Method, medium and equipment for dynamically estimating rotation center position of knee joint

Abstract

The invention discloses a method, medium and equipment for dynamically estimating the rotation center position of a knee joint, and belongs to the technical field of motion analysis. The method comprises the steps that a first inertial sensor and a second inertial sensor are used for collecting acceleration information and angular velocity information when a joint moves, and the first inertial sensor and the second inertial sensor are arranged on the two sides of the joint at a set distance; taking the distance between the first inertial sensor and the second inertial sensor as a constraint condition, bringing the constraint condition into an extended Kalman filtering observation model for state estimation, and correcting an estimation error by using Kalman gain so as to obtain an estimation error covariance matrix of the first inertial sensor and the second inertial sensor; and calculating an adaptive weight coefficient based on the estimation error covariance matrix, and fusing estimation results of the first inertial sensor and the second inertial sensor according to the weight coefficient to obtain an estimated knee joint center position. According to the method, the precision and stability of joint rotation center position estimation are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of kinematic analysis, and more specifically, to a method, medium, and device for dynamically estimating the position of the knee joint rotation center. Background Art

[0002] Soft Tissue Artifact (STA) is the core problem hindering the measurement accuracy of knee joint kinematics. This is mainly because of the non-rigid coupling between skin markers or inertial sensors and bones, resulting in a deviation between the true position and the measured position of bones during dynamic movement. Some studies have found through synchronous X-ray and optical marker comparison that the error caused by STA in the knee flexion angle during rapid movement can reach 8° - 12° (especially during movement outside the sagittal plane), and the error increases non-linearly with the increase of movement speed. Sangeetha et al. found through relevant research that during the movement of the knee joint, due to the influence of STA, there are obvious displacements in different directions of the knee joint center position. During knee flexion, the STA displacement of the lateral femoral condyle can reach 13.2 mm, and the STA displacement of the medial tibial condyle is 8.6 mm. This results in a large error between the estimated knee joint center position based on skin markers and the actual bone center position, with an average error of up to 9.4 mm and a maximum error of even up to 15.6 mm, seriously affecting the accurate estimation of the knee joint center position.

[0003] Existing knee joint kinematic measurement schemes mainly include optical motion capture systems, inertial sensor (IMU) methods, and self-calibration techniques, but they all have significant limitations. Optical motion capture systems estimate joint motion by tracking skin marker points with multiple cameras, but soft tissue sliding errors (such as knee joint angle errors reaching 8° - 12%) and the high cost of equipment make it difficult to achieve real-time monitoring and popular application. Inertial sensor methods rely on accelerometer and gyroscope data to calculate joint angles, but they ignore the dynamic offsets caused by skin sliding, resulting in a significant increase in errors with the increase of movement intensity (such as the adduction angle error exceeding 7% during running). Self-calibration techniques optimize joint parameters through preset actions or algorithms, but dynamic calibration requires strict user cooperation, while static calibration ignores the complexity of dynamic movement and will significantly amplify errors during complex movements (such as the error reaching up to 16% when going up and down stairs). The closest prior art is the IMU-based joint axis optimization method (such as magnetometer-assisted calibration), but it relies on a fixed biomechanical model and cannot adapt to the non-linear errors caused by STA in real time, further verifying the problem that traditional linear models are difficult to handle multi-degree-of-freedom coupling.

