CN-121987185-A - Human body lower limb joint angle calculation method and device

CN121987185ACN 121987185 ACN121987185 ACN 121987185ACN-121987185-A

Abstract

The application provides a human body lower limb joint angle calculation method and device, wherein the method comprises the steps of collecting surface electromyographic signals of a last moment in a walking state of a subject by using a surface electromyographic sensor, respectively calculating lower limb joint angle estimated values of the last moment, obtaining lower limb joint angle measured values of a current moment in the walking state of the subject by using an IMU, establishing a state equation by using the lower limb joint angle estimated values, establishing a fusion model based on extended Kalman filtering by using the lower limb joint angle measured values as observation equations, and obtaining an optimal estimated value of the lower limb joint angle of the current moment by using the fusion model based on the extended Kalman filtering on the lower limb joint angle estimated values of the last moment and the lower limb joint angle measured values of the current moment. According to the application, the measurement result of the IMU and the estimation result of the surface electromyographic signals are fused to obtain the stable and advanced joint angle, so that the problem of hysteresis of the measurement value of the joint angle of the lower limb of the IMU is solved.

Inventors

WANG XINGJIAN
TIAN XINYU
ZHANG YUWEI
ZHANG YIXIN
WANG SHAOPING
ZHOU XINGFEI

Assignees

天目山实验室

Dates

Publication Date: 20260508
Application Date: 20251222

Claims (10)

1. The method for calculating the angle of the joints of the lower limbs of the human body is characterized by comprising the following steps of: collecting surface electromyographic signals of the last moment in the walking state of the subject by using a surface electromyographic sensor, and respectively calculating the lower limb joint angle estimated value of the last moment, wherein the lower limb joint angle estimated value comprises a hip joint angle estimated value, a knee joint angle estimated value and an ankle joint angle estimated value; Acquiring a lower limb joint angle measurement value at the current moment under the walking state of a subject by using an IMU, wherein the lower limb joint angle measurement value comprises a hip joint angle measurement value, a knee joint angle measurement value and an ankle joint angle measurement value; Establishing a state equation by using the lower limb joint angle estimation value, establishing an observation equation by using the lower limb joint angle measurement value, and constructing a fusion model based on extended Kalman filtering; Based on the estimated value of the angle of the lower limb joint at the last moment and the measured value of the angle of the lower limb joint at the current moment, the optimal estimated value of the angle of the lower limb joint at the current moment under the walking state of the subject is obtained by utilizing a fusion model based on extended Kalman filtering.
2. The method according to claim 1, wherein the method further comprises: placing IMUs at the waist, thigh, shank and instep, respectively, over the pelvic bones of the subject; Surface myoelectric sensors were placed at the rectus femoris, biceps femoris, and tibialis anterior of the subject, respectively.
3. The method according to claim 2, wherein the step of acquiring surface electromyographic signals at a previous moment in the walking state of the subject by using the surface electromyographic sensors, and calculating the estimated lower limb joint angle at the previous moment respectively includes: Under the walking state of the subject acquired by three surface myoelectric sensors Surface electromyographic signals at time; Respectively calculating the first muscle activation degree according to the three surface electromyographic signals Second muscle activation And third muscle activation ; According to the following calculation Estimated hip joint angle value at time : Wherein, the 、 Are hip joint angle coefficients; According to the following calculation Time-of-day knee joint angle estimation : Wherein, the 、 Are knee joint angle coefficients; According to the following calculation Ankle joint angle estimation value at time : Wherein, the 、 Are ankle joint angle coefficients.
4. A method according to claim 3, characterized in that the method further comprises: based on a plurality of groups of hip joint angle values and corresponding muscle activation activities, five hip joint angle coefficients corresponding to the hip joint angle values are obtained through fitting: 、 ; based on a plurality of groups of knee joint angle values and corresponding muscle activation degrees, five knee joint angle coefficients corresponding to the knee joint angle values are obtained through fitting: 、 ; Based on a plurality of groups of ankle joint angle values and corresponding muscle activation degrees, five ankle joint angle coefficients corresponding to the ankle joint angle values are obtained through fitting: 、。
5. the method of claim 4, wherein the establishing a state equation with the estimated value of the angle of the joint of the lower limb and the measured value of the angle of the joint of the lower limb as an observation equation, the establishing a fusion model based on the extended Kalman filtering comprises: The state equation is: in the formula, Is that A state variable of time; Is that The state variable of moment is related to the lower limb joint angle estimated value; As a function of the state of the device, Is muscle activation; is process noise; The observation equation is: Wherein, the For observing the variables, correlating with the angle measurement of the lower limb joint, In order to observe the noise it is possible, For expanding the measurement function in kalman filtering: 。
6. the method according to claim 5, wherein obtaining the optimal estimated value of the angle of the lower limb joint at the current moment in the walking state of the subject by using the fusion model based on the extended kalman filter based on the estimated value of the angle of the lower limb joint at the previous moment and the measured value of the angle of the lower limb joint at the current moment comprises: the state update equation is: Wherein, the Is that Moment lower limb joint angle state vector: ; An estimated value of the angle of the human lower limb joint at the moment k; Is that Is a first-order derivative of (a), Is that Is a second order differential of (2); One-step prediction value for the state vector; As a one-step predictor of the covariance, Is that A state noise covariance matrix of the moment; Is process noise Is a variance matrix of (a); is a jacobian matrix; the observation update equation is: Wherein, the To observe noise Is a variance matrix of (a); in the form of a gain matrix, Is that Time lower limb joint angle observation vector: ; Is that The measured value of the angle of the lower limb joint at the moment, Is that Is a first-order derivative of (a), Is that Is a second order differential of (2); And The angle of the same joint is described; Is that A state noise covariance matrix of the moment; solved for Is that The time optimal estimation vector of the lower limb joint angle comprises the following steps: The second order differential of the optimal estimated value of the lower limb joint angle at the moment, First-order differential sum of time-of-day optimal estimation values of lower limb joint angles And the optimal estimated value of the lower limb joint angle at the moment.
7. The method according to claim 6, wherein when , Time, non-linear function The method comprises the following steps: Jacobian matrix The method comprises the following steps: Wherein, the Is a gait cycle.
8. The method according to claim 6, wherein when , Nonlinear function The method comprises the following steps: Jacobian matrix The method comprises the following steps: Wherein, the Is a gait cycle.
9. The method according to claim 6, wherein when , Nonlinear function The method comprises the following steps: Jacobian matrix The method comprises the following steps: Wherein, the Is a gait cycle.
10. A human lower limb joint angle calculating device, comprising: The estimating unit is used for acquiring surface electromyographic signals of the last moment in the walking state of the subject by utilizing the surface electromyographic sensors and respectively calculating lower limb joint angle estimated values of the last moment, and comprises a hip joint angle estimated value, a knee joint angle estimated value and an ankle joint angle estimated value; The measuring unit is used for acquiring a lower limb joint angle measured value at the current moment under the walking state of the subject by using the IMU, and comprises a hip joint angle measured value, a knee joint angle measured value and an ankle joint angle measured value; The construction unit is used for establishing a state equation by using the lower limb joint angle estimation value, establishing an observation equation by using the lower limb joint angle measurement value, and constructing a fusion model based on the extended Kalman filtering; And the fusion unit is used for obtaining the optimal estimated value of the lower limb joint angle at the current moment under the walking state of the subject by utilizing the fusion model based on the extended Kalman filtering based on the estimated value of the lower limb joint angle at the last moment and the measured value of the lower limb joint angle at the current moment.

