A decoupling measurement method for plantar pressure and shear force signals in deep learning

Abstract

The present invention discloses a novel decoupling measurement method for plantar pressure and shear force signals in deep learning, which relates to the technical field of plantar force and gait measurement, and includes the following steps: Step S1, fabricate a sensor structure; Step S2, integrate multi-dimensional perception; Step S3, perform pressure and shear force signal analysis. By adopting the above-mentioned novel plantar pressure and shear force sensor and its signal decoupling method in deep learning, the present invention realizes high-precision measurement of plantar pressure and shear force, significantly improves the diversity and accuracy of data acquisition, and provides reliable data support for intelligent wearable devices.

Description

Technical Field

[0001] The present invention relates to the technical field of plantar force and gait measurement, and in particular to a novel decoupling measurement method for plantar pressure and shear force signals based on deep learning. Background Art

[0002] Gait is one of the key patterns in human daily activities, containing rich physiological information of human movement. Plantar force is a treasure trove of information waiting to be explored. However, gait is a complex behavioral characteristic, involving central commands, body balance and coordination control, the coordinated movement of various joints and muscle groups in the lower limbs, and the postures of the upper limbs and torso. In addition, the combined action of the bony prominences and various soft tissues on the sole of the foot makes it often in a complex stress state with coupled pressure and shear force. How to effectively decouple pressure and shear force in this complex stress state and measure the uncertainty of the shear force direction has become a major challenge.

[0003] The plantar stress state is complex and changes dynamically with time and space. The coupled action of pressure and shear force makes accurate measurement extremely difficult. The following problems exist in the prior art:

[0004] 1) A preliminary consensus on plantar pressure sensors has been formed, but complex stress conditions are not involved.

[0005] 2) There has been preliminary research on the synchronous measurement of foot pressure and shear force, but there are deficiencies in the measurement, accuracy, and applicability of the shear force direction.

[0006] 3) The measurement of plantar force is combined with artificial intelligence, but there are still obvious deficiencies in the accurate acquisition of original plantar force data and feature extraction under the cross-action of gait and various diseases.

[0007] 4) The performance differences of the arch, bony prominences, and plantar tissues of the foot lead to a complex plantar stress state

[0008] In summary, due to the variability of human activities, the complexity of the plantar bony prominence structure, and the combined action of various soft tissue structures, the sole of the foot is often in a complex stress state with coupled pressure and shear force. At present, the ability to decouple and measure complex plantar forces is insufficient, resulting in the existing force sensors being unable to meet the requirements of intelligent wearable devices for accurately monitoring plantar forces. Summary of the Invention

[0009] The purpose of the present invention is to provide a novel decoupling measurement method for plantar pressure and shear force signals based on deep learning, to solve the problems proposed in the above background art, to be able to perform data analysis and prediction in real time and efficiently, to provide accurate and reliable results for users, to improve the intelligent level of the system, and to provide a solid foundation for decision-making support and optimization.

[0010] To achieve the above object, the present invention provides a novel decoupled measurement method for plantar pressure and shear force signals in deep learning, comprising the following steps:

[0011] Step S1, fabricate the sensor structure;

[0012] Step S2, integrate multi-dimensional perception;

[0013] Step S3, perform pressure and shear force signal analysis.

[0014] Preferably, the specific steps of the said Step S1 are as follows:

[0015] Step S11, design the pressure and shear force sensor structure for multi-dimensional perception;

[0016] Step S12, assemble and fabricate the sensor according to the structure design.

[0017] Preferably, in the said Step S2, data is collected through a data acquisition unit, and the collected data includes plantar pressure and shear force data during the user's daily activities. After the collection is completed, the sensor structure transmits the real-time data to the data processing unit via low-power Bluetooth.

[0018] Preferably, the specific steps of the analysis and processing in the said Step S3 are as follows:

[0019] Step S31: Perform intelligent feature extraction processing on multi-dimensional force signals: Adopt an intelligent feature extraction algorithm to identify key multi-dimensional mechanical features from the processed signals. The multi-dimensional mechanical features include pressure magnitude, shear force magnitude, and shear force direction;

[0020] Step S32: Perform high-precision data cleaning and standardization processing: Adopt a refined data cleaning and standardization processing method to ensure the integrity and consistency of all data.

