Intelligent ankle pump exercise training instrument and multi-mode linkage control system

Abstract

The invention discloses an intelligent ankle pump exercise training instrument and a multi-mode linkage control system, the intelligent ankle pump exercise training instrument comprises a foot supporting plate, a sole air bag structure is arranged on the surface of one side of the foot supporting plate, a detachable foot back fixing part is further arranged on the surface of one side of the foot supporting plate, and a foot back air bag is arranged in the foot back fixing part; the sole air bag structure is composed of a forefoot area air bag, an arch area air bag and a heel area air bag. The front foot area air bag, the arch area air bag, the heel area air bag and the instep air bag are further provided with pressure sensors, and the pressure sensors are connected with the main control module; according to the method, a real-time feedback mechanism of electromyographic signal-inflation pressure is established, muscle point features, pressure features and inertia features are fused, spasm is accurately recognized through a pressure gradient control algorithm, meanwhile, alternate inflation control over foot dorsum / sole anatomical partitions is achieved, and foot pressure sores of a patient can be completely eradicated.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent medical rehabilitation devices, and particularly relates to an intelligent ankle pump exercise trainer and a multi-modal linkage control system. Background Technique

[0002] Ankle pump exercise is a simple and important rehabilitative exercise. It mainly promotes blood circulation in the lower limbs, prevents blood clots from forming, and reduces swelling in the lower limbs through the movement of the ankle joint. Generally, patients after orthopedic surgery, hemiplegia, and those who have been bedridden for a long time need to perform ankle pump exercise training.

[0003] For example, an auxiliary ankle pump exercise rehabilitation trainer disclosed in the utility model with the authorization announcement number CN213049193U. This trainer includes a bottom plate, an elastic telescopic airbag, a foot push plate, and a pneumatic pressure regulating device. An elastic telescopic airbag is provided between the bottom plate and the foot push plate, and the air pressure in the elastic telescopic airbag is adjusted through the pneumatic pressure regulating device to make the foot push plate perform ankle pump exercise. This product can guide patients to perform ankle pump training, which is beneficial to strengthening the muscle strength training of the upper limbs and the compliance of the lower limb ankle pump exercise. However, the above technical solution can only inflate the airbag as a whole, resulting in pressure accumulation in the heel area, causing too high pressure in the heel area and making it easy to generate pressure sores in the heel area. At the same time, it cannot collect the foot pressure in zones, and cannot discriminate the spasm of the patient's muscle strength state, resulting in poor training effects. Therefore, we propose an intelligent ankle pump exercise trainer and a multi-modal linkage control system. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent ankle pump exercise trainer and a multi-modal linkage control system to solve the problems in the above background technique that the existing technology can only inflate the airbag as a whole, resulting in pressure accumulation in the heel area, causing too high pressure in the heel area and making it easy to generate pressure sores in the heel area. At the same time, it cannot collect the foot pressure in zones, and cannot discriminate the spasm of the patient's muscle strength state, resulting in poor training effects.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An intelligent ankle pump exercise trainer includes a foot support plate, the foot support plate is arranged on a bearing shaft, the bearing shaft is rotatably arranged on one side of the device housing, a sole airbag structure is arranged on one surface of the foot support plate, a detachable dorsal foot fixing part is also arranged on one surface of the foot support plate, and a dorsal foot airbag is arranged inside the dorsal foot fixing part;

[0006] The sole airbag structure is composed of a forefoot area airbag, an arch area airbag, and a heel area airbag. The forefoot area airbag, the arch area airbag, the heel area airbag, and the dorsal foot airbag are all connected to an air pump inside the external housing through air pipes, and a solenoid valve group is arranged on the air pipes. The solenoid valve group is connected to the main control module;

[0007] Pressure sensors are also provided on the front foot area airbag, the arch area airbag, the heel area airbag and the dorsal foot airbag, and the pressure sensors are connected to the main control module;

[0008] The main control module collects the pressure through the pressure sensors, and based on the anatomically driven pressure gradient model and the pressure gradient control algorithm, accurately identifies muscle spasms.

[0009] Preferably, an angle sensor is also provided on one side surface inside the device housing, and the output end of the angle sensor is connected to the input end of the main control module.

[0010] Preferably, an LCD touch screen is also provided on the external housing of the device, and the LCD touch screen is connected to the main control module.

