time. D is the customer behavior index. is the standard deviation of the customer behavior index D, which is used to measure the volatility of customer behavior changes. is the smoothing parameter, a and b are the upper and lower limits of the integral, indicating the time interval of the calculation, and dt represents a small time increment;

[0086] The mother wavelet function formula is:

[0087] ,

[0088] In the formula, is a sine function, often used to capture oscillatory behavior in signals. It is the Gaussian envelope of the wavelet basis function, which controls the decay speed of the wavelet transform;

[0089] The adaptive scaling function formula is:

[0090] ,

[0091] In the formula, is the initial scale, used to set the initial time scale benchmark, is the amplitude of adaptive adjustment, which is used to control the degree of change of time scale over time. It is the rate at which the time step changes and determines the speed of adaptive adjustment. It is the center point of the time step, which is used to determine the starting time for the time scale to be adjusted;

[0092] The weighting function formula is:

[0093] ,

[0094] In the formula, is the decay coefficient between time steps, which controls the rate of change of weights between time steps;

[0095] Through the integration operation, the formula can capture the changes in the entire time period, thereby more comprehensively reflecting the long-term dependencies between different time steps. This is particularly effective when it is necessary to analyze long-term trends and cumulative changes (such as seasonal fluctuations, long-term customer behavior changes, etc.);

[0096] Generate enhanced samples with different characteristics based on high-frequency and low-frequency reconstructed signals and add noise to different scale components to generate new samples. Store the generated enhanced sample data as a new training data set and re-input it into the inventory demand forecasting model for model training. Continue to use the mean square error as the loss function, iterate and optimize the model parameters based on the Adam optimizer until the loss is minimized and then stop the iteration. Input the iterated model parameters into the inventory demand forecasting model to obtain the inventory demand forecasting model.

[0097] Wavelet transform can generate more diverse data samples to capture feature changes at different time scales. The enhanced data samples generated by discrete wavelet transform will cover multiple time scales and frequency changes, making the data richer and more diverse. Through data enhancement, the model will be able to learn more time-dependent features and adapt to more changing scenarios, such as sudden changes in inventory demand and irregular customer behavior.

[0098] Input the inventory consumption and customer behavior index obtained by edge computing into the inventory demand forecasting model to obtain the inventory demand in the future period;

[0099] Generate an inventory demand forecast report based on the forecast results, including the future inventory demand quantity and time for each commodity, and automatically generate replenishment orders based on the generated forecast report and place replenishment orders.

[0100] The sliding window technology slices the inventory consumption data and customer behavior index according to the time series to generate multiple short-term time series segments for model training. This method can process large-scale time series data into multiple segments, which is convenient for model learning and training. It retains the sequential characteristics of the time series, helps the model capture the time dependency in the data, and enables the model to more accurately identify short-term changes and long-term trends when predicting. Through the design of forward and backward LSTM networks, it can handle the time dependency characteristics of inventory consumption and customer behavior and generate a comprehensive hidden state for each time step. This architectural design ensures that when the model processes the time series, it not only pays attention to past data, but also can use future information at the same time to make the prediction more accurate. This global time dependency analysis can significantly improve the overall performance of inventory demand forecasting. By defining the search space and setting the inertia weight, acceleration constant and other parameters of the particle swarm, the CRLPSO algorithm gradually optimizes the weight and learning rate of the model, and finally converges to the global optimal solution, which can effectively avoid the local optimal problem that may occur in traditional optimization algorithms. Through global search, the algorithm can find the optimal configuration in a complex parameter space, making the model more accurate and robust after multiple iterations. Introducing an attention layer into the inventory demand forecasting model, the key time steps with the greatest impact on the forecast are identified and given higher weights. Through this processing, the model can significantly enhance its attention to important time periods, such as promotional periods, holidays, and other time periods that have a special impact on inventory demand. The attention mechanism improves the model's ability to process complex time series data, reduces the model's error, and improves the accuracy of the forecast. Through adaptive discrete wavelet transform, the generated time series signal is decomposed at different time scales to generate enhanced samples, and noise is added to different components to further expand the sample space. The generated enhanced samples contain multi-scale features and dynamic noise, enriching the training data set, helping the model learn more complex time dependencies and uncertainties, and making it more adaptable when facing different changing scenarios.

