Intelligent control method and system for limb pressurization equipment

Abstract

The invention discloses an intelligent control method and system for limb pressurization equipment, and relates to the technical field of intelligent control. The current working state of the limb pressurization equipment is determined according to multiple pieces of working data of the limb pressurization equipment; the training report of the target limb is determined according to the current working state of the limb pressurization equipment, the multiple training pressure items and the overall form of the target limb, and the accuracy of the training report of the target limb is improved. Determining a plurality of training features according to the detection of the training report, and determining a pressure curve of the target limb according to the plurality of training features, the pressure condition of the target limb and the age of the user; determining an intelligent control event of the limb pressurization equipment based on the pressure curve of the target limb and the current working state of the limb pressurization equipment, and determining a dynamic early warning system of the limb pressurization equipment according to the intelligent control event, the abnormal signal of the limb pressurization equipment and the current body state of the user, and the accuracy of the dynamic early warning system of the limb pressurization equipment is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control, and in particular to an intelligent control method and system for a limb pressurizing device. BACKGROUND

[0002] With the development of science and technology, a limb pressurizing device is an intelligent medical or fitness device specially used for applying controllable pressure to a human limb (such as an arm or a thigh). The limb pressurizing device is worn on a target limb of a user and performs pressurizing training on the target limb. The target limb can be an arm. In the prior art, the limb pressurizing device is worn on the arm of the user and performs pressurizing training on the arm. The limb pressurizing device performs dynamic pressurizing on the target limb, collects multiple pulse data, and determines a regular pressure training of the limb pressurizing device based on the multiple pulse data. However, the current working state of the limb pressurizing device and multiple training pressure projects are ignored, which affects the intelligent control of the limb pressurizing device and results in low accuracy of a dynamic early warning system of the limb pressurizing device. SUMMARY

[0003] The present application aims to overcome the shortcomings of the prior art. The present application provides an intelligent control method and system for a limb pressurizing device.

[0004] The present application provides an intelligent control method for a limb pressurizing device, which comprises the following steps. When the limb pressurizing device is worn on a target limb of a user, a pressurizing mode of the limb pressurizing device on the target limb is determined based on the limb pressurizing device, the target limb, and corresponding detection projects, and dynamic pressurizing of the limb pressurizing device on the target limb is triggered to determine multiple pulse data of the target limb. A training pressure event of the limb pressurizing device is determined according to the multiple pulse data and a current physical state of the user, and multiple training pressure projects are determined based on recognition of the training pressure event. Multiple working data of the limb pressurizing device are collected, a current working state of the limb pressurizing device is determined according to the multiple working data of the limb pressurizing device, and a training report of the target limb is determined according to the current working state of the limb pressurizing device, the multiple training pressure projects, and an overall shape of the target limb. In the training report of the target limb, multiple training features are determined according to detection of the training report, a pressure curve of the target limb is determined according to the multiple training features, a compression condition of the target limb, and an age of the user. An intelligent control event of the limb pressurizing device is determined based on the pressure curve of the target limb and the current working state of the limb pressurizing device, and a dynamic early warning system of the limb pressurizing device is determined according to the intelligent control event, an abnormal signal of the limb pressurizing device, and a current physical state of the user.

[0005] The embodiment of the present application provides a kind of intelligent control system of limb pressurizing equipment, the intelligent control system of limb pressurizing equipment is applied to the intelligent control method of above-mentioned limb pressurizing equipment.

[0006] Compared with the prior art, the beneficial effects of the present application are: (1) when the limb pressurizing equipment is worn on the target limb of the user, the pressure mode of the limb pressurizing equipment on the target limb is determined based on the limb pressurizing equipment, the target limb and the corresponding detection item, and the dynamic pressure of the limb pressurizing equipment on the target limb is triggered to determine multiple pulse data of the target limb; the training pressure event of the limb pressurizing equipment is determined according to the multiple pulse data and the current physical state of the user, and multiple training pressure items are determined based on the identification of the training pressure event, which introduces multiple pulse data of the target limb, controls the training pressure event of the limb pressurizing equipment, and improves the identification accuracy of the multiple training pressure items.

[0007] (2) multiple working data of the limb pressurizing equipment are collected, the current working state of the limb pressurizing equipment is determined according to the multiple working data of the limb pressurizing equipment, and the training report of the target limb is determined according to the current working state of the limb pressurizing equipment, multiple training pressure items and the overall shape of the target limb, which introduces the current working state of the limb pressurizing equipment, and comprehensively considers the current working state of the limb pressurizing equipment, multiple training pressure items and the overall shape of the target limb, to improve the accuracy of the training report of the target limb.

[0008] (3) in the training report of the target limb, multiple training features are determined according to the detection of the training report, and the pressure curve of the target limb is determined according to the multiple training features, the pressure condition of the target limb and the age of the user; the intelligent control event of the limb pressurizing equipment is determined based on the pressure curve of the target limb and the current working state of the limb pressurizing equipment, and the dynamic early warning system of the limb pressurizing equipment is determined according to the intelligent control event, the abnormal signal of the limb pressurizing equipment and the current physical state of the user, which introduces the pressure curve of the target limb, further controls the intelligent control event of the limb pressurizing equipment, and improves the accuracy of the dynamic early warning system of the limb pressurizing equipment. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 is a flowchart of the intelligent control method of the limb pressurizing equipment in the embodiment of the present application; Figure 2 is a flowchart of step S11 in the intelligent control method of the limb pressurizing equipment in the embodiment of the present application; Figure 3 is a flowchart of step S12 in the intelligent control method of the limb pressurizing equipment in the embodiment of the present application; Figure 4is a flowchart of step S13 in the intelligent control method of the limb compression device in the embodiment of the application; Figure 5 is a flowchart of step S14 in the intelligent control method of the limb compression device in the embodiment of the application; Figure 6 is a flowchart of step S15 in the intelligent control method of the limb compression device in the embodiment of the application; Figure 7 is a structural composition diagram of the intelligent control system of the limb compression device in the embodiment of the application. DETAILED DESCRIPTION

[0010] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application.