[0004] In the prior art, Patent Application 202111203131.6 discloses a method for determining the center of a human joint based on a wearable sensor. In this solution, multiple inertial measurement units (IMUs) are strapped to the joint positions of the human waist, thigh, and calf to collect triaxial angular velocity values and triaxial acceleration values. The collected data is processed using a data fusion algorithm, and the loss function is optimized in combination with the optimal method to accurately determine the position of the human joint center. Although this solution attempts to solve the problems of low-frequency and high-frequency noise encountered in the process of angular velocity integration and acceleration angle solution through complementary filtering technology, this method relies on fixed filtering coefficients and is not applicable to all situations or individual differences, resulting in limited solution accuracy in some cases. In addition, this solution mainly targets conventional movements such as walking, and its applicability to complex movement states such as rapid running and jumping has not been clarified. Patent Application 202010117861.3 proposes a method for collecting human body movements based on inertial sensors without a standard correction posture. This method includes: arranging IMUs and establishing joint physiological kinematic constraints; when the number of IMU sampling points exceeds 100, an optimization program is run according to the constraint equation, and the Gaussian-Newton method is used to estimate the joint axis and joint position vector; using the solved joint axis and joint position vector, two sets of joint angles are respectively obtained by combining acceleration information and angular velocity integration; the weighted average of the two sets of joint angles is obtained through complementary filtering, and finally the joint angle is solved. However, this solution has specific requirements for the user group and cannot accurately obtain the joint angle information of people with inconvenient mobility or irregular limb movements; moreover, the complexity of multi-sensor deployment and algorithms not only increases the manufacturing and maintenance costs of the device, but also raises the requirements for computing resources, restricting the popularization and application of this technology.

[0005] Upon analysis, the prior art mainly has the following defects: 1) Although the optical motion capture system can accurately capture the knee joint angle, it needs to be used in a specific laboratory environment. The equipment is expensive, the installation is complex, and the marker points are easily blocked and interfered, resulting in data loss. This system has high requirements for the surrounding environmental conditions and needs to maintain certain lighting and space conditions, otherwise it will affect the effect of motion capture.

[0006] 2) Although the method based on electromyography data fusion can estimate the knee joint angle in real time, due to the fact that electromyography signals are easily interfered by noise, the feature extraction process is relatively complex, and this method requires wearing multiple electrodes. The wearing process is cumbersome and has a certain impact on the comfort of the user.

[0007] 3) The model-based method requires the establishment of a mathematical model of the knee joint. The process of establishing and solving the model has high requirements for computing resources and is difficult to achieve real-time estimation. In addition, the accuracy of the model depends on the quality of the input data and the accuracy of the model parameters. Incorrect data or inaccurate parameters will lead to deviations in the estimation results. At the same time, due to the limited generalization ability of the model, it is difficult to be widely applied to different individuals and complex motion situations. Summary of the Invention

[0008] The object of the present invention is to overcome the defects of the above-mentioned prior art and provide a method, medium and device for dynamically estimating the position of the knee joint rotation center.

[0009] According to the first aspect of the present invention, there is provided a method for dynamically estimating the position of the knee joint rotation center. The method includes the following steps: Collect acceleration information and angular velocity information during joint movement by using a first inertial sensor and a second inertial sensor, wherein the first inertial sensor and the second inertial sensor are arranged on both sides of the joint at a set distance; Take the distance between the first inertial sensor and the second inertial sensor as a constraint condition and incorporate it into the extended Kalman filter observation model for state estimation, and use the Kalman gain to correct the estimation error, thereby obtaining the estimation error covariance matrix of the first inertial sensor and the second inertial sensor, wherein the state vector of the extended Kalman filter observation model is characterized by the position vector of the joint center, the velocity vector of the joint center, the acceleration vector of the joint center, the angular velocity vector of the joint, and the angular acceleration vector of the joint; Based on the estimation error covariance matrix, calculate the adaptive weight coefficient, and fuse the estimation results of the first inertial sensor and the second inertial sensor according to the weight coefficient to obtain the estimated position of the knee joint center.

[0010] According to the second aspect of the present invention, there is provided a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the provided method for dynamically estimating the position of the knee joint rotation center are implemented.

[0011] According to the third aspect of the present invention, there is provided a computer device, including a memory and a processor. A computer program capable of running on the processor is stored on the memory. When the processor executes the computer program, the steps of the provided method for dynamically estimating the position of the knee joint rotation center are implemented.

[0012] Compared with the prior art, the advantages of the present invention are as follows. In order to solve the problems of unstable and inaccurate measurement caused by soft tissue artifacts and error accumulation of kinematic parameters in the existing knee joint center estimation technology, the present invention can effectively reduce the influence of soft tissue artifacts on the joint center position by improving the extended Kalman filter algorithm and combining fixed geometric constraints and adaptive weight coefficients, thereby improving the stability and accuracy of joint center estimation and the accuracy and stability of knee joint angle estimation. Traditional methods are difficult to calibrate the joint center position in real time in dynamic motion scenarios and have insufficient ability to handle non-linear errors, while the present invention improves the accuracy, robustness and adaptability of knee joint rotation center estimation and knee joint angle estimation, and also has good adaptability to joint center position estimation in non-fixed joints and complex motion scenarios.