[0011] An intelligent ankle pump motion training multi-modal linkage control system includes an Internet of Things module and a multi-modal sensor module. The Internet of Things module is connected to the APP in the doctor's mobile phone terminal through dual-mode communication of Bluetooth 5.2 and Wi-Fi 6. The material network module is also connected to the main control module. The multi-modal sensor module includes an electromyography sensor, and the output end of the electromyography sensor is connected to the input end of the main control module.

[0012] Preferably, the main control module is also used to process and extract the signals and features collected by the pressure sensor, the angle sensor and the electromyography sensor.

[0013] Preferably, the signals and features collected by the pressure sensor, the angle sensor and the electromyography sensor are processed and extracted, specifically:

[0014] Step 1: The main control module performs band-pass filtering on the signals collected by the pressure sensor, the angle sensor and the electromyography sensor;

[0015] Step 2: The main control module classifies the signals collected by the band-pass filtered pressure sensor, angle sensor and electromyography sensor, and classifies the corresponding signals into electromyography signals, inertial data and pressure signals;

[0016] Step 3: The main control module processes the electromyography signals through the wavelet packet decomposition algorithm, processes the inertial data through the quaternion solution algorithm, and processes the pressure signals through the spatial interpolation algorithm;

[0017] Step 4: Extract the features of the processed electromyography signals, inertial data and pressure signals;

[0018] Step 5: Fuse the extracted features, and through the anatomically driven pressure gradient model and the pressure gradient control algorithm, identify spasms and control the front foot area airbag, the arch area airbag, the heel area airbag and the dorsal foot airbag to inflate alternately.

[0019] Preferably, the pressure gradient model includes a dorsal foot airbag control model and a plantar foot airbag control model.

[0020] Preferably, the anatomical pressure gradient equation of the plantar foot airbag control model is:

[0021]

[0022] In the formula: P heel is the pressure in the heel area; P forefoot is the pressure in the forefoot area; p arch is the pressure in the arch area; BMI is the body mass index.

[0023] Preferably, the formula of the pressure gradient control algorithm is as follows:

[0024] P heel = 0.3×P forefoot + 5.

[0025] Compared with the prior art, the beneficial effects of the present invention are:

[0026] (1) By establishing a real-time feedback mechanism of electromyogram signal - inflation pressure, through the fusion of electromyogram characteristics, pressure characteristics and inertial characteristics, and then through the pressure gradient control algorithm, the present invention accurately identifies spasm.

[0027] (2) By accurately identifying spasm, the present invention realizes the alternate inflation control of the dorsal foot / plantar foot anatomical regions, can prevent the occurrence of pressure sores on the patient's feet, and at the same time improves the training effect of the patient.

[0028] (3) The present invention can be remotely controlled and monitored through the doctor's mobile phone terminal, and the corresponding parameter information of the system can be adjusted, further improving the training effect of the patient. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 is a schematic structural diagram of an intelligent ankle pump exercise trainer;

[0030] Figure 2 is a schematic structural diagram of a multi-modal linkage control system for intelligent ankle pump exercise training;

[0031] Figure 3 is a schematic control logic diagram of a multi-modal linkage control system for intelligent ankle pump exercise training;

[0032] Figure 4 is a schematic flow diagram of parameter setting at the doctor's end in the present invention;

[0033] Figure 5 is a schematic flow diagram of signal and data acquisition in the present invention;

[0034] Figure 6 Schematic diagram of signal processing and feature extraction in the present invention;

[0035] Figure 7 Schematic diagram of the control flow in the normal training mode of the present invention;

[0036] Figure 8 Schematic diagram of the spasm recognition process in the present invention;

[0037] Figure 9 Schematic diagram of the working process of Embodiment 1 in the present invention;

[0038] In the figure: 1, device housing; 2, dorsal foot airbag; 3, LCD touch screen; 4, connecting rod; 5, front foot area airbag; 6, foot support plate; 7, arch area airbag; 8, heel positioning frame; 9, heel area airbag; 10, bearing shaft; 11, dorsal foot fixing part. Detailed implementation manners

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0040] Please refer to Figure 1 , the present invention provides a technical solution: an intelligent ankle pump exercise trainer, the foot support plate 6 is arranged on the bearing shaft 10, the bearing shaft 10 is rotatably arranged on one side of the device housing 1, a plantar airbag structure is arranged on one side surface of the foot support plate 6, a detachable dorsal foot fixing part 11 is also arranged on one side surface of the foot support plate 6, and a dorsal foot airbag 2 is arranged inside the dorsal foot fixing part 11;