[0011] Please refer to Figures 1 to 7 An intelligent control method of a limb compression device is applied to an intelligent control scene. The intelligent control method of the limb compression device comprises the following steps. Step S11: When the limb compression device is worn on a target limb of a user, a compression mode of the limb compression device on the target limb is determined based on the limb compression device, the target limb, and corresponding detection items, and dynamic compression of the target limb by the limb compression device is triggered to determine a plurality of pulse data of the target limb. Step S12: A training pressure event of the limb compression device is determined according to the plurality of pulse data and a current body state of the user, and a plurality of training pressure items are determined based on recognition of the training pressure event. Step S13: A plurality of working data of the limb compression device are collected, a current working state of the limb compression device is determined according to the plurality of working data of the limb compression device, and a training report of the target limb is determined according to the current working state of the limb compression device, the plurality of training pressure items, and an overall morphology of the target limb. Step S14: In the training report of the target limb, a plurality of training features are determined according to detection of the training report, and a pressure curve of the target limb is determined according to the plurality of training features, a compression condition of the target limb, and an age of the user. Step S15: An intelligent control event of the limb compression device is determined based on the pressure curve of the target limb and the current working state of the limb compression device, and a dynamic early warning system of the limb compression device is determined according to the intelligent control event, an abnormal signal of the limb compression device, and the current body state of the user.

[0012] Reference Figure 2 In step S11, the specific steps are as follows: S111: The limb compression device is worn on the target limb of the user and is arranged around the outer periphery of the target limb of the user; detection items of the target limb of the user are collected, and a plurality of compression stage events are determined according to the limb compression device, the target limb and the corresponding detection items; S112: A compression mode of the limb compression device on the target limb is determined according to the plurality of compression stage events; dynamic compression of the limb compression device on the target limb is triggered based on the compression mode, and a plurality of pulse data of the target limb are collected during the compression process.

[0013] In the embodiment of the present application, the limb compression device is worn on the target limb of the user and is arranged around the outer periphery of the target limb of the user; detection items of the target limb of the user are collected, and a plurality of compression stage events are determined according to the limb compression device, the target limb and the corresponding detection items, which is compatible with the overall consideration of the limb compression device, the target limb and the corresponding detection items, and ensures the accuracy of the plurality of compression stage events.

[0014] At this time, the limb compression device is worn on the target limb of the user and is arranged around the outer periphery of the target limb of the user; the limb compression device confirms that the cuff has been correctly wrapped and adhered to the outer periphery of the target limb through the built-in pressure sensor or contact sensor, to ensure the effectiveness and safety of subsequent compression; the cuff adopts a multi-layer composite material, including an outer wear-resistant layer, an intermediate regional air bag layer and an inner anti-skid biological material, and is quickly and stably bound through magic tape and metal buckles; the central control module will detect whether the pressure is zero during initialization, and preliminarily judges the cuff binding state through air tightness detection.

[0015] In the limb compression device, the most core detection item is the measurement of arterial occlusion pressure, and the collection of auxiliary physiological parameters such as resting heart rate and blood pressure baseline is also included; the user can select the AOP automatic measurement process on the mobile terminal APP, and the system will activate the arterial occlusion pressure detection unit, start the pressure pulsation wave sensor, PPG photoelectric sensor or external Doppler ultrasound probe, and collect data from different sensors at the same time, to provide multi-dimensional information input for subsequent judgment.

[0016] The central control module generates a discrete and serialized compression control plan according to the determined detection items, and the plan is composed of a plurality of compression stage events, each event defines an explicit target pressure value and a corresponding monitoring task; a compression stage event includes a target pressure value, a pressure holding time and a pulse sampling requirement and other key parameters; the control algorithm will preset the pressure step value and the holding time, generate the event sequence until the safety upper limit is reached; the advanced algorithm can also dynamically adjust the subsequent pressure step according to the pulse signal strength, and the whole sequence is strictly constrained by the safety strategy, to ensure that the highest pressure will not cause harm to the user.

[0017] Specifically, a fitness person is ready to use the limb compression device for upper limb BFR training, and the device needs to automatically measure the arterial occlusion pressure of the person's arm; the fitness person binds the intelligent compression sleeve (about 10 cm wide) of the limb compression device above the biceps of the left upper arm, tightens the magic tape and buckles the metal buckle; the built-in micro pressure sensor of the device detects that the initial pressure inside the sleeve is 0 mmHg and the air path system has no leakage, sends a wearing ready signal to the central control module, and the APP interface displays that the sleeve has been correctly worn and the measurement can begin.

[0018] The fitness person clicks start training on the mobile phone APP, and the system automatically enters the AOP automatic measurement process; the central control module activates the AOP detection unit based on pressure pulse wave built in the sleeve, which starts to listen to the small pulse wave oscillation signal superimposed on the static pressure due to the change of the sleeve pressure at a high frequency of 100 Hz.

[0019] The control algorithm unit of the central control module generates a series of compression phase events based on the core detection item of AOP measurement in real time: starting from {target pressure: 60 mmHg, holding time: 8 seconds}, gradually increasing to {target pressure: 160 mmHg, holding time: 8 seconds}, each event contains the task of collecting and analyzing pulse wave amplitude; during execution, when the pressure reaches 100 mmHg, the algorithm analyzes that the pulse wave amplitude starts to significantly attenuate, so the subsequent pressure step is adjusted from 20 mmHg to 10 mmHg to improve the resolution of capturing AOP points; the safety upper limit of the entire event sequence is set to 200 mmHg.

[0020] Further, the compression mode of the limb compression device to the target limb is determined according to the plurality of compression phase events; the dynamic compression of the limb compression device to the target limb is triggered based on the compression mode, and a plurality of pulse data of the target limb are collected during the compression process, which takes into account the overall consideration of the plurality of compression phase events, and ensures the accuracy of the compression mode of the limb compression device to the target limb.

[0021] At this time, the control algorithm unit of the central control module converts each event into specific time sequence control instructions, such as converting the target pressure value into the PWM drive signal of the micro air pump and the specific drive time length; the closed-loop PID adjustment strategy is embedded in the compression mode, which dynamically adjusts the air pump power or the opening degree of the pressure relief valve by continuously receiving the feedback of the high-precision pressure sensor, to realize smooth and accurate pressure control, and ensure that the final pressure is stable within the error range of ±3 mmHg of the set value; at the same time, safety logic is integrated, and the global pressure change rate monitor will immediately reduce the air pump power when the pressure rises too fast to prevent impact on the user.

[0022] The central control module sends instructions to the automatic inflation and pressure relief module through the power drive circuit according to the compiled pressurization mode, and drives the physical execution component to start dynamic and controllable pressurization of the cuff. The GPIO pin output of the MCU controls the H-bridge or MOSFET circuit, thereby accurately controlling the rotation speed and start-stop of the miniature air pump. During the pressurization process, the one-way valve ensures that the gas can only flow to the cuff air bag, and the electronic pressure relief valve remains closed to maintain the pressure. If the cuff adopts a regional air bag design, the pressurization mode can also control multiple air path channels to achieve differential pressurization of different regions to achieve more uniform pressure distribution.

