Intelligent store Internet of Things method and system

Abstract

The invention discloses an intelligent store Internet of Things method, and the method comprises the following steps: receiving a connection request of store hardware equipment, and synchronously carrying out the collection of anti-theft data and fitting data; performing anti-theft processing on the anti-theft data, and performing anti-theft efficiency self-diagnosis analysis through a self-adaptive algorithm; and analyzing the fitting data to obtain customer behavior analysis data. According to the intelligent store Internet of Things method, three-in-one store management improvement of equipment efficient access, anti-theft accurate diagnosis and sales strategy optimization is realized.

Description

Technical Field

[0001] The present invention belongs to the intersection of Internet of Things technology and retail intelligent management, and in particular relates to an intelligent store Internet of Things method and system. Background Art

[0002] In the digital transformation of the retail industry, remote device management and accurate processing of anti-theft data in stores have become key challenges. Existing RFID anti-theft systems suffer from a high false alarm rate (the industry average is 2.1%), leading to a large number of invalid alarms that disrupt normal operations. Traditional methods that rely on manual on-site troubleshooting suffer from slow response times (average time of 48 hours) and low maintenance efficiency, severely restricting store management efficiency and safety. Summary of the Invention

[0003] In view of the above-mentioned technical defects of the current retail Internet of Things system, the present invention provides an intelligent store Internet of Things method system, which forms a closed-loop management of equipment maintenance, anti-theft processing, performance optimization and sales analysis through the coordinated operation of the central processing unit, significantly improving store operation efficiency and anti-theft performance.

[0004] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0005] An intelligent store Internet of Things method includes the following steps:

[0006] By receiving connection requests from store hardware devices, anti-theft data and fitting data are collected simultaneously;

[0007] Performing anti-theft processing on the anti-theft data and performing anti-theft effectiveness self-diagnosis analysis through an adaptive algorithm;

[0008] Analyze the fitting data and output a behavior analysis report.

[0009] According to one aspect of the present invention, the synchronous collection of anti-theft data and fitting data by receiving a connection request from a store hardware device specifically includes:

[0010] Use multiple communication protocols to support connection requests from store hardware devices;

[0011] All hardware devices in stores across the country are remotely connected via the Internet to achieve automatic device topology discovery, dual-channel redundant communication and dynamic load balancing.

[0012] According to one aspect of the present invention, the anti-theft processing of the anti-theft data includes one or more of the following steps:

[0013] Data cleaning and whitelist comparison;

[0014] Three-level cache mechanism;

[0015] POS machine request optimization;

[0016] Third-party request buffering;

[0017] EPC data transmission.

[0018] According to one aspect of the present invention, the self-diagnosis analysis of the anti-theft effectiveness using the adaptive algorithm includes one or more of the following steps:

[0019] Calculate the detection rate and false alarm rate in real time. The detection rate is calculated based on the missed reading statistics method of the time window. The formula is η = (number of effective alarms) / (number of expected triggers) × 100%, where the number of expected triggers = the number of EPCs on the sales whitelist × the anti-theft device coverage rate;

[0020] Based on alarm event pattern recognition technology and adaptive threshold adjustment mechanism, the false alarm rate is calculated as μ = (number of false alarms) / (total number of alarms) × 100%, and a time decay factor μ' = μ × e^(-λΔt) is introduced;

[0021] The anti-theft alarm is triggered when the detection rate is lower than 95% or the false alarm rate is higher than 5%;

[0022] Dynamically adjust the anti-theft alarm threshold, the formula is T_n=αT_(n-1)+(1-α)μ', where α∈[0.6,0.8].

[0023] According to one aspect of the present invention, the anti-theft alarm comprises:

[0024] When the detection rate or false alarm rate of an anti-theft device is lower than the set value, an alarm will be automatically pushed and a maintenance notification will be generated;

[0025] Query and count the detection rate and false alarm rate of each anti-theft device.

[0026] According to one aspect of the present invention, analyzing the fitting data and outputting a behavior analysis report includes:

[0027] Data linkage between RFID reading timestamp and door magnetic sensor triggering time;

[0028] Statistical modeling of fitting time and conversion rate;

[0029] Data push and query.

[0030] According to one aspect of the present invention, the data linkage between the RFID reading timestamp and the door magnetic sensor triggering time includes:

[0031] Matching analysis between RFID reading timestamp and door magnetic sensor triggering time;

[0032] Calculate the length of time customers spend trying on clothes by using the time difference.