[0013] Other features and advantages of the present invention will become clear from the following detailed description of exemplary embodiments of the present invention with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The drawings incorporated in and constituting a part of this specification illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0015] Figure 1 is a flowchart of a method for dynamically estimating the position of the knee joint rotation center according to an embodiment of the present invention; Figure 2 is a schematic diagram of the overall process of a method for dynamically estimating the position of the knee joint rotation center according to an embodiment of the present invention; Figure 3 is a schematic diagram of knee joint angle estimation according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that: unless otherwise specifically stated, the relative arrangements of components and steps, numerical expressions and values set forth in these embodiments do not limit the scope of the present invention.

[0017] The following description of at least one exemplary embodiment is merely illustrative in nature and in no way serves as a limitation on the present invention, its application or use.

[0018] Techniques, methods and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the techniques, methods and devices should be considered as part of the specification.

[0019] In all examples shown and discussed herein, any specific values should be construed as merely exemplary and not as a limitation. Thus, other examples of exemplary embodiments may have different values.

[0020] It should be noted that like reference numerals and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0021] The present invention focuses on the problems of measurement accuracy and stability of knee joint center estimation caused by soft tissue artifacts, and innovatively proposes a dynamic estimation method that combines a dual inertial measurement unit (IMU, or inertial sensor) with an improved extended Kalman filter (EKF) algorithm. By introducing the fixed geometric constraint between the two IMUs and the adaptive weight data fusion technology, the influence of soft tissue artifacts on the position estimation of the knee joint center of rotation (COR) is effectively reduced, and the accuracy and reliability of joint center position estimation in dynamic motion scenarios are significantly improved. As the key reference point for knee joint angle calculation, the accurate estimation of the knee joint center position provides a reliable reference coordinate system for the calculation of knee joint angles, and thus lays a foundation for in-depth research on the kinematic and dynamic characteristics of the knee joint, and has important application value for clinical applications such as the diagnosis of related diseases, rehabilitation monitoring, and motion analysis.

[0022] Combined Figure 1 and Figure 2 As shown, the provided method for dynamically estimating the position of the knee joint center of rotation includes the following steps: Step S110, using two inertial sensors to collect acceleration information and angular velocity information during knee joint movement.

[0023] For example, using a first inertial sensor and a second inertial sensor to collect acceleration information and angular velocity information during knee joint movement, where the first inertial sensor and the second inertial sensor are arranged on both sides of the knee joint at a set distance.

[0024] In one embodiment, two MPU9250 inertial sensors (including accelerometers and gyroscopes) are used, which are respectively installed on both sides of the knee joint, and triaxial acceleration and angular velocity data are synchronously collected at a sampling rate of 100 Hz or above to ensure time consistency. The high-frequency sampling of the MPU9250 (such as 200 Hz) can capture the details of the movement. By installing the inertial sensors on both sides of the knee joint respectively, the acceleration and angular velocity information during knee joint movement can be captured. The installation position of the sensors should be as close to the skin as possible to reduce the influence of soft tissue artifacts, and the two IMUs synchronously collect triaxial acceleration and angular velocity signals at a high sampling rate (usually 100 Hz or above) to ensure the time consistency of the data.

[0025] Step S120, incorporating the distance constraint between the two inertial sensors into the extended Kalman filter observation model for state estimation, and using the Kalman gain to correct the estimation error, thereby obtaining the estimation error covariance matrix of the two inertial sensors.

[0026] In this step, through state modeling and the improved extended Kalman filter (EKF) algorithm, the error covariance matrices of the two inertial sensors are obtained, which generally include: initializing the state vector and covariance matrix, predicting the state at the next moment according to the state transition equation, integrating the geometric constraints between the two IMUs, using the Kalman gain to correct the state estimate, adaptive weight data fusion, and updating the fusion result by correcting the data error, etc.

