CN-122004841-A - Data-driven intelligent gait health detection and localization system solution

CN122004841ACN 122004841 ACN122004841 ACN 122004841ACN-122004841-A

Abstract

The invention discloses a solution of a data-driven intelligent gait health detection and positioning system, and belongs to the technical field of wearable rehabilitation equipment. The system comprises a three-point flexible film pressure sensor array, an IMU inertial measurement unit, a GPS positioning module, a master-slave distributed control platform and a communication module, wherein sensor data are acquired by a slave control board and then transmitted to the master control board, the master control board realizes gait pattern recognition, fall risk early warning and user positioning based on a multi-sensor fusion and machine learning algorithm, a layered processing strategy is adopted, a primary model is deployed on the master control board to realize real-time state recognition and low power consumption control, and a secondary model is deployed on a cloud to realize gait type subdivision and health assessment. Through multi-source data fusion and edge cloud cooperative calculation, gait recognition accuracy and timeliness of fall early warning are improved, wearing comfort and system endurance are considered, and the method is suitable for daily health monitoring of the old, the recovered patient and other groups.

Inventors

PAN XIUJIANG
JIANG YING
LIAO ZHENG
HU WEI
LUO SHENGLI
SHU XIAOLONG
YU HONGLIU
MENG QIAOLING

Assignees

奥思特甲医疗科技(上海)有限公司
上海理工大学

Dates

Publication Date: 20260512
Application Date: 20260203

Claims (10)

1. Data-driven intelligent gait health monitoring and positioning system, characterized by comprising: The sensor unit comprises a three-point flexible film pressure sensor array, an IMU inertial measurement unit and a GPS positioning module which are arranged on the insole; the slave control boards are respectively arranged in the left foot insole and the right foot insole and are used for collecting sensor data of corresponding feet; The main control board is in communication connection with the slave control board and is used for receiving and fusing the multi-sensor data and executing gait analysis and state identification; the communication module is used for carrying out data interaction with an external terminal or a cloud platform; the main control board is internally provided with a primary identification model for judging the motion state of a user in real time and controlling the sampling frequency and the data transmission strategy.
2. The data-driven intelligent gait health detection and localization system of claim 1, wherein the three-point flexible film pressure sensor array is disposed on the medial forefoot, lateral forefoot and heel areas of the insole for collecting plantar pressure distribution timing data during a gait cycle.
3. The data-driven intelligent gait health detection and positioning system according to claim 1, wherein the slave control board adopts an STM32G4 series microcontroller, the master control board adopts an STM32H7 series microcontroller, and the data uploading and the instruction issuing are realized through serial communication.
4. The data-driven intelligent gait health detection and localization system of claim 1, wherein the main control board is further configured to perform the steps of: S1, preprocessing received sensor data and segmenting gait cycles; S2, extracting time domain and frequency domain features, and fusing pressure, gesture and acceleration multisource information; S3, judging whether the user is in a static state, a walking state, a running state or a suspected falling state based on the first-level recognition model; S4, entering a low power consumption mode in a static state, increasing the sampling rate in a motion or suspected falling state, and triggering data uploading.
5. The data-driven intelligent gait health detection and positioning system according to claim 1, further comprising a cloud server for running a secondary identification model, receiving multi-step continuous data uploaded by a main control board, performing gait pattern refinement and health assessment, and generating fall early warning information.
6. The data-driven intelligent gait health detection and localization system of claim 5, wherein the secondary identification model is constructed by: S1, collecting plantar pressure, IMU and positioning data of a plurality of subjects in different sports scenes as a training set; s2, marking the data, including gait type labels and falling event labels; And S3, performing model training by using a machine learning algorithm, and optimizing the recognition accuracy and recall rate on the verification set.
7. The data-driven intelligent gait health detection and positioning system according to claim 1, wherein the system is integrated at the tongue part, is detachably packaged, and the sensor unit is connected with the main control board by a flexible circuit, so that wearing comfort and system reliability are ensured.
8. The data-driven intelligent gait health detection and positioning system according to claim 1, wherein the communication module supports multiple communication modes of 4G/5G, wiFi or bluetooth for transmitting the early warning information, the positioning coordinates and the health data to the bound mobile terminal or the monitoring platform.
9. The data-driven intelligent gait health detection and localization system of claim 1, further comprising a fall early warning rule base that enables early recognition and early warning of fall precursors based on historical data and real-time feature matching.
10. The data-driven intelligent gait health detection and positioning system of claim 1, wherein the IMU inertial measurement unit comprises a tri-axis accelerometer, a tri-axis gyroscope and a tri-axis magnetometer for acquiring the attitude angle, angular rate and acceleration information of the user.