[0033] According to one aspect of the present invention, the statistical modeling of fitting time and conversion rate includes:

[0034] Calculate the average number of fittings and fitting duration for different product categories;

[0035] Establish a correlation analysis model between fitting behavior and sales conversion rate.

[0036] According to one aspect of the present invention, the data push and query includes pushing statistical results to the platform through the device for management personnel to browse and query, and optimize product layout and sales strategy. An intelligent store Internet of Things system includes:

[0037] The data collection module receives connection requests from store hardware devices and performs data collection and remote maintenance management simultaneously;

[0038] A data processing module performs anti-theft processing on the data and performs anti-theft effectiveness self-diagnosis analysis through an adaptive algorithm and outputs analysis results;

[0039] The fitting analysis module performs correlation analysis on the fitting behavior data based on the analysis results to obtain customer behavior analysis data.

[0040] Advantages of the present invention:

[0041] Efficient remote maintenance and management

[0042] Through the device access layer, remote connection to hardware equipment in stores across the country is achieved, and data collection and real-time maintenance are completed simultaneously, which greatly reduces on-site maintenance costs and response time (fault response is shortened from 48 hours to 2 hours, an 80% increase in efficiency).

[0043] It supports real-time monitoring of equipment status, fault code self-diagnosis, dual-channel redundant communication and hardware abnormality warning to ensure stable system operation.

[0044] Intelligent anti-theft performance optimization

[0045] Adaptive algorithms are used to dynamically calculate the detection rate (η) and false alarm rate (μ), and the performance of anti-theft devices is accurately quantified based on a 7-day / 30-day sliding window missed reading statistical model and alarm event pattern recognition technology.

[0046] Dynamically adjust the warning threshold (T_n=αT_(n-1)+(1-α)μ'), introduce a time decay factor to suppress historical false alarm interference, and reduce the false alarm rate to below 0.3% (industry average 2.1%).

[0047] When the detection rate is <95% or the false alarm rate is >5%, an L2 level warning is automatically triggered and a maintenance work order is generated to ensure the efficient operation of the anti-theft system.

[0048] Accurate fitting behavior analysis

[0049] By linking RFID with door magnetic sensor data, the fitting time (ΔT = avg(t2'-t1')) is accurately calculated, and a fitting-to-purchase conversion rate matrix is established to reveal the correlation between product fitting behavior and sales conversion.

[0050] Generate heat maps to visualize hot spots for fitting activities in stores, provide intelligent suggestions such as product display optimization and fitting room layout adjustments, and increase sales conversion rates by 40%.

[0051] Automated analysis of the entire process reduces manual intervention, improving store operation efficiency and scientific decision-making.

[0052] Layered architecture ensures system scalability

[0053] The three-tier architecture of device access layer - data processing layer - business application layer supports multi-protocol compatibility (MQTT / HTTP / CoAP) and multi-module expansion to meet the needs of stores of different sizes.

[0054] Local + distributed caching mechanism (memory LRU, SQL, Redis) and request priority queue management reduce server load pressure (actually measured concurrency pressure is reduced by 65%). BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0056] Figure 1 This is a system architecture topology diagram of an intelligent store Internet of Things method and system according to the present invention;

[0057] Figure 2 This is a flow chart of anti-theft data processing for an intelligent store Internet of Things method and system according to the present invention;

[0058] Figure 3 This is a flowchart of anti-theft data and server processing of an intelligent store Internet of Things method and system according to the present invention;

[0059] Figure 4 This is a schematic diagram of an anti-theft effectiveness analysis interface of an intelligent store Internet of Things method and system according to the present invention;

[0060] Figure 5 This is a fitting analysis statistical chart for the intelligent store Internet of Things method and system described in the present invention.

[0061] Figure 6 This is a schematic diagram of the fitting room sensor networking of the intelligent store Internet of Things method and system described in the present invention.

[0062] Figure 7 This is a schematic diagram of the working of fitting room sensors in an intelligent store Internet of Things method and system according to the present invention;

[0063] Figure 8 This is a basic flow chart of the intelligent store Internet of Things method and system described in the present invention. DETAILED DESCRIPTION

[0064] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0065] Example 1

[0066] like Figures 1 to 7 As shown in the figure, the system architecture of the intelligent store IoT method and system adopts a layered architecture design:

[0067] Device access layer:

[0068] Supports MQTT / HTTP / CoAP protocols to receive connection requests from store hardware devices (such as RFID readers and infrared sensors).