[0101] S3, encrypting and transmitting the data analysis results to the central server and generating a visual report for storage;

[0102] Specifically, encrypting and transmitting the data analysis results to the central server means adding noise to all data processed by edge computing to generate obfuscated data:

[0103] ,

[0104] Among them, R is the obfuscated data, Q is the original data, and w is the noise, which refers to the randomly generated disturbance value. is the confusion coefficient;

[0105] The formula for calculating the confusion coefficient is:

[0106] ,

[0107] Among them, j is an adjustment factor, which is used to adjust the intensity of the noise according to the specific application scenario and experimental results. It is initially set to 1 and optimized according to the experimental results. is the standard deviation of the normally distributed noise;

[0108] In this embodiment, the noise is normally distributed noise, and the standard deviation of the noise is calculated:

[0109] ,

[0110] ,

[0111] in, is the standard deviation of the normally distributed noise, is the data sensitivity, q is the privacy budget parameter, which is determined according to the specific application scenario and privacy requirements, and its value range is [0.01, 1]. is the result of statistical analysis on the data set W, the statistical analysis includes but is not limited to sum, average, count and maximum value, etc., in this embodiment, it is a summation calculation, r(W') is the result of statistical analysis on the data set W', the data set W' is a data set with one set of user data less than the data set W;

[0112] Divide the obfuscated data into blocks and mark each block with a serial number. Each block of data is 128 bits long. Use the AES-256 algorithm to encrypt each block of data to generate ciphertext.

[0113] Use the TLS protocol to establish a secure connection, and transmit the encrypted obfuscated data to the central server through the 5G network. After receiving the encrypted data, the central server uses the same key to decrypt the received data, sorts and synthesizes the decrypted block data according to the sequence number sequence marked in the database, restores the complete obfuscated data, removes the noise in the synthesized obfuscated data, and restores it to the original data:

[0114] ,

[0115] Among them, R is the obfuscated data, Q is the original data, and w is the noise, which refers to the randomly generated disturbance value. is the confusion coefficient.

[0116] During the data transmission process, noise is added to all data processed by edge computing to generate obfuscated data, and the AES-256 algorithm is used to encrypt it, ensuring the security and integrity of the data during transmission. The specific operations include calculating the standard deviation of the noise, generating perturbation values ​​based on the normal distribution, and using the TLS protocol to establish a secure connection. The encrypted data is transmitted to the central server via the 5G network, effectively preventing the data from being intercepted or tampered with during transmission, ensuring the confidentiality and integrity of the data, and protecting the sensitive information of enterprises and customers.

[0117] Furthermore, generating visual reports for storage refers to generating inventory level graphs based on inventory data and predicted inventory demand in the received data, generating customer behavior analysis graphs based on customer behavior indexes and interaction data, integrating the produced graphs into reports, storing the generated reports in a central server and implementing secure access control on the stored data, backing up the stored data in the cloud and regularly checking the integrity of the stored data and the cloud backup data, and using blockchain technology to record all data changes.

[0118] By generating inventory level graphs based on inventory data and predicted inventory demand in the received data, generating customer behavior analysis graphs based on customer behavior indexes and interaction data, integrating these graphs into visual reports, and storing the generated reports in a central server, data visualization and intuitive presentation are ensured, enabling managers to quickly understand and analyze business situations. The storage and backup of reports, as well as the use of blockchain technology to record all data changes, further enhance data security and traceability, and provide reliable data management and decision-making support.

[0119] Example 2, reference Figure 2 , which is the second embodiment of the present invention, which is different from the previous embodiment and provides a 5G chip-based Internet of Things and POS machine collaborative management system, which includes:

[0120] A data collection module is used to collect data from IoT devices and POS machines and pre-process the data after the connection between the IoT devices and POS machines is established through the 5G network;

[0121] Edge computing module, used to transmit IoT device data to POS machines through 5G networks for edge computing, calculating inventory consumption and customer behavior index;

[0122] Demand forecasting module, used to forecast inventory demand in real time based on edge computing results and replenish stocks based on the forecast results;

[0123] Data encryption module, used to encrypt and obfuscate data analysis results before transmitting them to the central server;

[0124] The data storage module is used to generate visual reports on the decrypted data and store them, and then perform access control and cloud backup on the data.