[0027] For example, the state vector includes the position, velocity, linear acceleration, angular velocity, and angular acceleration of the joint center. The kinematic relationship is described by the state transition equation, assuming that the acceleration and angular acceleration are constant in the short term.

[0028] To enhance the geometric consistency of the position estimate, the geometric constraints are integrated, and the fixed-distance constraint between the two IMUs is incorporated into the EKF observation model.

[0029] Covariance adaptive update adjusts the state covariance matrix through the prediction and update steps, combines the process noise ( ), and the observation noise ( ) to optimize the Kalman gain to correct the estimation error.

[0030] In the following text, the state vector, state transition equation, observation model, integration of geometric constraints, state prediction process, and update process of the improved extended Kalman filter (EKF) observation model will be specifically introduced.

[0031] 1. Define variables First, define the state vector and related variables, where the state vector is expressed as: (1) where, is the position vector of the joint center, representing the position of the joint center in three-dimensional space. is the velocity vector of the joint center, representing the linear velocity of the joint center. is the acceleration vector of the joint center, representing the linear acceleration of the joint center. is the angular velocity vector of the joint, representing the angular velocity of the joint rotation. is the angular acceleration vector of the joint, representing the angular acceleration of the joint rotation, k represents the time step.

[0032] 2. State transition equation The state transition equation is used to describe the change of the state vector from the time step k to k +1, and is expressed as: (2) These state transition equations describe how the state vector changes fromk evolves to the k +1 moment. Specifically, represents the next moment k position vector of the +1 joint center. represents the current moment k position vector of the joint center. represents the current moment k linear velocity vector of the joint center. represents the current moment k linear acceleration vector of the joint center. represents k the time interval between the k moment and the +1 moment. represents the next moment k linear velocity vector of the +1 joint center. represents the next moment k linear acceleration vector of the +1 joint center (assuming it remains unchanged in a short time). represents the next moment k angular velocity vector of the +1 joint. represents the current moment k angular velocity vector of the joint. represents the current moment k angular acceleration vector of the joint. represents the next moment k angular acceleration vector of the +1 joint (assuming it remains unchanged in a short time).

[0033] The functions of the state transition equation include: Position update: The position of the joint center at the next time step is determined by the current position, velocity, and acceleration. Velocity update: The velocity of the joint center is determined by the current acceleration. Acceleration maintenance: Assume the acceleration remains unchanged in a short time. Angular velocity update: The angular velocity of the joint is determined by the current angular acceleration. Angular acceleration maintenance: Assume the angular acceleration remains unchanged in a short time.

[0034] 3. Observation model The linear acceleration of each IMU can be expressed as: (3) where i is used to identify the IMU, for example i = 1 represents the first IMU, i = 2 represents the second IMU. is the linear acceleration of the joint center, is the angular acceleration, is the angular velocity, is the vector from the IMU to the joint center. It is the linear acceleration caused by angular acceleration. When the rotational speed of the joint is changing, the IMU will be affected by this acceleration. It is the centripetal acceleration. This is the acceleration caused by the rotational motion itself, and its direction points to the direction perpendicular to the axis of rotation.

[0035] 4. Geometric Constraint Integration The fixed distance between the two IMUs d can be used as a constraint condition and added to the Extended Kalman Filter (EKF) as follows: (4) It is expressed as: (5) where, represents k the position vector of the first IMU relative to the joint center at time represents k the position vector of the second IMU relative to the joint center at time d is the distance between the two IMUs, and || || represents the Euclidean norm (the magnitude of the vector). Through the fixed distance constraint, the distance between the two IMUs can be ensured to remain unchanged, providing additional geometric constraints.

[0036] 5. Prediction Process The prediction process uses the state transition equation to predict the state at the next time step and updates the uncertainty (covariance) of the state estimate. The prediction process includes state prediction and covariance prediction.

[0037] The state prediction is based on the current state and the state transition equation to predict the state at the next time step, which is expressed as: (6) where, represents the predicted state vector at time k , based on the information up to time k −1, represents the state vector at the previous time step k −1, f is the non-linear state transition function, is the control input at the previous time step k −1.