[0021] Therefore, the present invention adopts the above novel decoupled measurement method for plantar pressure and shear force signals in deep learning, realizing high-precision measurement of plantar pressure and shear force, significantly improving the diversity and accuracy of data collection, and providing reliable data support for intelligent wearable devices.

[0022] Next, through the accompanying drawings and embodiments, the technical solutions of the present invention will be further described in detail. Description of the Drawings

[0023] Figure 1 It is a schematic structural diagram of the sensor structure of a novel decoupled measurement method for plantar pressure and shear force signals in deep learning according to the present invention;

[0024] Figure 2Schematic diagram of the measurement principle of the sensor structure for a novel decoupling measurement method of plantar pressure and shear force signals in deep learning according to the present invention;

[0025] Figure 3 Flowchart of a novel decoupling measurement method of plantar pressure and shear force signals in deep learning according to the present invention. Detailed implementation manners

[0026] The technical solutions of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0027] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention belongs. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before this term cover the elements or objects listed after this term and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right" are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0028] Please refer to Figures 1-3 , the present invention provides a novel decoupling measurement method of plantar pressure and shear force signals in deep learning, which can measure the shear force in the 360° direction, including the following steps:

[0029] Step S1, fabricate a sensor structure that can simultaneously sense pressure and shear force to improve the diversity and accuracy of data acquisition. Embed a plurality of high-precision pressure-shear force sensor structures in the insole, and accurately distribute the sensors at the key force-bearing parts of the sole (such as the forefoot, midfoot, and hindfoot) according to the research results of human movement biomechanics.

[0030] Step S11, design a pressure and shear force sensor structure for multi-dimensional perception, and the design drawing is as Figure 1 shown.

[0031] Step S12, assemble and fabricate the sensor according to the structural design. The detailed steps are as follows:

[0032] Refer to human movement biomechanics to assemble and fabricate the pressure sensing unit in the sensor structure. As Figure 1As shown, the sensor structure includes a semi-circular contact block disposed below the contact surface. One end of an elastic support column and multiple pressure sensing units is connected to the lower part of the semi-circular contact block, and the other ends of the multiple pressure sensing units and the elastic support column are connected to the substrate. The elastic support column has excellent sensitivity and durability, and can provide stable support and rebound effects during multiple repeated loading and unloading processes. The pressure sensing unit is made of a polymer flexible material, with excellent sensitivity and durability, and can provide stable and accurate data acquisition under various foot force conditions. The changes and differences in the sensing signals can effectively reflect different compressive and shear force states. The flexible pressure sensing unit significantly reduces its interference with foot movement and effectively improves the wearing comfort. The pressure sensing units are arranged on the circumferential side ribs, and the accuracy of sensing the shear force direction is significantly enhanced as the number of units increases.

[0033] Use three-dimensional foot scanning technology to obtain accurate foot morphology data. Based on the foot morphology data, use computer-aided design (CAD) software to design the geometric shapes and sizes of the pressure sensing units and shear force sensing units. Consider mechanical properties and ergonomics during the design to ensure that the sensor units fit closely to the curved surface of the sole. Design the layout of the sensor units to ensure that the main force-bearing areas such as the forefoot, heel, and arch of the sole are covered by sensor units.

[0034] Step S2: Integrate multi-dimensional sensing. Data is collected through the data collection unit, and the collected data includes the plantar pressure and shear force data during the user's daily activities. After the collection is completed, the sensor structure transmits the real-time data to the data processing unit via low-power Bluetooth.

[0035] During the integration process of multi-dimensional sensing, the aim is to optimize the layout and performance of the sensor structure to achieve accurate measurement of plantar pressure and shear force. The detailed steps are as follows:

[0036] First, perform sensor positioning: When embedding the sensor structure in the insole, scientifically select the layout positions of the sensor units according to the mechanical distribution characteristics of the sole. Through the biomechanical analysis of the sole, it is determined that the forefoot, heel, and arch areas are the key areas where pressure and shear forces act intensively. Arrange the pressure sensing units at the forefoot and heel areas, which bear relatively large vertical pressures during walking and standing; while the shear force sensing units are preferentially arranged at the arch and the root of the toes to capture the shear forces generated by the sole under different motion states. This positioning scheme can ensure that the sensor structure covers all the main force-bearing areas of the sole, comprehensively capture the dynamic force conditions, and improve the accuracy of data collection.

