Gait monitoring method based on contact pressure sensor

Abstract

The invention discloses a gait monitoring method based on contact pressure sensors, which comprises an insole on which a plurality of contact pressure sensors are uniformly distributed, a three-dimensional inertial sensor is arranged at the bottom of the insole, and the contact pressure sensors and the inertial sensor are connected with a system end; comprising the following steps: acquiring data of foot pressure borne by a contact pressure sensor and a three-dimensional inertial sensor from a t0 moment to a t1 moment, preprocessing the data from the t0 moment to the t1 moment, denoising, filtering and calibrating, and marking as data S; the method has the advantages that the monitoring accuracy is improved, the data quality is enhanced, the calculation precision is optimized, personalized suggestions are provided, real-time data transmission and access are achieved, and a more comprehensive, accurate and personalized gait monitoring effect is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of sensor measurement and control, and particularly relates to a gait monitoring method based on a contact pressure sensor. Background Art

[0002] Gait analysis is crucial for evaluating human walking function, diagnosing movement disorders, guiding rehabilitation training, and optimizing sports performance. Traditional gait analysis techniques mostly adopt video monitoring and image processing methods. However, this method is greatly affected by the environment. The target tracking range is restricted by video equipment, and the data volume is huge, resulting in high complexity and long time required for data processing. In addition, although large gait analysis systems have advantages such as high precision and multi-function, they are expensive and restricted by location and space, and it is difficult to popularize and apply them under the current situation of scarce medical resources.

[0003] Traditional gait monitoring methods rely on a single type of sensor, such as a pressure sensor or an inertial sensor. Although each has its own advantages, there are limitations in comprehensiveness and accuracy. Contact pressure sensors can accurately capture the plantar pressure distribution, while inertial sensors are good at measuring kinematic parameters such as acceleration and angular velocity. Therefore, a gait monitoring method combining the two is expected to significantly improve the depth and accuracy of gait analysis. Summary of the Invention

[0004] To solve the above problems, the present invention provides a gait monitoring method based on a contact pressure sensor, including an insole uniformly distributed with a plurality of contact pressure sensors, a three-dimensional inertial sensor is arranged at the bottom of the insole, and the contact pressure sensor and the inertial sensor are connected to the system end; the method includes the following steps: Obtain the data of the foot pressure received by the contact pressure sensor and the three-dimensional inertial sensor from time t0 to time t1, and preprocess the data from time t0 to time t1: denoise, filter, and calibrate, and mark it as data S; In response to the preprocessed data S, analyze the corresponding plantar pressure distribution and gait cycle, and calculate the acceleration, angular velocity, displacement, velocity, and body posture changes during the gait; In response to the acceleration, angular velocity, displacement, velocity, and body posture changes during the gait of data S, calculate the gait parameters through a data fusion algorithm; In response to the gait parameters of data S, provide a display interface and a report of the gait analysis result.

[0005] Preferably, the preprocessing step further includes time synchronization processing to ensure that the data of the contact pressure sensor and the inertial sensor are consistent in time stamps.

[0006] Preferably, the data fusion algorithm adopts Kalman filtering, extended Kalman filtering, and particle filtering techniques to improve the calculation accuracy of gait parameters.

[0007] Preferably, the system terminal internally presets health standards and normal exercise performance indicators, compares them with the gait analysis results, and provides personalized health advice and training guidance.

[0008] Preferably, a wireless communication module is provided at the bottom of the insole, and the gait analysis results are transmitted to the system terminal in real time through the wireless communication module for storage, analysis, and access.

[0009] The advantages of the present invention are as follows: In this solution, by evenly distributing multiple contact pressure sensors on the insole and combining three-dimensional inertial sensors, it is possible to comprehensively and meticulously capture and analyze the plantar pressure distribution during walking or running and various dynamic parameters (acceleration, angular velocity, displacement, velocity, and body posture changes) in the gait cycle. This multi-dimensional data acquisition method greatly improves the accuracy and reliability of gait monitoring.

[0010] In this solution, through preprocessing steps such as denoising, filtering, and calibration, the noise and errors in the original data are effectively reduced, improving the usability and accuracy of the data. In addition, time synchronization processing ensures the consistency of different sensor data in time stamps, providing a reliable basis for subsequent gait analysis.

[0011] This solution adopts advanced data fusion algorithms such as Kalman filtering, extended Kalman filtering, and particle filtering, which can further optimize the calculation process of gait parameters, improve the calculation accuracy, and thus more accurately reflect the gait characteristics and exercise performance of users.

[0012] In this solution, the health standards and normal exercise performance indicators preset by the system terminal are compared with the gait analysis results, which can provide personalized health advice and training guidance for users. This helps users understand their own gait problems and potential health risks and take corresponding measures for improvement and enhancement.