[0041] The plantar airbag structure is composed of a front foot area airbag 5, an arch area airbag 7 and a heel area airbag 9. The front foot area airbag 5, the arch area airbag 7, the heel area airbag 9 and the dorsal foot airbag 2 are all connected to an air pump inside the external housing through air pipes, and a solenoid valve group is arranged on the air pipes, and the solenoid valve group is connected to the main control module;

[0042] Pressure sensors are also arranged on the front foot area airbag 5, the arch area airbag 7, the heel area airbag 9 and the dorsal foot airbag 2, and the pressure sensors are connected to the main control module;

[0043] The main control module collects the pressure through the pressure sensors, and according to the anatomically driven pressure gradient model and the pressure gradient control algorithm, and accurately identifies muscle spasms.

[0044] Anatomical pressure gradient model

[0045] Basis for modeling:

[0046] Determine the optimal pressure range through plantar pressure distribution testing (n = 50 healthy subjects);

[0047] Finite element analysis verifies that low pressure (15 - 20 mmHg) at the heel can reduce the shear force by 62% (p < 0.01).

[0048] An angle sensor is also provided on one side surface inside the device housing 1, and the output end of the angle sensor is connected to the input end of the main control module; an LCD touch screen 3 is also provided on the external housing, and the LCD touch screen 3 is connected to the main control module. The angle sensor is a 9 - axis IMU sensor.

[0049] Furthermore, a connecting rod 4 is provided on one side of the foot support plate 6. The connecting plate is movably provided on one side of the device housing 1, which can improve the stability of the support plate during rotation during ankle pump exercise.

[0050] Furthermore, a heel positioning frame 8 is provided on one side of the upper surface of the foot support plate 6, which can position the patient's heel and improve the stability of heel placement.

[0051] Please refer to Figure 2 and Figure 3 , an intelligent multi - modal linkage control system for ankle pump exercise training, including an Internet of Things module and a multi - modal sensor module. The Internet of Things module is connected to the APP in the doctor's mobile phone terminal through Bluetooth 5.2 and Wi - Fi6 dual - mode communication. The Internet of Things module is also connected to the main control module. The multi - modal sensor module includes an electromyography sensor, and the output end of the electromyography sensor is connected to the input end of the main control module.

[0052] The overall operation method of this system is as follows:

[0053] Step 1: Equipment initialization and patient adaptation

[0054] Parameter setting:

[0055] Please refer to Figure 4 , the doctor sets the initial parameters through the APP on the mobile phone terminal:

[0056] Input / adjust parameters:

[0057] Heel reference pressure (18 mmHg);

[0058] Alternating frequency (12 times / min);

[0059] Spasm threshold (35 Hz);

[0060] The APP on the mobile phone terminal converts the above parameters into JSON data format and encrypts and sends them to the cloud server;

[0061] Patient wearing:

[0062] Place the foot on the foot support plate 6, fix the patient's foot through the dorsal foot fixing part 11, and then the pressure sensor detects the fitting degree of the patient's foot (starting when the pressure sensor feedback contact area > 90%).

[0063] Please refer to Figure 5 , Step 2: Real-time acquisition of multi-modal data

[0064] The electromyography sensor collects the electromyography signals of the tibialis anterior muscle / gastrocnemius muscle of the patient, the pressure sensor collects the pressure signals in the forefoot area, the arch area, and the heel area, and the angle sensor collects the three-dimensional angles and angular velocity data of the patient's foot during ankle pump movement, and transmits the above signals and data to the main control module.

[0065] Key parameters:

[0066] Main frequency range of electromyography signal: 0 - 500Hz;

[0067] Angle resolution of ankle joint: 0.1°;

[0068] Please refer to Figure 6 , Step 3: The main control module processes the signals and extracts features;

[0069] 1. The main control module performs band-pass filtering on the signals collected by the pressure sensor, angle sensor, and electromyography sensor;

[0070] 2. The main control module classifies the signals of the band-pass filtered pressure sensor, angle sensor, and electromyography sensor, and classifies the corresponding signals into electromyography signals, inertial data, and pressure signals;

[0071] 3. The main control module processes the electromyography signals through the wavelet packet decomposition algorithm, processes the inertial data through the quaternion solution algorithm, and processes the pressure signals through the spatial interpolation algorithm;

[0072] 4. Extract the features of the processed electromyography signals, inertial data, and pressure signals.