[0069] It complies with the standard protocol of the Internet of Things industry and is suitable for the compatibility requirements of different hardware devices (for example, RFID readers often use MQTT, while sensors may use CoAP). In this embodiment, the RFID readers deployed in clothing stores are connected through the MQTT protocol, and the access control sensors use the CoAP protocol to upload timestamps.

[0070] In this embodiment, a clothing store deploys 10 RFID readers and 5 access control sensors, and sends connection requests to the devices through the MQTT protocol. The device access layer maintains the link status through heartbeat packets (interval ≤ 5 seconds), and the system monitors the device status through the heartbeat packet status code (0 / 1 / 2).

[0071] Data processing layer:

[0072] Contains data cleaning, whitelist comparison and request queue management modules.

[0073] RFID data cleaning:

[0074] RFID data cleaning rules

[0075] Remove duplicate EPC codes: This solves the redundancy caused by reading the same product multiple times.

[0076] Timestamp calibration (±50ms): Corrects sensor clock deviation.

[0077] Whitelist comparison:

[0078] EPC whitelist source: a list of legal product labels in the store’s product database.

[0079] Blacklist rules: Combine brand characteristics (such as the serial number patterns of luxury goods labels) and a dynamically updated blacklist library.

[0080] In this embodiment, the EPC code E200345678 of women's clothing in the store is searched. When the EPC code E200345678 does not hit the memory cache, the SQLite database is automatically queried to verify the legitimacy.

[0081] Cache mechanism: Use three-level cache (memory LRU → SQLite → Redis) to reduce database access frequency;

[0082] Memory LRU cache: capacity 10,000 entries, refresh cycle 5 seconds.

[0083] SQLite response time: ≤500ms.

[0084] Redis response time: ≤100ms.

[0085] Request queue management: Prioritize POS and third-party API requests (QoS grading) to ensure core business processes.

[0086] Business application layer:

[0087] Provides equipment management, performance analysis and sales-related modules.

[0088] Device management: supports remote login, parameter configuration and firmware upgrade;

[0089] Sales association analysis: Optimize product display strategies based on fitting behavior data (such as RFID reading and door magnetic trigger time) to increase fitting conversion rates (for example, the women's clothing category increased from 32% to 38%).

[0090] like Figure 1 As shown, a smart store Internet of Things method includes the following steps:

[0091] Step S1: By receiving the connection request of the store hardware equipment, data collection and remote maintenance management are carried out synchronously;

[0092] Step S11: device access and protocol adaptation;

[0093] Multi-protocol support: The device access layer is compatible with MQTT / HTTP / CoAP protocols through a dynamic protocol parsing engine.

[0094] Technical implementation: Deploy a protocol identification module to automatically identify the protocol type based on data packet features (such as MQTT fixed header and CoAPOption field); use a protocol conversion gateway to convert heterogeneous protocols into a unified internal format (such as JSON over MQTT).

[0095] In this embodiment, the RFID reader sends the product tag EPC code (topic: / store001 / RFID / data stream) through the MQTT protocol and the access control sensor uploads the trigger timestamp through the CoAP protocol. The system dynamically converts it into the MQTT format and enters the data processing layer to realize device access and protocol adaptation.

[0096] Heartbeat keepalive and status monitoring:

[0097] The device sends a heartbeat packet containing a status code (0 = normal / 1 = offline / 2 = abnormal) at intervals of ≤ 5 seconds;

[0098] Step S12: data collection and preprocessing;

[0099] Real-time data diversion: High-priority data (such as anti-theft alarms and POS transaction records) is directly transmitted to the data processing layer through the main TCP / IP channel; non-real-time data (such as logs and configuration files) is processed asynchronously through the Kafka message queue.

[0100] Data cleaning and standardization: The cleaning rules are to remove duplicate EPC codes and correct timestamp deviation (calibration within ±50ms).

[0101] Step S13: remote maintenance collaboration;

[0102] Fault self-check logic: Health check items: CPU usage > 80%, network delay > 500ms, firmware version expired;

[0103] Diagnostic rules: Single abnormality → P2 level alarm; Two or more abnormalities → P1 level alarm;

[0104] Remote command channel: Maintenance commands are encrypted via TLS 1.3, and support is provided for SSH tunnel penetration to perform firmware upgrades (e.g., to fix CVE vulnerabilities).