[0023] In each stage of dynamic pressurization, the system synchronously starts the arterial occlusion pressure detection unit and the pressure acquisition component to perform high-frequency and multi-channel data acquisition of the physiological signals of the target limb and the physical state of the cuff. The system synchronously acquires two types of core data: digital pressure signals converted by the high-precision pressure sensor through the ADC acquisition module, and pressure pulsation wave or PPG photoelectric plethysmogram signals captured by the AOP detection unit. The original pulse signals acquired are filtered, amplified, and feature-extracted in real time to determine the blood flow state. The system timestamps each frame of data to ensure that the pulse data and the corresponding pressure values are strictly aligned in time, which is the key to accurately determining the AOP point.

[0024] Specifically, the MCU of the central control module compiles the event sequence generated by S111 into a complete pressurization mode script. The script contains a series of instructions: start the PID controller and set the first target pressure to 60 mmHg, while limiting the pressure rise slope to 5 mmHg / s; when the pressure reaches 60 mmHg and stabilizes, start the pulse wave acquisition module for 8 seconds and mark the data packet as Pressure_60; then update the PID target pressure to 80 mmHg and repeat the acquisition process, marking the data packet as Pressure_80, and so on. The script also integrates safety instructions that require monitoring of the pressure sensor feedback throughout the process. If the instantaneous pressure exceeds the set value of 10 mmHg, the micro pressure relief valve is immediately triggered for correction.

[0025] The MCU starts executing the pressurization mode script and sends signals to the power drive circuit to drive the miniature air pump inside the cuff to rotate. The gas is pumped into the intermediate air bag layer through the one-way valve and air path pipeline. The pressure sensor feeds back the pressure value to the PID controller in real time, and the controller dynamically adjusts the PWM duty cycle to make the pressure rise smoothly and linearly to the first target value of 60 mmHg, and accurately maintain stability thereafter.

[0026] When the pressure is stabilized at 60 mmHg, the AOP detection unit based on the pressure pulsation wave technology is activated to collect the pressure oscillation signals in the cuff at a frequency of 200 Hz; at the same time, the pressure collection component also records the current static pressure value at the same frequency; the system performs real-time Fourier transform on the collected oscillation signals, extracts the main frequency component synchronized with the heartbeat and calculates its amplitude, which is identified as the pulse wave amplitude at the current pressure; the system records the data set {timestamp: T1, pressure: 60 mmHg, pulse wave amplitude: A1}; then, the system pressurizes to 80 mmHg, repeats the collection process, and obtains {timestamp: T2, pressure: 80 mmHg, pulse wave amplitude: A2}; in this way, the cycle continues until at a certain pressure value (for example, 150 mmHg), the pulse wave amplitude A_n approaches zero or is lower than the preset threshold; at this time, the system determines that the AOP point has been found, and the dynamic pressurization and data collection process of S112 is ended, providing a solid data foundation for the next step of accurately calculating the AOP and setting the training pressure.

[0027] Reference Figure 3 In step S12, the specific steps are: S121: Collecting a plurality of physiological data of the user, determining a current physical state of the user according to the plurality of physiological data of the user and a corresponding age, determining a compression coefficient of a target limb based on a plurality of pulse data and the current physical state of the user, and determining a training pressure event of the limb compression device according to a mapping relationship between the compression coefficient of the target limb and the training pressure event; S122: Identifying the training pressure event and outputting a plurality of project marks, and determining a plurality of training pressure projects according to a matching between the plurality of project marks and a training database of the limb compression device.

[0028] In the embodiments of the present application, a plurality of physiological data of the user are collected, a current physical state of the user is determined according to the plurality of physiological data of the user and a corresponding age, a compression coefficient of a target limb is determined based on a plurality of pulse data and the current physical state of the user, and a training pressure event of the limb compression device is determined according to a mapping relationship between the compression coefficient of the target limb and the training pressure event, which is compatible with the overall consideration of the mapping relationship between the compression coefficient of the target limb and the training pressure event, and ensures the accuracy of the training pressure event of the limb compression device.

[0029] At this time, the system synchronously collects resting heart rate, heart rate variability, recent sleep quality and duration, skin electrical response or skin temperature and other physiological data through device body sensor and Bluetooth communication with external devices such as smart watches, and inputs these data into a built-in multi-parameter weighted evaluation model. The model normalizes each data, gives different weights according to its correlation with the training readiness, and adjusts the expected range of each indicator according to the user's age. The model outputs a comprehensive current physical state quantitative score or a classification label to scientifically evaluate whether the user is suitable for high-intensity training.

[0030] Based on the physical state score evaluated in the previous step and the arterial occlusion pressure of the target limb, the standard training pressure is individually fine-tuned to generate a compression coefficient. The calculation logic can use linear mapping to map the physical state score to a pre-set coefficient range, or use a piecewise function to set fixed coefficient values for different state threshold intervals. To ensure safety, the compression coefficient usually has an upper and lower limit to prevent overtraining or ensure minimum effective stimulation, ensuring that the training is both safe and effective, and matches the user's real-time bearing capacity.

[0031] The system internally stores a compression coefficient-training pressure event mapping table based on sports science theory and expert experience. The system matches the calculated compression coefficient with the table to determine the most suitable training target, such as mapping coefficients greater than or equal to 1.1 to high-intensity muscle-building training events, mapping coefficients between 0.9 and 1.1 to standard strength training events, and mapping coefficients less than 0.9 to recovery endurance training events. A training pressure event is a high-level instruction that defines the core goal of this training, marking the system's formal entry into the planning phase from the evaluation phase.

[0032] Specifically, the system has measured the AOP of the exerciser's arm as 150mmHg, and the limb compression device has synchronized the following data from the exerciser's smart watch through Bluetooth: resting heart rate 52bpm, RMSSD (HRV indicator) 85ms, sleep duration last night 8.2 hours, and deep sleep proportion 25%; the system's evaluation model receives these data and combines them with the exerciser's age of 30 to make a comprehensive calculation; the model determines that the HRV level is far above the average of people of the same age, the sleep quality is excellent, and the resting heart rate is low; the model outputs a current physical state score of 95 / 100 and a label of excellent.

[0033] The system obtains the physical state score of 95 and the AOP of 150mmHg as input parameters; during calculation, the system uses piecewise function logic; since the score of 95 is greater than 90, the system determines that the user is in excellent condition; according to the built-in rules, the compression coefficient is determined to be 1.15, which means that the system believes that today can withstand a training load that is 15% higher than the standard.

[0034] The system inputs the pressure coefficient 1.15 into the mapping table for querying; the query result shows that 1.15 falls within the coefficient interval of the high-intensity muscle-building training event; therefore, the system formally determines that the training pressure event of this training is the high-intensity muscle-building training event; through the three sub-steps of S121, the system not only objectively evaluates the real-time physiological state of the exerciser, but also converts it into a clear and scientific training target, laying a solid foundation for subsequent generation of specific and executable training pressure items.