Description

Human body lower limb joint angle calculation method and device Technical Field The application relates to the technical field of human engineering, in particular to a method and a device for calculating the angle of a human lower limb joint. Background At present, an inertial measurement unit (Inertial Measurement Unit, IMU) is generally adopted for measuring the joint angle when a human body walks, namely, the IMU is respectively placed at the waist, the thigh, the shank and the instep above the pelvic bone, and the joint angle is obtained by calculating the relative direction difference of the two IMUs. But the joint angle measurement obtained by using the IMU alone has the technical problem of hysteresis. Disclosure of Invention In view of the above, the present application provides a method and apparatus for calculating the angle of the joints of the lower limbs of a human body, so as to solve the above-mentioned technical problems. In a first aspect, an embodiment of the present application provides a method for calculating a joint angle of a lower limb of a human body, including: collecting surface electromyographic signals of the last moment in the walking state of the subject by using a surface electromyographic sensor, and respectively calculating the lower limb joint angle estimated value of the last moment, wherein the lower limb joint angle estimated value comprises a hip joint angle estimated value, a knee joint angle estimated value and an ankle joint angle estimated value; Acquiring a lower limb joint angle measurement value at the current moment under the walking state of a subject by using an IMU, wherein the lower limb joint angle measurement value comprises a hip joint angle measurement value, a knee joint angle measurement value and an ankle joint angle measurement value; Establishing a state equation by using the lower limb joint angle estimation value, establishing an observation equation by using the lower limb joint angle measurement value, and constructing a fusion model based on extended Kalman filtering; Based on the estimated value of the angle of the lower limb joint at the last moment and the measured value of the angle of the lower limb joint at the current moment, the optimal estimated value of the angle of the lower limb joint at the current moment under the walking state of the subject is obtained by utilizing a fusion model based on extended Kalman filtering. In one possible implementation, the method further comprises: placing IMUs at the waist, thigh, shank and instep, respectively, over the pelvic bones of the subject; Surface myoelectric sensors were placed at the rectus femoris, biceps femoris, and tibialis anterior of the subject, respectively. In one possible implementation, the method for acquiring the surface electromyographic signals of the last moment in the walking state of the subject by using the surface electromyographic sensors, and respectively calculating the estimated value of the joint angle of the lower limb at the last moment comprises the following steps: Under the walking state of the subject acquired by three surface myoelectric sensors Surface electromyographic signals at time; Respectively calculating the first muscle activation degree according to the three surface electromyographic signals Second muscle activationAnd third muscle activation; According to the following calculationEstimated hip joint angle value at time: Wherein, the 、Are hip joint angle coefficients; According to the following calculation Time-of-day knee joint angle estimation: Wherein, the 、Are knee joint angle coefficients; According to the following calculation Ankle joint angle estimation value at time: Wherein, the 、Are ankle joint angle coefficients. In one possible implementation, the method further comprises: based on a plurality of groups of hip joint angle values and corresponding muscle activation activities, five hip joint angle coefficients corresponding to the hip joint angle values are obtained through fitting: 、; based on a plurality of groups of knee joint angle values and corresponding muscle activation degrees, five knee joint angle coefficients corresponding to the knee joint angle values are obtained through fitting: 、; Based on a plurality of groups of ankle joint angle values and corresponding muscle activation degrees, five ankle joint angle coefficients corresponding to the ankle joint angle values are obtained through fitting: 、。 In one possible implementation, a state equation is established by using the estimated value of the angle of the lower limb joint, and a fusion model based on the extended Kalman filtering is established by using the measured value of the angle of the lower limb joint as an observation equation, wherein the method comprises the following steps: The state equation is: in the formula, Is thatA state variable of time; Is that The state variable of moment is related to the lower limb joint angle estimated value; As a function of