[0013] The wireless communication module provided at the bottom of the insole in this solution can transmit the gait analysis results to the system terminal in real time for storage, analysis, and access. This real-time nature not only improves the monitoring efficiency but also enables users to view their gait analysis reports at any time and make corresponding adjustments and improvements based on the analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic flowchart of the method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0015] 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.

[0016] A gait monitoring method based on a contact pressure sensor includes the following steps; S1. Hardware preparation: Design and manufacture an insole. Multiple contact pressure sensors are evenly distributed at the bottom of the insole to capture the pressure distribution when the foot contacts the ground. A three-dimensional inertial sensor is also integrated at the bottom of the insole to measure motion parameters such as acceleration and angular velocity during gait, ensuring the comfort and durability of the insole while ensuring that the sensors can work accurately and stably. At the same time, system-side devices (smartphones, tablets, and professional analysis devices) are equipped, which need to have the capabilities of communicating with the insole sensors, data processing, and result display. The core is to ensure that both the comfort and durability requirements for daily wear are met, and key data related to gait can be accurately captured. The contact pressure sensors are evenly distributed at the bottom of the insole, which can finely sense the pressure distribution when different areas of the foot contact the ground, which is crucial for understanding the mechanical characteristics in gait. The integration of the three-dimensional inertial sensor is used to measure the acceleration and angular velocity during gait, and these data help analyze the dynamic characteristics of gait and the changes in body posture.

[0017] S2. Data acquisition: The user wears the insole with sensors to perform activities such as walking or running. The sensors capture the foot pressure data and three-dimensional inertial data from time t0 to time t1 (such as a complete gait cycle) in real time, and the data is transmitted to the system-side device by wired or wireless means. The data acquisition stage relies on the sensors built into the insole to capture gait information in real time. When the user performs activities such as walking or running, the contact pressure sensors and the three-dimensional inertial sensors start to work and continuously record data from time t0 to time t1 (such as a complete gait cycle). These data include the pressure distribution of each area of the foot, acceleration, and angular velocity, etc., which together form the basis for gait analysis. The data transmission method can be wired or wireless, and wireless transmission (including Bluetooth and Wi-Fi) is more common in modern smart devices due to its convenience. Ensuring the real-time and integrity of data transmission is the key in this stage for subsequent data processing and analysis.

[0018] S3. Data Preprocessing: Denoise the received data to eliminate the interference of environmental noise and sensor self-noise on the data. Apply a filtering algorithm (such as low-pass filtering) to smooth the data, improve the accuracy of the data, calibrate the data to ensure the consistency and accuracy of data between different sensors. Perform time synchronization processing. Adopt the IEEE 1588 PTP protocol to achieve a synchronization accuracy at the microsecond level. After filtering, the signal-to-noise ratio is improved: SNR ≥ 40dB; full-scale calibration accuracy: ≤0.5%FS; time synchronization error: ≤100μs; cross-sensor data consistency: correlation coefficient ≥ 0.98; ensure that the data of the contact pressure sensor and the inertial sensor are exactly matched in terms of timestamps.

[0019] S4. Gait Analysis: Based on the preprocessed data S, analyze the plantar pressure distribution, identify different phases in the gait cycle (such as touchdown, stance, push-off, etc.), calculate the acceleration, angular velocity during gait, and parameters such as displacement and velocity obtained by integration. Combine the data of the three-dimensional inertial sensor to analyze the changes in body posture. Regarding the plantar pressure distribution, the trajectory of the center of pressure (COP) needs to be considered, including changes on the x and y axes. There is also zonal statistics, dividing the sole of the foot into the forefoot, midfoot, and hindfoot, and calculating the pressure peak and time proportion of each area. The identification status is specified as the touchdown, stance, and push-off phases through the gait phase. The gait cycle is divided into the stance phase and the swing phase, where the stance phase includes initial touchdown, loading response, mid-stance, terminal stance, and push-off. It is necessary to determine the start and end time points of each phase. The threshold method is used to mark touchdown when the pressure exceeds a certain value and mark lift-off when it is below. Dynamic time warping (DTW) is used to match with the standard gait template, and automatic classification of phases can be based on machine learning models: SVM and LSTM. For parameter calculation, quaternions or rotation matrices are used to convert the acceleration to the global coordinate system, and then the gravity component is subtracted and integrated.

[0020] In the visualization and quantification index part, specific charts and parameters need to be given, such as the pressure distribution heat map, COP trajectory map, time proportion of each phase, step frequency, step length, etc.

[0021] For the calculation of the gait trajectory of the center of pressure COP, the calculation formula for the dynamic coordinates is: COP_x = \frac{\sum_{i=1}^n (p_i \cdot x_i)}{\sum_{i=1}^n p_i}, COP_y= \frac{\sum_{i=1}^n (p_i \cdot y_i)}{\sum_{i=1}^n p_i}; Among them, \(p_i\) is the pressure value of the \(i\)-th sensing unit, and \((x_i, y_i)\) is the sensor position coordinate. The corresponding trajectory characteristic parameters are: swing amplitude: \(max(COP_x)-min(COP_x)\); pressure asymmetry: DTW distance of the COP trajectories of the left and right feet.