[0125] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0126] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0127] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0128] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit with a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit with a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

Claims

1. A collaborative management method of the Internet of Things and POS machines based on 5G chips, characterized in that: include, Establish connections between IoT devices and POS machines through 5G networks, collect real-time data from IoT devices and POS machines, and perform pre-processing; The collected IoT device data is transmitted to the POS machine through the 5G network for edge computing, and inventory demand is predicted and response measures are executed based on the edge computing results; The transmission of the collected IoT device data to the POS machine through the 5G network for edge computing refers to transmitting the collected IoT device data to the POS machine through the 5G network at a predetermined time interval. After receiving the data from the IoT device, the POS machine stores the data in a local database and calculates the real-time inventory consumption based on the received inventory data and transaction data: , Among them, P is inventory consumption, C is the initial inventory quantity, V is the purchase quantity, and M is the sales quantity. Calculate the customer behavior index based on customer behavior data: , Among them, D is the customer behavior index, F is the customer's stay time in the store, I is the number of interactions between customers and products, and B is the total number of visits; The inventory demand forecasting and response measures based on edge computing results include: After data cleaning and standardization of the calculated inventory consumption data and customer behavior index, the standardized data is divided into different channels to process the inventory consumption data and customer behavior index respectively, and the separated inventory consumption data and customer behavior index are divided into multiple time series segments in chronological order using the sliding window technology; Construct a bidirectional LSTM model as an inventory demand forecasting model and design the inventory demand forecasting model architecture to handle the time-dependent characteristics of inventory consumption and customer behavior data respectively; Construct forward LSTM and backward LSTM, concatenate the outputs of forward and backward LSTM at each time step, generate the comprehensive hidden state of the time step and generate preliminary prediction values ​​through the fully connected layer; The segmented time series data is input into the inventory demand forecasting model for model training. The mean square error is used as the loss function to evaluate the model prediction error. The weight and bias of the model are updated through the back propagation algorithm to obtain the inventory demand forecasting model that has been initially trained. After the initial training of the model is completed, the parameters that need to be optimized are determined, the search space of the particle swarm is defined and initialized using the CRLPSO algorithm, and the inertia weight, acceleration constant, search range, and objective function are set; Through the iteration of the CRLPSO algorithm, the weight and learning rate of the model are gradually adjusted. After multiple iterations, when the model error converges to the global optimal solution, the iteration is stopped and the optimized weight and learning rate are confirmed; An attention layer is added to the inventory demand forecasting model. The attention mechanism is used to identify and process the time step that has the greatest impact on the forecast. The time step with the greatest impact is taken as the key time step and given a higher weight. After all time steps are weighted, the implicit state of the key time step is generated and an adaptive discrete wavelet transform is performed to generate enhanced samples: ， In the formula, It is the decomposition result of time series signal at multiple time scales. are the wavelet coefficients, is the mother wavelet function, is the adaptive scaling function of the time step, is a weighting function, which represents the importance weight of time step k at time t. is the time derivative of the customer behavior index, D is the customer behavior index, is the standard deviation of the customer behavior index D, which is used to measure the volatility of customer behavior changes. is the smoothing parameter, a and b are the upper and lower limits of the integral, indicating the time interval of the calculation, and dt represents a small time increment; Generate enhanced samples with different characteristics based on high-frequency and low-frequency reconstructed signals and add noise to different scale components to generate new samples. Store the generated enhanced sample data as a new training data set and re-input it into the inventory demand forecasting model for model training. Continue to use the mean square error as the loss function, iterate and optimize the model parameters based on the Adam optimizer until the loss is minimized and then stop the iteration. Input the iterated model parameters into the inventory demand forecasting model to obtain the inventory demand forecasting model. Input the inventory consumption and customer behavior index obtained by edge computing into the inventory demand forecasting model to obtain the inventory demand in the future period; Generate an inventory demand forecast report based on the forecast results, including the future inventory demand quantity and time of each commodity, and automatically generate replenishment orders based on the generated forecast report and place replenishment orders; The data analysis results are encrypted and transmitted to the central server, and a visual report is generated for storage.