[0038] The covariance matrix P represents the uncertainty of the state estimate. The formula for predicting the covariance matrix is: (7) where, represents the covariance matrix of the predicted state error, representsk The state error covariance matrix at time - 1 denotes k the state transition matrix at time f with respect to x the Jacobian matrix), denotes k the noise covariance matrix at time is the state transition matrix, which is the Jacobian matrix of the state transition equation with respect to the state vector and is used to linearize the non - linear state transition model. is the process noise covariance matrix, reflecting the impact of process noise on state estimation.

[0039] 6. Update process The update process uses actual observations to correct the predicted state and update the covariance matrix to improve the estimation accuracy.

[0040] 1) Residual calculation The residual is the difference between the observation and the prediction value, used to measure the accuracy of the prediction: (8) where is k the residual at time denotes the actual measurement value (observation) at time k, h (.) denotes the non - linear observation function.

[0041] 2) Covariance update Update the covariance matrix to reflect the new uncertainty: (9) where denotes the covariance matrix of the residual. is k the observation matrix at time h with respect to x the Jacobian matrix), which is the Jacobian matrix of the observation equation with respect to the state vector and is used to linearize the non - linear observation model. is k the observation noise covariance matrix at time

[0042] 3) Kalman gain calculation The Kalman gain determines the degree of correction of the observation value to the state estimation: (10) where is kThe Kalman gain at a moment, represents the inverse of the residual covariance matrix.

[0043] 4) State update Use the Kalman gain and the residual to correct the state estimate: (11) 5) Covariance update Finally, update the covariance matrix to reflect the new estimation uncertainty: (12) where, I represents the identity matrix.

[0044] Through the above steps, the improved EKF can estimate the knee joint center position in real time and accurately, and effectively cope with various challenges in complex motion scenarios.

[0045] Step S130, calculate the adaptive weight coefficient based on the estimation error covariance matrices of the two inertial sensors, and then fuse the estimation results of the two inertial sensors according to the weight coefficient to obtain the estimated knee joint center position.

[0046] To further improve the estimation accuracy and robustness, the weights are dynamically allocated according to the trace (trace value) of the estimation error covariance matrix of the dual IMU, and then the joint center position estimation results of the two are weighted and fused.

[0047] For example, the calculated estimation error covariance matrices of each IMU are respectively expressed as and . Then, the weight coefficient is dynamically adjusted according to the trace of the covariance matrix, which is expressed as: (13) where, represents k the trace of the estimation error covariance matrix of the first IMU at the moment. The trace is the sum of the elements on the main diagonal of the matrix and is a measure of the overall uncertainty, represents k the trace of the estimation error covariance matrix of the second IMU at the moment. represents the weight coefficient at the moment k .

[0048] In one embodiment, the estimation results of the two IMUs are fused by weighted averaging to obtain a more accurate knee joint center position, which is expressed as: (14) where, represents k the more accurate joint center position vector after fusion at the moment, denotes k the vector of the joint center position estimated by the first IMU at time denotes k the vector of the joint center position estimated by the second IMU at time

[0049] Step S140: Apply the estimated knee joint center position to the motion analysis.

[0050] The knee joint center position is a key reference point in knee joint kinematic analysis. Its determination is crucial for the accurate estimation of knee joint angles. In sports biomechanics research, the calculation of knee joint angles depends on the relative motion analysis of the femur and tibia, and the knee joint center position provides an accurate reference coordinate system for this relative motion. As Figure 3 shown, based on the calculation of the angle between these vectors, the vector dot product formula can be used to determine the magnitude and variation law of the knee joint angle. This accurate calculation method is crucial for in-depth study of the kinematic and dynamic characteristics of the knee joint. It not only helps to evaluate the normality and abnormality of lower limb movements, but also provides a reliable quantitative analysis basis for the optimization of sports training, the formulation of rehabilitation treatment plans, and the diagnosis and surgical planning of knee joint diseases. Therefore, it has wide application value in the fields of sports biomechanics and clinical medicine.

[0051] In one embodiment, based on the estimated knee joint center position, the thigh segment (hip-knee) and calf segment (knee-ankle) vectors are defined, and the knee joint angle is calculated by vector dot product.