[0105] Security control: two-step verification (OTP+biometrics), operation logs synchronized to blockchain for evidence storage.

[0106] Step S2: performing anti-theft processing on the data, and performing anti-theft effectiveness self-diagnosis analysis using an adaptive algorithm to output analysis results;

[0107] Step S21: data theft prevention processing mechanism;

[0108] EPC whitelist comparison:

[0109] Three-level cache architecture:

[0110] In-memory LRU cache (10,000 entries / 5-second refresh) → SQLite database (500ms response) → Redis cluster (100ms response).

[0111] In this embodiment, when the RFID reader uploads the EPC code E200345678, the system prioritizes a hit in the memory cache, and queries the database if a hit is not found.

[0112] Abnormal data interception:

[0113] Blacklist rule: The EPC code is not on the whitelist and matches common stolen tag features (such as luxury brands);

[0114] Step S22: Quantitative analysis of adaptive anti-theft effectiveness;

[0115] Detection rate (η) calculation:

[0116] Detection rate η = (number of effective alarms) / (number of times that should be triggered) × 100%;

[0117] Number of triggers = number of EPCs on the sales whitelist × anti-theft device coverage rate.

[0118] In this embodiment, the specific implementation is:

[0119] Number of times the trigger should occur = number of whitelisted products × covered store area × time window (e.g. 95% within 1 hour);

[0120] When η=92%, the system is marked as "to be optimized" state (below the 95% threshold).

[0121] Dynamic adjustment of false alarm rate (μ):

[0122] Introducing a time decay factor: μ' = μ × e^(-λΔt) (λ = 0.05 / day) to suppress occasional false alarms;

[0123] Step S23: early warning and disposal process;

[0124] Multi-level early warning mechanism:

[0125] Level 0 (Normal): η ≥ 95% and μ ≤ 5%;

[0126] Level 1 (Attention): η < 95% for 3 consecutive days or μ > 5%;

[0127] Level 2 (Emergency): η < 90% or 5 consecutive false alarms;

[0128] Set a sliding time window (7 days / 30 days). When η < 95% or μ > 5% in 3 consecutive window periods, trigger a Level 2 warning;

[0129] Obtain the dynamic adjustment formula based on the statistical dynamic threshold adjustment model: T_n = αT_(n - 1)+(1 - α)μ'(α ∈ [0.6, 0.8]).

[0130] Automatically generate an analysis report:

[0131] Output a PDF report to visualize the data analysis results. The report includes:

[0132] Anti-theft device health score (0 - 100); Heat map of high-risk areas (clustered based on historical false alarm points); Maintenance suggestions (such as adjusting the angle of the RFID reader to reduce the missed reading rate).

[0133] Step S3: Based on the analysis results, conduct an association analysis on the fitting behavior data and output a behavior analysis report.

[0134] Step S31: Multi-sensor data fusion;

[0135] Spatio-temporal alignment algorithm:

[0136] Align the RFID reading timestamp t1 with the access control sensor trigger timestamp t2, and calculate the fitting duration ΔT = avg(t2' - t1');

[0137] Specifically, customer A scans the code to start fitting at 10:00:00 and triggers the door magnetic at 10:15:30. The system identifies t1' = 10:00:05 (first scan), t2 = 10:15:30, and ΔT = 925 seconds.

[0138] Repeat fitting filtering:

[0139] Define a time threshold T (such as 30 minutes): If t1'' - t1' < T for the same customer, mark it as a repeated fitting and exclude it from the statistics.

[0140] Step S32: Association analysis model;

[0141] Purchase conversion rate matrix:

[0142]

[0143] Establish a fitting-sales association model: correlation degree K = number of purchases after fitting / (total number of fittings - number of repeated fittings).

[0144] Step S33: data push and query;

[0145] The statistical results are pushed to the platform through the device for managers to browse and query, and optimize product layout and sales strategies.

[0146] Advantages of the present invention: Through the above solution, the present invention has achieved significant technological breakthroughs in equipment access and management, data security and anti-theft, sales behavior analysis, etc. The specific advantages are as follows:

[0147] 1. Efficiency of device access and protocol adaptation

[0148] Multi-protocol compatibility: Supports MQTT / HTTP / CoAP protocols to adapt to the communication needs of devices from different manufacturers and avoid access difficulties caused by protocol incompatibility.

[0149] Example verification: A clothing store deployed 10 RFID readers and 5 access control sensors, achieved stable connection through the MQTT protocol, and achieved a 100% device online rate.