Description

Data-driven intelligent gait health detection and localization system solution Technical Field The invention belongs to the technical field of wearable rehabilitation equipment and health monitoring, and particularly relates to a solution of a data-driven intelligent gait health detection and positioning system. Background With the acceleration of the progress of social aging, the elderly fall has become an important public health problem. The falling not only causes the body injury, but also can cause the consequences of long-term bedridden, psychological fear and the like, and brings heavy burden to families and society. Health monitoring equipment such as intelligent bracelet, fall detector etc. in the present market relies on single acceleration sensor to carry out action recognition more, and recognition accuracy is limited under the complex environment (going up and down stairs, ramp walking), and can't acquire key gait characteristics such as plantar pressure distribution, is difficult to realize the risk early warning before falling. In addition, the existing gait analysis system such as optical motion capturing and pressure measuring tables is high in accuracy, but the device is large in size, high in cost and not applicable to daily continuous monitoring, and cannot be used outdoors. Disclosure of Invention The invention aims to overcome the defects of the prior art and provides a data-driven intelligent gait health detection and positioning system and method. The system combines an innovative hardware architecture and a software algorithm to realize real-time acquisition, intelligent edge processing and cloud deep analysis of multi-source gait data, so that the aims of high-precision gait recognition, early fall risk early warning and accurate user positioning are fulfilled. In order to achieve the above purpose, the invention provides the following technical scheme that the data-driven intelligent gait health monitoring and positioning system comprises: The sensor unit comprises a three-point flexible film pressure sensor array, an IMU inertial measurement unit and a GPS positioning module which are arranged on the insole; the slave control boards are respectively arranged in the left foot insole and the right foot insole and are used for collecting sensor data of corresponding feet; The main control board is in communication connection with the slave control board and is used for receiving and fusing the multi-sensor data and executing gait analysis and state identification; the communication module is used for carrying out data interaction with an external terminal or a cloud platform; the main control board is internally provided with a primary identification model for judging the motion state of a user in real time and controlling the sampling frequency and the data transmission strategy. Further, the three-point flexible film pressure sensor array is disposed on the medial forefoot, lateral forefoot and heel areas of the insole for acquiring plantar pressure distribution timing data during a gait cycle. Further, the slave control board adopts STM32G4 series microcontrollers, the master control board adopts STM32H7 series microcontrollers, and the slave control board and the master control board realize data uploading and instruction issuing through serial port communication. Further, the main control board is further configured to perform the following steps: S1, preprocessing received sensor data and segmenting gait cycles; S2, extracting time domain and frequency domain features, and fusing pressure, gesture and acceleration multisource information; S3, judging whether the user is in a static state, a walking state, a running state or a suspected falling state based on the first-level recognition model; S4, entering a low power consumption mode in a static state, increasing the sampling rate in a motion or suspected falling state, and triggering data uploading. Further, the intelligent monitoring system further comprises a cloud server, wherein the cloud server is used for running the secondary identification model, receiving multi-step continuous data uploaded by the main control board, performing gait pattern fineness and health assessment, and generating fall early warning information. Further, the secondary identification model is constructed by: S1, collecting plantar pressure, IMU and positioning data of a plurality of subjects in different sports scenes as a training set; s2, marking the data, including gait type labels and falling event labels; And S3, performing model training by using a machine learning algorithm, and optimizing the recognition accuracy and recall rate on the verification set. Further, the system is integrated at the tongue part, adopts detachable encapsulation, adopts flexible circuit to connect between sensor unit and the main control board, ensures wearing travelling comfort and system reliability. Further, the communication module supports multiple communica