[0052] For example, the thigh segment vector is defined as: (15) The calf segment vector is defined as: (16) where denotes the position vector of the knee joint center, denotes the position vector of the hip joint center, denotes the position vector of the ankle joint center.

[0053] The knee joint angle is calculated according to the following formula: (17) In summary, the advantages of the present invention are mainly reflected in the following aspects: 1) The present invention innovatively integrates the linear displacement and angular displacement of the knee joint center and their corresponding derivatives (velocity, acceleration, angular velocity, angular acceleration) into a multi-dimensional motion state vector. This multi-motion state fusion modeling method breaks through the limitation of only focusing on single-dimensional motion characteristics and accurately depicts the complex biomechanical characteristics of the knee joint. Furthermore, by establishing a high-precision state transition equation, the dynamic mapping of the knee joint motion state is realized, providing a solid theoretical basis for joint center estimation and angle estimation.

[0054] 2) The observation model constructed by the present invention deeply binds the linear acceleration of each IMU with the kinematic parameters of the knee joint center by introducing kinematic geometric constraint relations. This precise correlation can maximize the extraction of useful information from the IMU measurement data, effectively reducing the observation error even in complex motion scenarios, ensuring the high fidelity of the observation data to the true motion state of the knee joint, and thus enhancing the observability and stability of the entire estimation.

[0055] 3) In the data fusion process, the present invention introduces an adaptive algorithm based on the trace of the covariance matrix to dynamically adjust the fusion weight. This algorithm calculates the estimated error covariance matrix of each IMU measurement data and uses its trace as a key indicator to measure the data reliability, and then realizes the real-time optimal allocation of the fusion weight. This unique covariance adjustment algorithm can ensure that the dual-IMU data is reasonably weighted under different motion states, thereby improving the accuracy and reliability of the fused data.

[0056] In summary, compared with the prior art, the present invention, by fusing the data of dual inertial measurement units (IMUs) and improving the extended Kalman filter (EKF) algorithm, innovatively introduces the fixed geometric constraint between the dual IMUs and the adaptive weight data fusion technology, can effectively reduce the influence of soft tissue artifacts in dynamic motion scenarios, and significantly improve the accuracy and stability of the knee joint rotation center position estimation. The improved EKF algorithm can more comprehensively and accurately capture the complex motion characteristics of the knee joint by defining a multi-dimensional state vector including the knee joint center position, velocity, acceleration, angular velocity, and angular acceleration, and an observation model that closely associates the linear acceleration of each IMU with the kinematic parameters of the joint center. In addition, the adaptive weight data fusion technology is introduced, and the optimal fusion of the estimation results of the two IMUs is realized by dynamically adjusting the weight coefficient, further improving the adaptability to different motion scenarios.

[0057] The present invention can be a system, a method, and / or a computer program product. The computer program product may include a computer-readable storage medium having thereon computer-readable program instructions for causing a processor to implement various aspects of the present invention.

[0058] A computer-readable storage medium can be a tangible device that can hold and store instructions for use by an instruction execution device. A computer-readable storage medium may be, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), a memory stick, a floppy disk, a mechanically encoded device such as a punch card or raised structures in grooves having instructions stored thereon, and any suitable combination of the foregoing. The computer-readable storage medium as used herein is not construed to be an instantaneous signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., an optical pulse through an optical fiber cable), or an electrical signal transmitted through a wire.

[0059] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include a copper transmission cable, an optical fiber transmission, a wireless transmission, a router, a firewall, a switch, a gateway computer, and / or an edge server. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0060] The computer program instructions for performing the operations of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state-setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages such as Smalltalk, C++, Python, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect through the Internet). In some embodiments, by using the state information of the computer-readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field-programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer-readable program instructions to implement various aspects of the present invention.

[0061] Aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0062] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, when executed by the processor of the computer or other programmable data processing apparatus, create a means for implementing the functions / acts specified in one or more blocks of the flowchart illustrations and / or block diagrams. These computer-readable program instructions may also be stored in a computer-readable storage medium, which instructions cause a computer, a programmable data processing apparatus, and / or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises a manufacture, including instructions for implementing various aspects of the functions / acts specified in one or more blocks of the flowchart illustrations and / or block diagrams.