[0035] Further, the training pressure event is identified, and multiple item tags are output, and the multiple training pressure items are determined according to the matching of the multiple item tags and the training database of the limb compression equipment, which is compatible with the overall consideration of the matching of the multiple item tags and the training database of the limb compression equipment, ensures the accuracy of the multiple training pressure items, and at the same time, introduces multiple pulse data of the target limb, controls the training pressure event of the limb compression equipment, and improves the identification accuracy of the multiple training pressure items.

[0036] At this time, the macro training pressure event determined in S121 is structurally analyzed and decomposed into a set of machine-readable and semantically clear item tags; the intelligent exercise prescription module in the central control module analyzes the input event and adds a series of structured key-value pair tags according to the preset rule base, which comprehensively describes the core attributes of the training, including the target limb, the training target, the intensity level, the metabolic type, and the user state inherited from S121; the system outputs a set containing multiple item tags, such as [{limb part: upper limb}, {training target: muscle building}, {intensity level: high}], and this tag set is the query condition for database matching in the next step.

[0037] Using the item tags generated in the previous step, efficient querying and matching are performed in a structured training database, and one or more specific and executable training pressure items are finally output; the training database is a pre-set knowledge base that stores a large number of scientifically verified training scheme templates; the system uses a weighted matching algorithm to compare the input item tags with the tags of each template in the database, calculates the similarity score and selects the template with the highest matching degree; after successful matching, the system instantiates the parameters in the template, such as replacing {pressure: %AOP} in the template with a specific value calculated according to the user AOP and the pressure coefficient; the system outputs a training pressure item set containing all execution details, and each item is an independent parameter module that together constitutes a complete training instruction.

[0038] Specifically, the system has determined a high-intensity muscle-building training event for the user's arms. After receiving the instruction of the high-intensity muscle-building training event, the intelligent exercise prescription module analyzes the event and outputs the following item tag set: {body part: upper limbs}, {training goal: muscle building}, {intensity level: high}, {user status: excellent}. This tag set will be used as a query condition to find the most suitable training program in the database.

[0039] The system uses the above-mentioned tag set as a query condition to search in the training database. A template named upper limb high-intensity muscle-building program V2.1 in the database matches all the tags. The system reads the parameters of the template and performs instantiation calculation combined with the user's specific data: the pressure percentage 50% AOP in the template is instantiated as 150 mmHg (user AOP) * 50% * 1.15 (pressure coefficient) ≈ 86 mmHg. Since the user's status is excellent, the system decides to use the upper limit values of the number of groups and times in the template.

[0040] The system generates and outputs the following specific training pressure items, which will directly guide the subsequent training execution: item 1 (core pressure) {target pressure: 86 mmHg}; item 2 (action mode) {training mode: super slow resistance, centripetal phase: 5-6 seconds, centrifugal phase: 5-8 seconds}; item 3 (group arrangement) {training group number: 4 groups, number of repetitions per group: 6-8 times, group interval: 90 seconds}; item 4 (closed-loop feedback) {real-time monitoring: pressure fluctuation ± 3 mmHg, action quality: evaluated by inertial sensor}. Through this process, the system successfully converts the abstract training intention into a complete, personalized, and accurately executable training program by the device.

[0041] Reference Figure 4 In step S13, the specific steps are: S131: Real-time monitoring of the limb compression device and obtaining a plurality of working data of the limb compression device, determining the current working state of the limb compression device based on the plurality of working data of the limb compression device and the working history of the limb compression device; S132: Collecting a plurality of training pressure items, determining a training data set based on the plurality of training pressure items and the current working state of the limb compression device, and determining a training report of the target limb based on the training data set, the overall morphology of the target limb, and the corresponding compression condition.

[0042] In the embodiment of the present application, the limb compression device is monitored in real time, and a plurality of working data of the limb compression device is acquired. The current working state of the limb compression device is determined based on the plurality of working data of the limb compression device and the working history of the limb compression device. The plurality of working data of the limb compression device and the working history of the limb compression device are considered as a whole, and the accuracy of the current working state of the limb compression device is ensured.

[0043] At this time, the system monitors the running state of the device itself in real time through multi-sensor fusion in high frequency and multi-dimension; the pressure acquisition component continuously acquires the real-time pressure value in the cuff at a high frequency, which is both the feedback signal of the PID closed-loop control and the core data for judging the stability of the pressure; at the same time, the system monitors the electrical parameters such as the PWM duty cycle, working current and voltage of the power driving circuit of the central control module driving the miniature air pump to identify potential load abnormalities; in addition, the system indirectly judges the air path integrity by analyzing the trend of the pressure data, and monitors the system environment and power consumption in real time through the built-in temperature sensor and power meter, providing bottom protection for the safety of the training process and the reliability of the data.

[0044] The real-time acquired working data and the historical data of the device are analyzed by fusion, and the current working state of the device is accurately and dynamically evaluated and classified through an intelligent state machine model; the system retrieves the working history data of the device from the non-volatile memory, including the cumulative running time, total use time, historical fault records, etc.; then, the real-time data and the historical data are comprehensively processed by using a weighted fusion algorithm, and different risk weights are given to different data; the state machine model in the system divides the current working state of the device into three states of normal, sub-health / warning or abnormal / failure according to the results of the fusion analysis, and each state has a clear definition and corresponding subsequent processing strategy.

[0045] Specifically, the exerciser has started to perform 86mmHg, 4 sets x 6-8 times, super slow arm BFR training, and the central control module of the limb compression device continuously performs multi-dimensional monitoring within 15 minutes of the exerciser's bicep curl training; the pressure sensor acquires data at a frequency of 100Hz, and the system calculates the pressure average value as 86.2mmHg and the standard deviation as 1.1mmHg in real time, indicating that the pressure is highly stable; the system monitors that the PWM duty cycle of the air pump maintaining the pressure is dynamically adjusted between 10%-15%, and the average working current is 150mA, which is far below the protection threshold of 300mA; during the resting period between sets, the system detects that the pressure only decreases by 0.5mmHg within 60 seconds, determining that the air path sealing is good; at the same time, the MCU temperature is stable at 41℃, and the battery power consumption is about 5%, all parameters are within the normal range.

[0046] The system fuses the real-time data above with the work history retrieved from the storage; the total usage time of the device is 200 hours, no historical failure record, and it has been used for 50 hours after the last maintenance; the state machine model of the system comprehensively evaluates all the data: the pressure stability is better than the target value, the air pump current and temperature are in the healthy interval, the leakage rate is extremely low, and the work history shows that the device is in the early stage of the life cycle; based on all the above information, the state machine model determines that the current working state of the limb compression device is normal; on the mobile APP, the device status icon is displayed in green, the user can continue training with peace of mind, and all the monitoring data are recorded synchronously to generate subsequent training reports.