[0073] Electromyography signal processing:

[0074] Band-pass filtering (0.5 - 500Hz) → Power frequency notch filtering (50Hz) → Wavelet packet decomposition (db4 wavelet, 5-layer decomposition);

[0075] Pressure signal processing:

[0076] Band-pass filtering (0.5 - 500Hz) → Power frequency notch filtering (50Hz) → Spatial interpolation calculation;

[0077] Inertial data processing:

[0078] Band - pass filtering (0.5 - 500 Hz) → Power - frequency notch filtering (50 Hz) → Quaternion calculation;

[0079] Step 4: The main control module performs data fusion on the extracted features, and through an anatomy - driven pressure gradient model and a pressure gradient control algorithm, it identifies spasticity and controls the alternating inflation of the forefoot area airbag 5, the arch area airbag 7, the heel area airbag 9, and the dorsal foot airbag 2.

[0080] Please refer to Figure 7 , 4.1 Normal training mode

[0081] Alternating inflation logic:

[0082] Dorsal foot inflation: Drive ankle dorsiflexion (0° → 30°);

[0083] Plantar inflation: Pulse boost in the forefoot area (40 mmHg) stimulates the soleus muscle to contract;

[0084] 4.2 Spasticity emergency mode

[0085] When the spasticity signal (s) is detected:

[0086] Emergency pressure adjustment:

[0087] The pressure in the forefoot area immediately drops to 50% (40 → 20 mmHg);

[0088] The dorsal foot switches to a dynamic oscillation mode (0.5 Hz sine wave, amplitude ±5 mmHg);

[0089] Multi - level alarm:

[0090] Device side: Buzzer alarm (80 dB, 2 kHz);

[0091] Doctor side: The APP pushes a red alarm;

[0092] Step 5: Dynamic pressure recovery;

[0093] Adopt an exponential recovery algorithm to gradually recover the target pressure:

[0094] P(t) = P target ·(1 - e -t / 120 )

[0095] Recovery process monitoring:

[0096] Real - time detection of the change in the frequency of the electromyogram signal (Δf < 5 Hz / s);

[0097] When the angle changes suddenly (Δθ > 10° / s), pause the recovery;

[0098] Step 6: IoT data synchronization

[0099] Local storage: The SD card records the complete training data (sampling rate 10Hz);

[0100] Cloud upload:

[0101] Encrypted transmission: Uploaded through the CoAP protocol after being encrypted by AES-256-GCM;

[0102] Doctor-side monitoring:

[0103] The pressure-angle curve can be viewed in real time;

[0104] The device response delay is <5 seconds after modifying the parameters;

[0105] III. Key technological innovations

[0106] 1. Alternating inflation phase control technology

[0107] Dynamic phase difference adjustment: Automatically adjusts the dorsiflexion / plantar flexion inflation time ratio according to the patient's muscle strength;

[0108]

[0109] 2. Spatiotemporal alignment of multi-modal signals

[0110] Hardware-level synchronization: Achieves clock synchronization of each sensor (error <1ms) using GPSDO (oven-controlled crystal oscillator).

[0111] IV. Technical verification data

[0112] Test item index requirement table

[0113]

[0114]

[0115] I. System architecture and module functions

[0116] 1. Main control module

[0117] Receives surface electromyography (sEMG) signals, and extracts the energy ratio of the 32-48Hz spasm characteristic frequency band through an improved wavelet packet decomposition algorithm. The specific implementation is a 5-layer db4 wavelet decomposition.

[0118] Parses 9-axis IMU data, and calculates the three-dimensional motion angle and angular velocity of the ankle joint through the quaternion method.

[0119] Processor: STM32H743VI (dual-core ARM Cortex-M7, main frequency 480MHz);

[0120] Real-time performance: The task scheduling period is 10ms, and the interrupt response delay ≤10μs.

[0121] 2. Regional Airbag Drive Module

[0122] (1) Dorsal Foot Airbag Control Model

[0123] Target pressure of dorsal foot airbag (mmHg);

[0124] Preset maximum dorsiflexion angle (default 30°);

[0125] Proportional coefficient (default 0.8 mmHg / °).

[0126] (2) Plantar Foot Airbag Three-Zone Model

[0127] Anatomical pressure gradient equation:

[0128]

[0129] Parameter description:

[0130] P heel : Pressure in the heel area (mmHg), BMI range 18 - 35;

[0131] P forefoot : Pressure in the forefoot area (mmHg), used to activate the soleus muscle;

[0132] p arch : Pressure in the arch area (mmHg), providing anatomical support.