[0063] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device, causing a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process such that the instructions executed on the computer, other programmable data processing apparatus, or other device implement the functions / acts specified in one or more boxes of the flowchart and / or block diagram.

[0064] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, a segment of code, or a portion of an instruction, which contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block of the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or acts, or by a combination of dedicated hardware and computer instructions. As will be apparent to those of ordinary skill in the art, implementations by hardware, by software, and by a combination of software and hardware are equivalent.

[0065] The embodiments of the present invention have been described above. The above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the market, or to enable other ordinary skill in the art to understand the embodiments disclosed herein. The scope of the present invention is defined by the appended claims.

Claims

1. A method for dynamically estimating the position of the knee joint rotation center, characterized in that The method includes the following steps: Collect the acceleration information and angular velocity information during knee joint movement by using a first inertial sensor and a second inertial sensor, wherein the first inertial sensor and the second inertial sensor are arranged on both sides of the knee joint at a set distance; Take the distance between the first inertial sensor and the second inertial sensor as a constraint condition and incorporate it into the extended Kalman filter observation model for state estimation, and use the Kalman gain to correct the estimation error, and then obtain the estimation error covariance matrix of the first inertial sensor and the second inertial sensor, wherein the state vector of the extended Kalman filter observation model is characterized by the position vector of the knee joint center, the velocity vector of the knee joint center, the acceleration vector of the knee joint center, the angular velocity vector of the knee joint, and the angular acceleration vector of the knee joint; Based on the estimation error covariance matrix, calculate the adaptive weight coefficient, and fuse the estimation results of the first inertial sensor and the second inertial sensor according to the weight coefficient to obtain the estimated knee joint center position.

2. The method according to claim 1, characterized in that, The estimated knee joint center position is expressed as: Among them, represents the moment k estimated knee joint center position vector, represents the moment k weight coefficient of, represents the moment k knee joint center position vector estimated by the first inertial sensor, represents the moment k knee joint center position vector estimated by the second inertial sensor.

3. The method according to claim 2, wherein The weight coefficient at time k is calculated according to the following formula: Among them, represents the moment k the trace of the error covariance matrix estimated by the first inertial sensor, represents the moment k the trace of the error covariance matrix estimated by the second inertial sensor.

4. The method according to claim 1, wherein The constraint condition is set as: Among them, represents the moment k the position vector of the first inertial sensor relative to the center of the knee joint, represents the moment k the position vector of the second inertial sensor relative to the center of the knee joint, d is the distance between the first inertial sensor and the second inertial sensor.

5. The method according to claim 1, wherein It further includes: Calculate the knee joint angle based on the estimated knee joint center position , expressed as: where, || || represents the Euclidean norm, is the thigh segment vector, is the calf segment vector, and they are respectively set as: Among them, represents the estimated knee joint center position vector, represents the hip joint center position vector, represents the ankle joint center position vector.

6. The method according to claim 1, characterized in that, Both the first inertial sensor and the second sensor include an accelerometer and a gyroscope, and synchronously collect the three-axis acceleration information and angular velocity information during knee joint movement at a sampling rate of 100 Hz or above.

7. The method according to claim 1, characterized in that The correction of the estimation error by using the Kalman gain is based on the following formula: Among them, is the Kalman gain at time k . represents the state vector at time k predicted based on the state at time k −1. is the residual at time k .

8. The method according to claim 1, wherein For the extended Kalman filter observation model, the linear accelerations of the first inertial sensor and the second inertial sensor are expressed as: Among them, i is used to identify the inertial sensor, i = 1 indicates the first inertial sensor, i = 2 indicates the second inertial sensor, is the linear acceleration of the knee joint center, is the angular acceleration, is the angular velocity, is the inertial sensor i to the vector of the knee joint center.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it realizes the steps of the method according to any one of claims 1 to 8.

10. A computer device, comprising a memory and a processor, wherein a computer program capable of running on the processor is stored on the memory, characterized in that, When the processor executes the computer program, it realizes the steps of the method according to any one of claims 1 to 8.

Images