[0047] Further, a plurality of training pressure projects are collected, a training data set is determined according to the plurality of training pressure projects and the current working state of the limb compression device, and a training report of the target limb is determined according to the training data set, the overall morphology of the target limb and the corresponding compression condition, which comprehensively considers the training data set, the overall morphology of the target limb and the corresponding compression condition, ensures the accuracy of the training report of the target limb, and introduces the current working state of the limb compression device, which comprehensively considers the current working state of the limb compression device, the plurality of training pressure projects and the overall morphology of the target limb, and improves the accuracy of the training report of the target limb.

[0048] At this time, the discrete data points generated during the training process and from different sources are integrated into a structured, multi-dimensional training data set; the system collects and integrates the preset training scheme (plan baseline) from S12, the device state evaluation result (execution quality label) from S131, and the high-frequency time series data recorded during the training process (such as pressure-time curve, heart rate-time curve and action data); through a unified time stamp reference, all time series data are accurately aligned and structured according to the preset mode to form a data set containing multiple dimensions such as plan, actual, physiological response and device state, providing a unified and high-quality data source for subsequent deep analysis and report generation.

[0049] A comprehensive and visual training report is finally generated by using the built-in intelligent analysis model to deeply mine the training data set and combining the individual characteristics of the user; the analysis engine of the system performs multi-level calculation and evaluation on the data set, including execution degree analysis by comparing the plan baseline with the actual data, physiological load analysis by analyzing the heart rate curve to quantify the physiological load intensity, and individual adaptability analysis by combining the user's limb morphology to evaluate the adaptability of the scheme; then, the analysis engine fills all the calculation results and evaluation conclusions into a visual report template to generate the final report containing multiple modules such as training abstract, key indicator dashboard, execution quality evaluation, training effect estimation and personalized suggestions.

[0050] Specifically, the exerciser has successfully completed the arm BFR training, the device state is normal, and the central control module of the limb compression device immediately performs data integration to build a complete training data set; this set contains {target pressure: 86 mmHg, planned group number: 4, planned number: 6-8} as the baseline of the plan, {state: normal, pressure stability: excellent} as the execution quality, and the complete pressure curve, heart rate curve and action rhythm data as the actual data; all these data are accurately aligned through a unified timestamp to form a structured data object, preparing for subsequent analysis.

[0051] The analysis engine of the system processes this data set and generates a detailed training report in the APP; in the execution degree analysis, the system calculates that the actual average pressure is 86.2 mmHg, with a very small deviation from the target value, and the pressure execution precision score is as high as 98%, and the user has completed all 4 groups of training, with a plan completion degree of 100%; in the physiological load analysis, the system extracts the heart rate data and calculates the average heart rate of 120 bpm and the peak heart rate of 138 bpm, determining that the heart rate interval has reached the expected medium-high intensity physiological stimulation; in the individual adaptability analysis, the system combines the arm circumference data of 32 cm in the user's profile to assess that the pressure of 86 mmHg is well matched with the user's limb morphology; the report is presented in a visual form: the summary page shows that the training has been completed, the dashboard displays smooth pressure and heart rate curves, the evaluation page gives a training completion degree score: 98 points (excellent) evaluation, and the suggestion page prompts that the training effect is excellent, and suggests that the same pressure can be maintained or the concentric time can be shortened to further improve the strength performance next time; through this process, the system converts a complex physiological and physical process into an intelligent report that is understandable, traceable and can guide future actions for the user.

[0052] Reference Figure 5 In step S14, the specific steps are: S141: Collect the training report of the target limb and detect the training report of the target limb, determine a plurality of key content areas according to the detection of the training report of the target limb, determine corresponding training features based on the recognition of each key content area, and collect a plurality of training features; S142: Collect the compression condition of the target limb, determine a first sub-pressure curve based on the plurality of training features and the compression condition of the target limb, determine a second sub-pressure curve based on the plurality of training features and the age of the user, and determine the pressure curve of the target limb based on the first sub-pressure curve and the second sub-pressure curve.

[0053] In the embodiments of the present application, a training report of a target limb is collected, and the training report of the target limb is detected, a plurality of key content regions are determined according to the detection of the training report of the target limb, and corresponding training features are determined based on the recognition of each key content region, so as to collect a plurality of training features. The overall consideration of the recognition of each key content region is compatible, and the accuracy of the corresponding training features is ensured.

[0054] At this time, the system parses the underlying format of the report, whether it is structured data or semi-structured layout, and builds a document object model or data tree; then, through a combination of template matching, natural language processing or computer vision technologies, the system divides different content blocks according to key fields, text topics or visual features; after recognition, the system defines clear boundaries for each key content region and labels it with semantic tags, such as {region ID:01, region type: execution data}, laying the foundation for subsequent accurate data extraction.

[0055] The system has a built-in feature extraction rule engine for the BFR training field, which includes a series of regular expressions, keyword matching lists and numerical extraction patterns; the engine performs multi-modal feature extraction, such as extracting average pressure and other numerical features in the execution data region, identifying excellent rating words in the evaluation region and mapping them to numerical scores, and even parsing the peak value of the heart rate curve from the chart; all the training features extracted from different regions are finally collected into a structured data set, forming a complete training feature set for subsequent steps.

[0056] Specifically, the exerciser has seen a detailed training report on the APP, and the central control module of the limb compression device has obtained the UI layout data of the training report on the APP through the Bluetooth communication module; the template matching algorithm of the system starts to work and identifies three blocks with clear IDs in the report: a block with the title of execution overview is marked as key content region 1 (execution data); a block with the title of physiological response is marked as key content region 2 (physiological data); a block with the title of AI evaluation and suggestion is marked as key content region 3 (evaluation and suggestion).

[0057] The system's feature extraction rule engine enters each identified area in turn and begins to extract specific training features; in the execution data area, the engine matches to the target pressure: 86 mmHg and extracts the feature {feature: target pressure, value: 86}, and matches to the actual average pressure: 86.2 mmHg and extracts the feature {feature: actual average pressure, value: 86.2}; in the physiological data area, the engine matches to the average heart rate: 120 bpm and extracts the feature {feature: average heart rate, value: 120}, and matches to the peak heart rate: 138 bpm and extracts the feature {feature: peak heart rate, value: 138}; in the evaluation and recommendation area, the engine matches to the training completion score: 98 points (excellent) and extracts the feature {feature: training completion score, value: 98}; the system aggregates all the extracted features into a highly structured training feature set, providing accurate and reliable input data for the next step of dual sub-pressure curve modeling.