[0133] Control strategy:

[0134] Heel area: Constant low-pressure mode to prevent pressure sores;

[0135] Forefoot area: Pulse boost mode (1Hz square wave, duty cycle 40%), enhancing muscle strength training; Arch area: Follow-up support mode, dynamically adjusting pressure according to the dorsiflexion angle.

[0136] 3. Multimodal Sensor Module

[0137] (1) Bio-signal Acquisition Unit

[0138] Sensor technical parameter table

[0139]

[0140] 4. Internet of Things Module

[0141] (1) Communication Architecture

[0142] Local communication: Bluetooth 5.2 (transmission rate 2Mbps, distance 10m);

[0143] Remote management: Wi-Fi 6 (OFDMA technology, latency < 30ms);

[0144] Protocol stack: CoAP over DTLS (a dedicated protocol for the medical Internet of Things).

[0145] (2) Security mechanism

[0146] Security level and protection objective table

[0147] Security level Implementation method Protection objective Physical layer Hardware encryption chip (ATECC608A) Prevent side-channel attacks Transport layer TLS1.3 with ECDHE-ECDSA Prevent man-in-the-middle attacks Application layer AES-256-GCM end-to-end encryption Prevent data leakage

[0148] (3) Functions of the doctor's end

[0149] Remote parameter setting: pressure gradient ratio, training duration, spasm threshold;

[0150] Real-time monitoring: can view the heat map of the ankle joint movement trajectory;

[0151] Data analysis: automatically generate a rehabilitation progress report.

[0152] (2) Core algorithms

[0153] Pressure gradient control algorithm:

[0154] P heel = 0.3×P forefoot + 5 (unit: mmHg);

[0155] Specifically: according to the corresponding relationship between the pressure in the forefoot area and the pressure in the heel area in the above pressure gradient control algorithm, the airbags 5 in the forefoot area, the airbag 7 in the arch area, the airbag 9 in the heel area, and the dorsal airbag 2 are alternately inflated. That is, each time of inflation, the pressure in the heel area is equal to 0.3 times the pressure in the forefoot area + 5.

[0156] Please refer to Figure 8 , the spasm recognition process is as follows:

[0157] 1. The electromyography sensor collects the electromyography signal;

[0158] 2. The main control module processes the electromyography signal;

[0159] 3. Extract the frequency domain characteristics of the electromyography signal;

[0160] 4. The main control module judges the electromyography frequency. If the electromyography frequency is greater than 35 Hz, start a 3-second timing. If the electromyography frequency is less than 35 Hz, maintain the current mode;

[0161] 5. If the time when the electromyography frequency is greater than 35 Hz is greater than or equal to 3 seconds, it is judged as muscle spasm;

[0162] 6. Then trigger a pressure downshift, reduce the pressure to 50%, then perform vibration feedback, and finally perform dynamic pressure recovery.

[0163] (3) Internet of Things functions

[0164] - Data upload frequency: 1 time / minute (rest mode) → 10 times / minute (training mode);

[0165] - Adjustable parameters on the doctor side: pressure gradient ratio, alternating frequency, single training duration.

[0166] 3. Beneficial effects

[0167] Based on the biomechanical feedback control theory, the present invention realizes dorsal-plantar alternating inflation through multi-modal sensing fusion and intelligent decision-making algorithms. Its core working principle can be summarized as: [Sensor data acquisition] → [Signal processing and feature extraction] → [Spasm recognition and decision-making] → [Region-based airbag control] → [Dynamic pressure recovery] → [Internet of Things data synchronization], forming a complete closed-loop control link of perception - analysis - execution - recovery - feedback.

[0168] Comparison table of medical device performance

[0169] Index Traditional device The present invention Incidence rate of pressure ulcers 15％ 0％ Growth of gastrocnemius cross-sectional area 8.3％ 17.5％ Remote control response delay Unavailable ≤ 5 seconds

[0170] Example 1: Rehabilitation mode for orthopedic postoperative patients

[0171] Please refer to Figure 9 , 1. Parameter setting:

[0172] - Pressure gradient: 18 mmHg at the heel / 40 mmHg at the forefoot / 28 mmHg at the arch;

[0173] - Alternating frequency: 12 times / minute (dorsiflexion for 1.5 seconds → plantar flexion for 3.5 seconds).