[0022] S5. Data fusion and gait parameter calculation: Apply Kalman filter, extended Kalman filter, and particle filter data fusion algorithms, and combine data on plantar pressure distribution, position, speed, acceleration, and angle to accurately calculate gait parameters, specifically the user's step length, step frequency, and body tilt angle.

[0023] (1) First, construct the state vector. Let the state vector be: , Among them, is the three-dimensional position, is the speed, is the Euler angle, is the angular velocity, and \(T\) is a constant.

[0024] (2) Secondly, perform the Kalman filter (KF) algorithm, which is used to estimate the state of the dynamic system. Through two steps of prediction and update, combine the system model and observation data. For gait analysis, the state variables include position, speed, acceleration, and angle: Based on the setting of the state vector, establish the system dynamic model: , Among them, the acceleration is obtained by compensating the IMU measurement value for gravity: , \(R\) is the rotation matrix, is the gravity vector.

[0025] Establish the system observation data: Observation data of plantar pressure, with constraints on the position of the touchdown point: ; Observation data of IMU attitude: .

[0026] Extended Kalman filter (EKF) for handling nonlinearity: Attitude quaternion model: When using the quaternion to represent the attitude, the system becomes nonlinear: ; Jacobian matrix calculation: Take the partial derivatives of the state transition function and the observation function : , When the observation includes the center of plantar pressure, it is necessary to include the partial derivative of .

[0027] Particle Filter (PF) realizes multi-modal estimation: Particle initialization: Generate N particles , with the initial weight .

[0028] Importance sampling: Propagate particles according to the motion model: ; Weight update: , thus, the likelihood function of plantar pressure observation is set as: ; Resampling: When the number of effective particles , resample particles according to the weights.

[0029] Gait parameter calculation: Stride length calculation: During the plantar contact phase (when the pressure > threshold), reset the speed to avoid integral drift; the specific single-step stride length is: ; Stride frequency calculation: Detect the time interval between consecutive contact events : ; Body tilt angle estimation: Static tilt: Measure the gravity component through the accelerometer: ; Dynamic compensation: Use EKF to fuse gyroscope data to eliminate the influence of motion acceleration: , is the gyroscope zero bias, is the acceleration compensation term.

[0030] S6. Result display and analysis: Provide a display interface for gait analysis results on the system-side device, intuitively presenting information such as gait parameters, plantar pressure distribution maps, gait cycle diagrams, etc., and generating a gait analysis report, including specific values of gait parameters, comparison results with health standards, personalized health suggestions, and training guidance.

[0031] S7. System-side interaction: Through the wireless communication module, the gait analysis results are transmitted to the system-side device in real time for storage, analysis, and access. Users can view the gait analysis results and reports at any time through mobile devices such as mobile phones, and adjust their exercise methods in a timely manner to improve the training effect. The system-side device can also dynamically adjust health suggestions and training guidance according to the real-time transmitted data to provide more personalized services.

[0032] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A gait monitoring method based on a contact pressure sensor, including an insole evenly distributed with a plurality of contact pressure sensors, a three-dimensional inertial sensor is arranged at the bottom of the insole, and the contact pressure sensor and the inertial sensor are connected to the system end; characterized in that: It includes the following steps: Obtain the data of the foot pressure received by the contact pressure sensor and the three-dimensional inertial sensor from time t0 to time t1, and preprocess the data from time t0 to time t1: denoise, filter, and calibrate, and mark it as data S; In response to the preprocessed data S, analyze the corresponding plantar pressure distribution and gait cycle, and calculate the acceleration, angular velocity, displacement, speed, and body posture changes during the gait; In response to the acceleration, angular velocity, displacement, speed, and body posture changes during the gait of data S, calculate the gait parameters through a data fusion algorithm; In response to the gait parameters of data S, provide a display interface and report of the gait analysis results.

2. The gait monitoring method based on a contact pressure sensor according to claim 1, characterized in that: The preprocessing step further includes time synchronization processing to ensure that the data of the contact pressure sensor and the inertial sensor are consistent in time stamps.

3. The gait monitoring method based on a contact pressure sensor according to claim 1, wherein: The data fusion algorithm adopts Kalman filtering, extended Kalman filtering, and particle filtering techniques to improve the calculation accuracy of gait parameters.

4. The gait monitoring method based on a contact pressure sensor according to claim 1, characterized in that: The system end internally presets health standards and normal exercise performance indicators, compares them with the gait analysis results, and provides personalized health advice and training guidance.

5. The gait monitoring method based on a contact pressure sensor according to claim 4, characterized in that: A wireless communication module is provided at the bottom of the insole, and the gait analysis results are transmitted to the system end in real time through the wireless communication module for storage, analysis, and access.

Images