[0058] Further, the compression condition of the target limb is collected, the first sub-pressure curve is determined according to the plurality of training features and the compression condition of the target limb, the second sub-pressure curve is determined according to the plurality of training features and the age of the user, and the pressure curve of the target limb is determined based on the first sub-pressure curve and the second sub-pressure curve, which takes into account the overall consideration of the first sub-pressure curve and the second sub-pressure curve, and ensures the accuracy of the pressure curve of the target limb.

[0059] At this time, the system takes the execution features extracted in S141 and the compression condition of the target limb (such as cuff width, material, arm circumference) as input parameters, and uses a physical-empirical hybrid model based on fluid mechanics, material mechanics and a large amount of experimental data for calculation; the model outputs a curve with theoretical pressure value as the horizontal axis and physical performance score as the vertical axis, and the score is a weighted function of pressure stability, distribution uniformity and energy efficiency, and the peak point of the curve represents the pressure point or interval with the optimal physical performance under the current compression condition.

[0060] The system takes the physiological features, evaluation features and user age extracted in S141 as input parameters, and uses an AI prediction model based on large-scale training data for analysis; the model masters the complex nonlinear relationship between pressure, physiological response and training effect, and outputs a curve with theoretical pressure value as the horizontal axis and physiological benefit score as the vertical axis, and the score is a comprehensive function of muscle growth potential, cardiovascular system load and subjective fatigue prediction, and the peak point of the curve represents the pressure point or interval with the maximum physiological benefit and relative safety under the current user state.

[0061] The first and second sub-pressure curves are integrated to generate a more robust and personalized target pressure curve; the system uses a weighted fusion algorithm to assign different weights to the two sub-curves according to the user's training goals, for example, a muscle gain goal will give a higher weight to the physiological curve, while a rehabilitation goal will give a higher weight to the physical curve and set a stricter safety upper limit; the algorithm weights and sums the scores of the two sub-curves at each pressure point according to the weights to generate a new fused curve; the final target pressure curve is presented to the user, which usually includes an optimal recommended interval, an acceptable interval, and a clear safety warning line, providing intuitive, fine, and safe pressure adjustment guidance for the user.

[0062] Specifically, the system has extracted all the key features from the training report, and the system obtains the training features {pressure stability indicator: 1.2 mmHg} and the compression situation {cuff width: 10 cm, arm circumference: 32 cm} as input; the physical-experience model calculates that for a combination of 32 cm arm circumference and 10 cm cuff, the limiting effect is insufficient when the pressure is below 80 mmHg, and when the pressure is above 95 mmHg, the sleeve tension is too large, causing uneven pressure distribution and decreased stability; based on this, the system generates a first sub-pressure curve, the peak of which has the highest physical efficiency in the 84-90 mmHg interval.

[0063] The system obtains the training features {average heart rate: 120 bpm}, {peak heart rate: 138 bpm}, and the user's age of 30 years as input; the physiological model analysis shows that for a 30-year-old user, the heart rate response of this training indicates that a pressure of 86 mmHg can already stimulate the user's heart rate into an efficient muscle gain stimulation interval; the model predicts that if the pressure exceeds 92 mmHg, the heart rate will enter a high-risk interval, and the recovery will be significantly slower; based on this, the system generates a second sub-pressure curve, the peak of which has the highest physiological benefit in the 85-91 mmHg interval.

[0064] Since the exerciser's goal is to gain muscle, the system sets the weights as follows: the second (physiological) curve weight is 70%, and the first (physical) curve weight is 30%; the system performs weighted calculation on the two curves at each pressure point to generate the final target pressure curve, which is visualized on the APP; the curve clearly identifies the green optimal recommended interval as 86-89 mmHg (the region with the highest overlap of the peaks of the two curves), the blue acceptable interval as 84-92 mmHg, and the red warning line (<83 mmHg insufficient effect, >93 mmHg high risk), this fused target pressure curve provides a more scientific, safer, and more personalized decision basis for the exerciser's next training than a single numerical value.

[0065] Reference Figure 6 In step S15, the specific steps are: S151: determine a plurality of key pressure nodes according to the identification of the pressure curve of the target limb, determine intelligent control content of each key pressure node based on the node position, corresponding node form and current working state of the limb compression device of each key pressure node, and construct an intelligent control event of the limb compression device based on the intelligent control content of each key pressure node; S152: real-time monitoring of the limb compression device and collection of abnormal signals of the limb compression device, determination of dynamic early warning content of the limb compression device according to the intelligent control event and the current body state of the user, and determination of a dynamic early warning system of the limb compression device according to the dynamic early warning content of the limb compression device and the abnormal signals of the limb compression device.

[0066] In the embodiment of the present application, a plurality of key pressure nodes are determined according to the identification of the pressure curve of the target limb, intelligent control content of each key pressure node is determined based on the node position, corresponding node form and current working state of the limb compression device of each key pressure node, and an intelligent control event of the limb compression device is constructed based on the intelligent control content of each key pressure node, which is compatible with the overall consideration of the node position, corresponding node form and current working state of the limb compression device of each key pressure node, and ensures the accuracy of the intelligent control content of each key pressure node.

[0067] At this time, the continuous target pressure curve generated in S14 is discretized and characterized, and key pressure nodes with specific tactical or safety significance are automatically identified from the curve through algorithm analysis; the system uses a numerical analysis algorithm to calculate the first and second derivatives of the curve to identify slope change points and inflection points; the built-in node classifier identifies several types of key pressure nodes such as optimal interval boundary nodes, safety warning nodes, slope mutation nodes and efficiency peak nodes according to the geometric shape of the curve and the preset threshold, and these nodes are the trigger nodes of the subsequent intelligent control logic.

[0068] Each identified key pressure node is assigned a specific and executable intelligent control content, and this process will consider the intrinsic properties of the node and the real-time health status of the device; the system establishes an attribute triple for each key node, including node position, node form and trigger direction; at the same time, the system obtains the current working state of the device from S131, which will affect the aggressiveness of the control strategy; the built-in rule engine maps the node attributes and device state to specific control actions, and the control content can include sending interaction instructions to the APP, highlighting on the APP interface, sending control commands to the actuator or marking events in the log.

[0069] All the intelligent control contents are integrated into one or more complete intelligent control events that can be monitored and executed by the central control module in real time; each intelligent control event is constructed as a condition-action (IF-THEN) structure, the condition part is the trigger logic of the key node, and the action part is the specific control content; all the constructed intelligent control events are loaded into the real-time event monitoring queue of the central control module, and the main loop of the MCU checks whether the trigger conditions of these events are met at a high frequency; different events can be assigned different priorities, for example, the event related to the safety warning node has the highest priority and will be executed immediately once triggered.