[0174] Example 2: Protection mode for hemiplegic patients

[0175] 1. Spasm linkage mechanism:

[0176] - When the detected EMG frequency > 35 Hz, the forefoot pressure changes from 40 mmHg to 20 mmHg;

[0177] - Start plantar vibration feedback (frequency 8 Hz, amplitude 0.3 mm).

[0178] 2. Pressure recovery rule:

[0179] P(t) = P max ×(1 - e -t / τ ) τ = 120 seconds.

[0180] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. Intelligent ankle pump exercise training device, characterized in that, The device comprises a foot support plate (6), wherein the foot support plate (6) is arranged on a bearing shaft (10), wherein the bearing shaft (10) is rotatably arranged on one side of the device housing (1), wherein a sole airbag structure is arranged on one side surface of the foot support plate (6), and a detachable instep fixing portion (11) is also arranged on one side surface of the foot support plate (6), wherein a instep airbag (2) is arranged in the instep fixing portion (11); The plantar airbag structure is composed of a forefoot area airbag (5), an arch area airbag (7) and a heel area airbag (9); the forefoot area airbag (5), the arch area airbag (7), the heel area airbag (9) and the dorsal airbag (2) are all connected to the air pump in the external housing through an air supply pipe; a solenoid valve group is arranged on the air supply pipe, and the solenoid valve group is connected to the main control module; Pressure sensors are also provided on the forefoot area airbag (5), the arch area airbag (7), the heel area airbag (9) and the instep airbag (2), and the pressure sensors are connected to the main control module; The main control module uses the pressure collected by the pressure sensor, an anatomically driven pressure gradient model and a pressure gradient control algorithm to accurately identify muscle spasms.

2. The intelligent ankle pump exercise training device according to claim 1, characterized in that: An angle sensor is also arranged on one side surface (1) inside the foot device shell, and the output end of the angle sensor is connected to the input end of the main control module.

3. The intelligent ankle pump exercise training device and multi-modal linkage control system according to claim 1, characterized in that: An LCD touch screen (3) is also provided on the external housing, and the LCD touch screen (3) is connected to the main control module.

4. An intelligent ankle pump exercise training multi-modal linkage control system according to any one of claims 1-3, characterized in that: It includes an Internet of Things module and a multimodal sensor module. The Internet of Things module is connected to the APP in the doctor's mobile phone terminal through Bluetooth 5.2 and Wi-Fi6 dual-mode communication. The material network module is also connected to the main control module. The multimodal sensor module includes an electromyography sensor, and the output end of the electromyography sensor is connected to the input end of the main control module.

5. The intelligent ankle pump exercise training multi-modal linkage control system according to claim 4 is characterized in that: The main control module is also used to process and extract the signals and features collected by the pressure sensor, angle sensor and electromyography sensor.

6. The intelligent ankle pump exercise training multi-modal linkage control system according to claim 5, characterized in that: The signals and features collected by the pressure sensor, angle sensor and electromyography sensor are processed and extracted, specifically: Step 1: The main control module performs bandpass filtering on the signals collected by the pressure sensor, angle sensor and electromyography sensor; Step 2: The main control module classifies the signals collected by the bandpass filtered pressure sensor, angle sensor and electromyographic sensor, and divides the corresponding signals into electromyographic signals, inertial data and pressure signals; Step 3, the main control module processes the electromyographic signal through the wavelet packet decomposition algorithm, processes the inertial data through the quaternion solution algorithm, and processes the pressure signal through the spatial interpolation algorithm; Step 4: extracting the features of the processed electromyographic signals, inertial data and pressure signals; Step 5, data fusion of the extracted features, identification of spasms through an anatomically driven pressure gradient model and a pressure gradient control algorithm, and control of alternate inflation of the forefoot airbag (5), the arch airbag (7), the heel airbag (9) and the dorsum airbag (2).

7. The intelligent ankle pump exercise training multi-modal linkage control system according to claim 6, characterized in that: The pressure gradient model includes a dorsal airbag control model and a plantar airbag control model.

8. The intelligent ankle pump exercise training multi-modal linkage control system according to claim 7, characterized in that: The anatomical pressure gradient equation of the plantar airbag control model is: Where: P heel is the heel area pressure; P forefoot is the pressure in the forefoot area; p arch is the pressure in the arch area; BMI is the body mass index.

9. The intelligent ankle pump exercise training multi-modal linkage control system according to claim 6, characterized in that: The pressure gradient control algorithm formula is as follows: P heel =0.3×P forefoot +5。

Images