2. The method for collaborative management of the Internet of Things and POS machines based on a 5G chip as claimed in claim 1, characterized in that: Establishing the connection between the IoT device and the POS machine through the 5G network refers to configuring network settings on the POS machine and the IoT device, selecting and connecting to the 5G network, using a network diagnostic tool to detect the connection status of the device, installing a blockchain node on the POS machine after the connection is established, adding the blockchain node to the newly created blockchain network, and synchronizing the latest data through the blockchain network.

3. The method for collaborative management of the Internet of Things and POS machines based on a 5G chip as claimed in claim 2, characterized in that: The collecting of real-time data from IoT devices and POS machines and preprocessing thereof refers to collecting inventory data through sensors, monitoring customer behavior through smart cameras and RFID tags, recording customer stay time and number of interactions, recording the transaction amount, commodity type and sales time after the transaction is completed from the POS machine in real time, marking the transaction ID with a unique identifier, storing the transaction data in a local database, cleaning the collected data, and removing outliers using the 3-times standard deviation method; De-duplicate the data based on its timestamp and unique identifier, and standardize the cleaned data; Align inventory data, customer behavior data, and transaction data based on timestamps, aggregate the aligned data, and generate comprehensive data records.

4. The method for collaborative management of the Internet of Things and POS machines based on a 5G chip as claimed in claim 3, characterized in that: The encrypted transmission of the data analysis results to the central server refers to adding noise to all the data processed by edge computing to generate obfuscated data: , where R is the obfuscated data, Q is the original data, and w is the noise, which refers to the randomly generated perturbation value. is the confusion coefficient; Divide the obfuscated data into blocks and mark each block with a serial number. Each block of data is 128 bits long. Use the AES-256 algorithm to encrypt each block of data to generate ciphertext. Use the TLS protocol to establish a secure connection, and transmit the encrypted obfuscated data to the central server through the 5G network. After receiving the encrypted data, the central server uses the same key to decrypt the received data, sorts and synthesizes the decrypted block data according to the serial number order marked in the database, restores the complete obfuscated data, removes the noise in the synthesized obfuscated data, and restores it to the original data.

5. The method for collaborative management of the Internet of Things and POS machines based on a 5G chip as claimed in claim 4, characterized in that: Generating visual reports for storage refers to generating inventory level graphs based on inventory data and predicted inventory demand in the received data, generating customer behavior analysis graphs based on customer behavior indexes and interaction data, integrating the produced graphs into reports, storing the generated reports in a central server and implementing secure access control on the stored data, backing up the stored data in the cloud and regularly checking the integrity of the stored data and the cloud backup data, and using blockchain technology to record all data changes.

6. A 5G chip-based Internet of Things and POS machine collaborative management system based on the 5G chip-based Internet of Things and POS machine collaborative management method according to any one of claims 1 to 5, characterized in that: include, A data collection module is used to collect data from IoT devices and POS machines and pre-process the data after the connection between the IoT devices and POS machines is established through the 5G network; Edge computing module, used to transmit IoT device data to POS machines through 5G networks for edge computing, calculating inventory consumption and customer behavior index; Demand forecasting module, used to forecast inventory demand in real time based on edge computing results and replenish stocks based on the forecast results; Data encryption module, used to encrypt and obfuscate data analysis results before transmitting them to the central server; The data storage module is used to generate visual reports on the decrypted data and store them, and then perform access control and cloud backup on the data.

7. A computer device comprising: Memory and processor; The memory stores a computer program, which is characterized in that when the processor executes the computer program, the steps of the 5G chip-based Internet of Things and POS machine collaborative management method described in any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps of the method for collaborative management of the Internet of Things and POS machines based on a 5G chip as described in any one of claims 1 to 5 are implemented.

Images