[0070] Specifically, the system has generated a target pressure curve for the exerciser, including an optimal interval (86-89 mmHg) and a warning line (93 mmHg), and the system's curve analysis algorithm scans the target pressure curve and successfully identifies three key pressure nodes: node A, located at 89 mmHg, is an upper limit node of the optimal interval; node B, located at 86 mmHg, is a lower limit node of the optimal interval; and node C, located at 93 mmHg, is a safety warning node.

[0071] The system obtains the current working state of the device as normal and determines the control content for each node; for node A (89 mmHg), when the pressure crosses upwards, since the device state is normal, the system decides to adopt a gentle reminder strategy, and the control content is to send a vibration and text prompt to the APP that the upper limit of the optimal interval has been reached; for node C (93 mmHg), when the pressure crosses upwards, this is the highest safety level and must be intervened immediately, and the control content is to immediately open the electronic pressure relief valve to 88 mmHg and send a red pop-up window and alarm sound to the APP to prompt that the pressure is too high and has been automatically relieved.

[0072] The system encapsulates the above control content into two intelligent control events and loads them into the monitoring queue of the MCU; the logic of event 1 (reminder event) is: if the real-time pressure is greater than or equal to 89 mmHg and the trigger direction is upward, send a reminder to the APP; the logic of event 2 (emergency intervention event) is: if the real-time pressure is greater than or equal to 93 mmHg and the trigger direction is upward, activate the pressure relief valve and send an emergency alarm to the APP; through this process, the system successfully converts a static pressure curve into a set of dynamic intelligent control logic with forward-looking and safety protection capabilities, laying a foundation for the automation and safety of the next training.

[0073] Further, the limb compression equipment is monitored in real time, and an abnormal signal of the limb compression equipment is collected, a dynamic early warning content of the limb compression equipment is determined according to the intelligent control event and the current body state of the user, a dynamic early warning system of the limb compression equipment is determined according to the dynamic early warning content of the limb compression equipment and the abnormal signal of the limb compression equipment, the dynamic early warning content of the limb compression equipment and the abnormal signal of the limb compression equipment are considered as a whole, the accuracy of the dynamic early warning system of the limb compression equipment is ensured, and meanwhile, the pressure curve of the target limb is introduced, the intelligent control event of the limb compression equipment is further controlled, and the accuracy of the dynamic early warning system of the limb compression equipment is improved.

[0074] At this time, the system predefines multiple abnormal signal modes at the firmware level, which has a higher priority than the regular working state evaluation, including a pressure sudden drop signal, a pressure out-of-control signal, a sensor failure signal and a power abnormal signal; the central control module polls each sensor and self-checking module at a high frequency, and once a signal conforming to any mode is detected, it is immediately marked as an abnormal signal, with a timestamp and signal type, and stored in a high-priority system event queue.

[0075] According to the intelligent control event generated in S151 and the current body state of the user, it is determined what kind of early warning should be issued in a specific situation; the system has a built-in multi-modal early warning content library, including different levels of visual, auditory and tactile feedback templates; when the intelligent control event is triggered, the system will call the corresponding warning content of the warning level; at the same time, if the user's heart rate and other physiological data suddenly rise to a dangerous range, even if the device pressure has not reached the warning line, the system will generate a physiological overload early warning content; when multiple conditions occur at the same time, the system will perform content fusion to generate composite information.

[0076] All early warning contents are layered and graded to form a structured and rapid response dynamic early warning system; the system adopts a multi-level layered architecture to ensure the appropriateness and efficiency of the response; the first level (warning level) is usually triggered by the intelligent control event, and uses a non-intrusive APP prompt; the second level (warning level) is triggered by the user's physiological state abnormality or the first level warning being ignored, and the response mode is upgraded to sound plus vibration; the third level (emergency level) is triggered only by the collected abnormal signal, which is the highest priority and will bypass all other logic to immediately execute the most intense physical intervention and the strongest alarm; all early warning rules, content templates and grading logic are solidified into firmware, and once the device starts working, the system is activated.

[0077] Specifically, the exerciser starts the next arm BFR training, the system has loaded the intelligent control event generated in S151, and the dynamic early warning system in S152 is activated; during the training process, the central control module of the limb compression device continuously performs monitoring at a high frequency; during the second bicep curl of the third set of training, due to the excessive amplitude of the action, the sleeve accidentally slips off, and the pressure sensor detects that the pressure drops from 88 mmHg to 15 mmHg within 0.2 seconds, the system immediately marks and generates a pressure drop abnormal signal; at the same time, the system also continuously monitors whether there is a pressure runaway signal that the air pump works at full power but the pressure does not change, and other hardware failure signals.

[0078] The exerciser exerts excessive force, causing the pressure in the sleeve to slowly rise to 89.5 mmHg, triggering the intelligent control event in S151; the system immediately generates a first-level warning content from the content library: the pressure value in the APP interface turns yellow, accompanied by a short vibration and a text prompt that the pressure is close to the upper limit; in another scenario, although the pressure is within the normal range, the exerciser's watch synchronously displays that the heart rate suddenly rises to a dangerous interval; based on the current physical state of the user, the system generates a second-level warning content: the APP pops up a red warning window, the device emits a beeping alarm sound, the APP continuously vibrates, and displays that the heart rate is too high, please check the status!

[0079] The system has built-in a third-level warning system and responds in real time; when the pressure reaches 89 mmHg, the system performs a first-level response, triggering only the APP vibration and text prompt; when the heart rate is too high, the system performs a second-level response, triggering sound, vibration, and a red pop-up window; when the pressure drop abnormal signal occurs, the warning system immediately identifies it as the highest priority event, bypasses all other logic, and directly performs a third-level response: the APP displays a full-screen red light, the device emits the longest and sharpest alarm sound, the APP continuously vibrates at maximum intensity, and the safety mode is immediately activated physically, locking all buttons, and displaying on the APP that the sleeve has fallen off / severe leakage, and the training is terminated! Through this process, the system builds a complete closed loop from reminders to warnings to emergency interventions, ensuring that the most timely and appropriate protection is provided in any abnormal situation.