[0037] Secondly, ensure tight contact: To ensure the close contact between the sensor structure and the sole of the foot, and to ensure that the sensor can closely adhere to the sole surface under various foot movement conditions, thereby ensuring the accuracy of data collection. The fit between the sensor and the sole of the foot is improved, effectively preventing the sensor from shifting in position during dynamic movements and ensuring continuous high-quality data collection.

[0038] Finally, signal transmission and power management: The sensor structure adopts a low-power design to extend the service life of the device. Through a low-power wireless communication method, the collected pressure and shear force data are transmitted in real time to an external device for processing. This design not only reduces the consumption of the sensor structure battery, improves the overall efficiency of the sensor system, but also ensures the immediacy and stability of data transmission, enhancing the practicality and reliability of the system.

[0039] Step S3: Parse the pressure and shear force signals, and the specific steps of the parsing process are as follows:

[0040] Step S31: Perform intelligent feature extraction processing on multi-dimensional force signals. First, perform denoising processing: Use the adaptive Kalman filter method to perform high-precision denoising on the real-time collected pressure and shear force signals. By using this method, the filtering parameters can be dynamically adjusted according to the real-time collected pressure and shear force signals, thereby effectively removing the high-frequency noise in the signals and retaining the true mechanical characteristics. Different from traditional static filters, the adaptive Kalman filter can intelligently identify and filter different types of noise, such as environmental noise and sensor noise, ensuring the accuracy and stability of the output signal. Secondly, use an intelligent feature extraction algorithm to identify the key multi-dimensional mechanical characteristics from the processed signals. The multi-dimensional mechanical characteristics include the spatio-temporal distribution characteristics such as the magnitude of pressure, the magnitude of shear force, and the direction of shear force. The intelligent feature extraction algorithm can identify the key parameters from the denoised multi-dimensional force signals, including the maximum pressure point, the direction of shear force, etc., which can reflect the real situation of the sole of the foot under force. The intelligent feature extraction algorithm analyzes the multi-dimensional signals collected by the sensor, extracts the most representative features, and uses these features for further analysis and application. It not only improves the speed and accuracy of feature extraction, but also shows good robustness in various complex sole of the foot force environments.

[0041] Step S32: Perform high-precision data cleaning and standardization processing: Use a refined data cleaning and standardization processing method to ensure the integrity and consistency of all data. The data cleaning steps include handling missing values and outliers to prevent data noise from affecting the analysis results. The standardization step ensures that data from different sources and natures can be compared and analyzed under the same standard by unifying the dimension and scale of the data. Through this processing, high-quality data input is guaranteed, providing a reliable basis for subsequent analysis and model training.

[0042] Therefore, the present invention adopts the above-mentioned novel decoupling measurement method for plantar pressure and shear force signals in deep learning, realizes high-precision measurement of plantar pressure and shear force, significantly improves the diversity and accuracy of data acquisition, and provides reliable data support for intelligent wearable devices.

[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A deep learning method for measuring plantar pressure and shear force signal decoupling, characterized in that: The following steps are involved: Step S1, making a sensor structure; The sensor structure includes a semicircular contact block arranged below the contact surface, an elastic support column and one end of a plurality of pressure sensing units are connected below the semicircular contact block, and the other end of the plurality of pressure sensing units and the elastic support column are connected to the substrate; Step S2, integrating multi-dimensional perception; Step S3, analyzing pressure and shear force signals; The specific steps of step S1 are as follows: Step S11, designing a multi-dimensional pressure and shear force sensor structure; Step S12: assembling and manufacturing the sensor according to the structural design; In step S2, data is collected by a data collection unit, and the collected data includes plantar pressure and shear force data in the user's daily activities. After the collection is completed, the sensor structure transmits the real-time data to the data processing unit via low-power Bluetooth; The specific steps of the analysis process in step S3 are as follows: Step S31: performing intelligent feature extraction processing of multi-dimensional force signals: using an intelligent feature extraction algorithm to identify key multi-dimensional mechanical features from the processed signals, the multi-dimensional mechanical features including pressure magnitude, shear force magnitude and shear force direction; Step S32: Perform high-precision data cleaning and standardization processing: Use refined data cleaning and standardization processing methods to ensure the integrity and consistency of all data.

Images