[0150] Heartbeat keepalive mechanism: The device sends heartbeat packets at intervals of ≤5 seconds to ensure link stability and monitor device status (normal / offline / abnormal) in real time.

[0151] Effect comparison: Compared with the traditional fixed heartbeat interval (such as 30 seconds), the fault detection time is shortened to within 5 seconds, reducing the risk of business interruption due to device offline.

[0152] Protocol conversion gateway: Automatically identify heterogeneous protocols (such as CoAP→MQTT), unify internal data formats, and reduce system integration complexity.

[0153] 2. Efficiency and security of data processing

[0154] Three-level cache mechanism: Use memory LRU → SQLite → Redis three-level cache to reduce database access frequency and improve data processing speed.

[0155] Performance comparison: RFID data query response time is optimized from 500ms in traditional solutions to ≤100ms, and the cache hit rate reaches over 95%.

[0156] QoS priority scheduling: hierarchical processing of POS machine and third-party API requests to ensure the real-time performance of core business processes (such as payment transactions).

[0157] Data cleaning and standardization: Automatically remove duplicate EPC codes and correct timestamp deviations to ensure data quality.

[0158] 3. Intelligent and adaptive optimization of anti-theft performance

[0159] EPC whitelist comparison: The three-level cache architecture (memory LRU → SQLite → Redis) enables fast comparison with a response time of <50ms.

[0160] Quantitative analysis of adaptive anti-theft effectiveness:

[0161] Detection rate (η) calculation: A time-window-based missed read statistics method is used to dynamically evaluate the performance of anti-theft devices.

[0162] Verification of the embodiment: When η=92%, the system is marked as "to be optimized" and it is recommended to adjust the position of the RFID reader.

[0163] Dynamic adjustment of false alarm rate (μ): Introduce the time decay factor μ' = μ × e^(-λΔt) to suppress the impact of occasional false alarms.

[0164] Effect comparison: A sensor falsely triggered an alarm due to a thunderstorm in the store. After three days of attenuation, μ dropped from 2.5% to 0.8%, reducing false alarm interference.

[0165] Multi-level warning mechanism: L0 (normal) → L1 (concern) → L2 (emergency) graded warning, combined with sliding window statistics to improve warning accuracy.

[0166] 4. Accuracy of sales behavior analysis and decision support

[0167] Multi-sensor data fusion: The RFID reading timestamp t1 is linked with the access control sensor trigger timestamp t2 to calculate the fitting time ΔT and filter repeated fitting behaviors.

[0168] Purchase conversion rate matrix: Optimize product display strategies based on fitting behavior data to increase sales conversion rates.

[0169] Data push and query: Statistical results are pushed to the management platform in real time, supporting online browsing and historical data query to assist operational decision-making.

[0170] Managers' decision-making cycles were shortened from weeks to hours, and inventory turnover increased by 15%.

[0171] 5. Remote maintenance and improved operation and maintenance efficiency

[0172] Fault self-diagnosis and diagnosis: Automatically detects health status such as CPU usage, network latency, and firmware version, and triggers graded alarms when abnormalities occur.

[0173] Remote command channel: supports SSH tunnel penetration to perform firmware upgrades, ensuring safe and stable system operation.

[0174] Optimized operation and maintenance costs: Remote maintenance coverage increased to 95%, and on-site maintenance needs were reduced by 80%.

[0175] Example 2

[0176] An intelligent store Internet of Things system, based on the intelligent store Internet of Things method described in Example 1, includes:

[0177] The data collection module collects anti-theft data and fitting data simultaneously by receiving connection requests from store hardware devices;

[0178] A data processing module performs anti-theft processing on the anti-theft data and performs self-diagnosis analysis of the anti-theft effectiveness through an adaptive algorithm;

[0179] The fitting analysis module performs correlation analysis on the fitting behavior data based on the analysis results to obtain customer behavior analysis data.

[0180] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general-purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing this type of system. In addition, the present invention is not directed to any specific programming language. It should be understood that various programming languages can be utilized to realize the content of the present invention described herein, and the above description of specific languages is for the purpose of disclosing the best mode of the present invention.

[0181] In the description provided herein, numerous specific details are described. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques are not shown in detail so as not to obscure the understanding of this description.

[0182] Similarly, it should be understood that in order to streamline the present disclosure and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the claims below, inventive aspects lie in less than all the features of the individual embodiments disclosed above. Accordingly, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of the invention.