[0080] Please refer to Figure 7 , Figure 7 is a structural composition diagram of the intelligent control system of the limb compression device in the embodiment of the present application; the intelligent control system of the limb compression device comprises: a pulse data module 21, configured to, when the limb compression device is worn on a target limb of a user, determine a compression mode of the limb compression device on the target limb based on the limb compression device, the target limb, and corresponding detection items, and trigger dynamic compression of the limb compression device on the target limb to determine a plurality of pulse data of the target limb; The training pressure item module 22 is configured to determine a training pressure event of the limb compression device according to the plurality of pulse data and the current physical state of the user, and determine a plurality of training pressure items based on the identification of the training pressure event; The training report module 23 is configured to collect a plurality of working data of the limb compression device, determine a current working state of the limb compression device according to the plurality of working data of the limb compression device, and determine a training report of the target limb according to the current working state of the limb compression device, the plurality of training pressure items and the overall shape of the target limb. The pressure curve module 24 is configured to determine a plurality of training features according to the detection of the training report in the training report of the target limb, and determine a pressure curve of the target limb according to the plurality of training features, the compression state of the target limb and the age of the user. The dynamic early warning system module 25 is configured to determine an intelligent control event of the limb compression device based on the pressure curve of the target limb and the current working state of the limb compression device, and determine a dynamic early warning system of the limb compression device according to the intelligent control event, an abnormal signal of the limb compression device and the current physical state of the user.

[0081] Any combination of the technical features of the above embodiments is possible. In order to make the description simple, not all combinations of the technical features in the above embodiments are described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present disclosure.

Claims

1. An intelligent control method for a limb compression device, characterized in that, include: When the limb compression device is worn on the user's target limb, the compression mode of the limb compression device on the target limb is determined based on the limb compression device, the target limb and the corresponding detection items, and the dynamic compression of the target limb by the limb compression device is triggered to determine multiple pulse data of the target limb. The training stress events of the limb compression device are determined based on multiple pulse data and the user's current physical state, and multiple training stress items are determined based on the identification of the training stress events. Collect multiple working data from the limb compression device, determine the current working status of the limb compression device based on the multiple working data, and determine the training report of the target limb based on the current working status of the limb compression device, multiple training pressure items, and the overall shape of the target limb. In the training report of the target limb, multiple training features are determined based on the detection of the training report, and the pressure curve of the target limb is determined based on the multiple training features, the pressure status of the target limb, and the age of the user. Based on the pressure curve of the target limb and the current working status of the limb compression device, the intelligent control event of the limb compression device is determined. Based on the intelligent control event, the abnormal signal of the limb compression device and the user's current physical state, the dynamic early warning system of the limb compression device is determined.

2. The intelligent control method for the limb compression device according to claim 1, characterized in that, When the limb compression device is worn on the user's target limb, the compression mode of the limb compression device on the target limb is determined based on the limb compression device, the target limb, and the corresponding detection items, and the dynamic compression of the target limb by the limb compression device is triggered to determine multiple pulse data of the target limb, including: The limb compression device is worn on the user's target limb and is placed around the outer periphery of the user's target limb; the detection items of the user's target limb are collected, and multiple compression stage events are determined based on the limb compression device, the target limb and the corresponding detection items; The compression mode of the limb compression device on the target limb is determined based on multiple compression phase events; the dynamic compression of the target limb is triggered based on the compression mode, and multiple pulse data of the target limb are collected during the compression process.

3. The intelligent control method for the limb compression device according to claim 1, characterized in that, The process involves determining training stress events for the limb compression device based on multiple pulse data points and the user's current physical state, and then identifying multiple training stress items based on the recognition of these training stress events, including: Collect multiple physiological data from the user, determine the user's current physical state based on the multiple physiological data and the corresponding age, determine the pressure coefficient of the target limb based on multiple pulse data and the user's current physical state, and determine the training pressure event of the limb pressure device based on the mapping relationship between the pressure coefficient of the target limb and the training pressure event. The training stress event is identified, and multiple item tags are output. Multiple training stress items are determined based on the matching of multiple item tags with the training database of limb stress devices.

4. The intelligent control method for the limb compression device according to claim 1, characterized in that, The process involves collecting multiple operational data points from the limb compression device, determining the current operational status of the limb compression device based on this data, and generating a training report for the target limb based on the current operational status of the limb compression device, multiple training pressure items, and the overall morphology of the target limb. This report includes: The system monitors the limb compression device in real time and acquires multiple operational data points from the device. Based on these data points and the device's operational history, the system determines the device's current operational status.

5. The intelligent control method for the limb compression device according to claim 4, characterized in that, The process of collecting multiple operational data points from the limb compression device, determining the current operational status of the limb compression device based on these data points, and generating a training report for the target limb based on the current operational status of the limb compression device, multiple training pressure items, and the overall morphology of the target limb, also includes: Multiple training stress items are collected, and a training data set is determined based on the multiple training stress items and the current working status of the limb pressure device. A training report for the target limb is then determined based on this training data set, the overall shape of the target limb, and the corresponding pressure conditions.

6. The intelligent control method for the limb compression device according to claim 1, characterized in that, In the training report for the target limb, multiple training features are determined based on the detection in the training report. A pressure curve for the target limb is determined based on these multiple training features, the pressure status of the target limb, and the user's age, including: The training report of the target limb is collected and detected. Based on the detection of the training report of the target limb, multiple key content regions are identified. Based on the identification of each key content region, the corresponding training features are determined to collect multiple training features.

7. The intelligent control method for the limb compression device according to claim 6, characterized in that, The training report for the target limb, which determines multiple training features based on the training report, and determines the pressure curve of the target limb based on the multiple training features, the pressure status of the target limb, and the user's age, also includes: The pressure conditions of the target limb are collected. A first-level pressure curve is determined based on multiple training features and the pressure conditions of the target limb. A second-level pressure curve is determined based on multiple training features and the user's age. The pressure curve of the target limb is determined based on the first-level and second-level pressure curves.

8. The intelligent control method for the limb compression device according to claim 1, characterized in that, The intelligent control event of the limb compression device is determined based on the pressure curve of the target limb and the current working status of the limb compression device. A dynamic early warning system for the limb compression device is then determined based on this intelligent control event, abnormal signals from the limb compression device, and the user's current physical state. This system includes: Based on the identification of the pressure curve of the target limb, multiple key pressure nodes are determined. Based on the node position, corresponding node shape and current working status of the limb pressurization device, the intelligent control content of each key pressure node is determined. Based on the intelligent control content of each key pressure node, the intelligent control events of the limb pressurization device are constructed.

9. The intelligent control method for the limb compression device according to claim 8, characterized in that, The method of determining intelligent control events for the limb compression device based on the pressure curve of the target limb and the current working status of the limb compression device, and determining a dynamic early warning system for the limb compression device based on the intelligent control events, abnormal signals of the limb compression device, and the user's current physical state, further includes: The system monitors limb compression devices in real time and collects abnormal signals from them. Based on the intelligent control event and the user's current physical state, it determines the dynamic warning content of the limb compression devices and establishes a dynamic warning system for them.

10. An intelligent control system for a limb compression device, characterized in that, The intelligent control system of the limb compression device is applied to the intelligent control method of the limb compression device as described in any one of claims 1-9.

Images