[0183] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition may be divided into multiple submodules or subunits or subcomponents. All features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed herein may be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0184] Furthermore, those skilled in the art will appreciate that although some embodiments herein include certain features included in other embodiments but not other features, combinations of features from different embodiments are intended to be within the scope of the present invention and to form different embodiments. For example, in the claims below, any of the claimed embodiments may be used in any combination.

[0185] It should be noted that the above embodiments illustrate rather than limit the invention, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between brackets should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising several different elements and by means of appropriately programmed computers. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

[0186] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. An intelligent store Internet of Things method, characterized in that: The method comprises the following steps: By receiving connection requests from store hardware devices, anti-theft data and fitting data are collected simultaneously; Performing anti-theft processing on the anti-theft data and performing anti-theft effectiveness self-diagnosis analysis through an adaptive algorithm; The fitting data is analyzed to obtain customer behavior analysis data.

2. The intelligent store Internet of Things method according to claim 1, characterized in that: The method of synchronously collecting anti-theft data and fitting data by receiving a connection request from a store hardware device specifically includes: Use multiple communication protocols to support connection requests from store hardware devices; All hardware devices in stores across the country are remotely connected via the Internet to achieve automatic device topology discovery, dual-channel redundant communication and dynamic load balancing.

3. The intelligent store Internet of Things method according to claim 1, characterized in that: The anti-theft processing of the anti-theft data includes one or more of the following steps: Data cleaning and whitelist comparison; Three-level cache mechanism; POS machine request optimization; Third-party request buffering; EPC data transmission.

4. The intelligent store Internet of Things method according to claim 1, characterized in that: The self-diagnosis analysis of the anti-theft effectiveness using the adaptive algorithm includes one or more of the following steps: Calculate the detection rate and false alarm rate in real time. The detection rate is calculated based on the missed reading statistics method of the time window. The formula is η = (number of effective alarms) / (number of expected triggers) × 100%, where the number of expected triggers = the number of EPCs on the sales whitelist × the anti-theft device coverage rate; Based on the alarm event pattern recognition technology and the adaptive threshold adjustment mechanism, the false alarm rate is calculated as μ = (number of false alarms) / (total number of alarms) × 100%, and the time decay factor is introduced. μ'=μ×e^(-λΔt); When the detection rate is lower than 95% or the false alarm rate is higher than 5%, the anti-theft alarm is triggered; Dynamically adjust the anti-theft alarm threshold, the formula is T_n=αT_(n-1)+(1-α)μ', where α∈[0.6,0.8].

5. The intelligent store Internet of Things method according to claim 4 is characterized in that: The anti-theft alarm includes: When the detection rate or false alarm rate of an anti-theft device is lower than the set value, an alarm will be automatically pushed and a maintenance notification will be generated; Query and count the detection rate and false alarm rate of each anti-theft device.

6. The intelligent store Internet of Things method according to claim 1, characterized in that: Analyzing the fitting data to obtain customer behavior analysis data includes: Data linkage between RFID reading timestamp and door magnetic sensor trigger time; Statistical modeling of fitting time and conversion rate; Data push and query.

7. The intelligent store Internet of Things method according to claim 6, characterized in that: The data linkage between the RFID reading timestamp and the door magnetic sensor trigger time includes: Matching analysis between RFID reading timestamp and door magnetic sensor triggering time; Calculate the length of time customers spend trying on clothes by using the time difference.

8. The intelligent store Internet of Things method according to claim 6, characterized in that: The statistical modeling of fitting time and conversion rate includes: Calculate the average number of fittings and fitting duration for different product categories; Establish a correlation analysis model between fitting behavior and sales conversion rate.

9. The intelligent store Internet of Things method according to claim 6, characterized in that: The data push and query include pushing statistical results to the platform through the device for management personnel to browse and query, and optimize product layout and sales strategies.

10. An intelligent store Internet of Things system, characterized by: The intelligent store Internet of Things method according to any one of claims 1 to 9 is implemented, comprising: The data collection module collects anti-theft data and fitting data simultaneously by receiving connection requests from store hardware devices; A data processing module performs anti-theft processing on the anti-theft data and performs self-diagnosis analysis of the anti-theft effectiveness through an adaptive algorithm; The fitting analysis module performs correlation analysis on the fitting behavior data based on the analysis results to obtain customer behavior analysis data.

Images