阅读说明：本技术 一种实时追踪人体下肢运动的穿戴设备 (Wearing equipment for tracking motion of lower limbs of human body in real time ) 是由 刘礼 李小虎 廖军 雍滋蕊 钱爽 邓正巧 于 2021-08-24 设计创作，主要内容包括：本发明公开一种实时追踪人体下肢运动的穿戴设备,包括可穿戴主体(1)、感知模块(2)、主控模块(3)和上位机；本发明提供了一种有线/无线双模的实时追踪人体下肢运动的穿戴设备,配合上位机算力,结合高识别率的态势感知算法和机器学习状态预测算法,在考虑穿戴舒适性的情况下,有效平衡了设备成本和运动捕捉准确率的问题,能实时地感知人体大腿与小腿姿态变化,追踪下肢移动轨迹、识别下肢动作,兼具灵活性、低延时、结构简单、制作成本低等优点。(The invention discloses wearable equipment for tracking the motion of lower limbs of a human body in real time, which comprises a wearable main body (1), a sensing module (2), a main control module (3) and an upper computer; the invention provides a wired/wireless dual-mode wearing device for tracking the motion of lower limbs of a human body in real time, which is matched with the computational power of an upper computer, is combined with a situation perception algorithm with high recognition rate and a machine learning state prediction algorithm, effectively balances the problems of device cost and motion capture accuracy rate under the condition of considering wearing comfort, can sense the posture change of thighs and shanks of the human body in real time, tracks the motion track of the lower limbs and identifies the motion of the lower limbs, and has the advantages of flexibility, low time delay, simple structure, low manufacturing cost and the like.)

1. The utility model provides a wearing equipment of human low limbs motion is tracked in real time which characterized in that: the wearable sensing device comprises a wearable main body (1), a sensing module (2), a main control module (3) and an upper computer.

The wearable main body (1) comprises n wearable parts, wherein n-1 wearable parts are provided with sensing modules (2), and 1 wearable part is provided with a main control module (3);

the sensing module (2) monitors a motion signal of the wearable main body (1) worn by a user in real time and transmits the motion signal to the main control module (3);

the main control module (3) performs attitude calculation on the received motion signal to obtain an attitude calculation result;

the main control module (3) sends the attitude calculation result and the motion signal to an upper computer;

and the upper computer calculates the attitude calculation result and the motion signal to obtain the user space motion attitude.

2. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 1, wherein: the sensing module (2) comprises a nine-axis inertial sensor; the nine-axis inertial sensor comprises a three-axis gyroscope, a three-axis accelerometer and a three-axis magnetometer;

the nine-axis inertial sensor monitors a motion signal of a user; the motion signals comprise 3-axis linear acceleration sensing signals, 3-axis gyroscope sensing signals and 3-axis magnetometer sensing signals;

the nine-axis inertial sensor can move within the range of human lower limb joint kinematics; the six modules are detachable.

3. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 2, wherein the step of performing posture calculation on the received motion signals by the main control module (3) comprises:

1) establishing a user posture information equation, namely:

in the formula, parameter i²＝j²＝k²＝ijk＝-1；u＝[u_x，u_y，u_z]^TAn object rotating shaft under ground coordinates; theta is a rotation angle; (q) a₀，q₁，q₂，q₃) Is a quaternion;

2) establishing a rotation matrix R_GNamely:

3) utilizing a three-axis gyroscope to update the attitude of the quaternion to obtain an updated quaternion Q of the gyroscope_w，tNamely:

in the formula, Q_t-1The estimated quaternion which is calibrated at the last moment; Δ t is the time interval;

wherein the quaternion derivative of the gyroscopeAs follows:

in the formula, w_t＝[0，w_x，w_y，w_z]^T3-axis gyroscope sensing signals monitored by a gyroscope;

3) obtaining a gradient quaternion by using a gradient descent method, comprising the following steps of:

3.1) separately calculating the accelerometer estimatesAnd error function f of accelerometer_aNamely:

in the formula (I), the compound is shown in the specification,is the value of the accelerometer measurement after normalization; rotation matrix R_b＝(R_G)^-1；g_G＝[0，0，1]^TIs the acceleration of gravity;

3.2) calculating the error function f_aObtaining a Jacobian matrix J for partial derivatives of the quaternion Q_aNamely:

3.3) calculating the gradient of the error function of the accelerometerNamely:

3.4) setting of the intermediate variablesAmount h_mNamely:

in the formula (I), the compound is shown in the specification,is the value of the normalized magnetometer measurement; h is_x，h_y，h_zAn element that is an intermediate variable;

3.5) calculating an estimate of the intermediate variableAnd error function f of the magnetometer_m；

3.6) calculating the error function f_mObtaining a Jacobian matrix J for partial derivatives of the quaternion Q_mNamely:

3.7) calculating the gradient of the error function of the magnetometerNamely:

3.9) calculating the overall gradientNamely:

in the formula, vector(Vector)

3.10) according to the overall gradientIteratively updating gradient quaternion in opposite direction of gradientObtaining:

wherein the optimal iteration step size mu_tAs follows:

wherein α is a gain; alpha is more than 1;

4) and fusing the quaternion of the gyroscope and the quaternion of the gradient, and updating the final quaternion.

In the formula, gamma_tIs a weight;

5) updating the quaternion through a formula (17) to obtain a user attitude calculation result, namely:

in the formula (I), the compound is shown in the specification,representing the spatial angle.

4. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 1, wherein the step of calculating the posture calculation result and the motion signal by the upper computer comprises:

1) and outputting the postures of the leg joints of the user by using the D-H transformation matrix, namely:

loc_{ankle-to-knee}＝g(θ₁，θ₂，θ₃，θ₄) (20)

loc_ankle＝loc_knee+loc_{ankle-to-knee} (21)

in the formula, the parameter l_u＝||loc_knee||；loc_kneeIs the knee position obtained by D-H transformation; theta₁、θ₂、θ₃The thigh femur pitch angle, the roll angle and the yaw angle are respectively; theta₄Is the calf shank pitch angle; loc_{ankle-to-knee}Is the relative position of the ankle and knee; loc_ankleIs the absolute position of the ankle;

2) calculating the velocity v of the knee_knee(t), namely:

velocity v of knee movement_knee(t) bringing the formulas (20) to (21) to obtain:

in the formula: acecel_l，tIs the acceleration value measured by the inertial sensor at the lower leg; rot is a rotation matrix converted to a lower limb coordinate system; v. of_v-1The speed at the previous moment;

3) predict the current state X (k | k-1), i.e.:

X(k|k-1)＝A·X(k-1|k-1)+B·U(k) (24)

wherein A and B are parameter matrices; the states include knee position and knee velocity; x (k-1| k-1) is the optimal predicted value of the last state; u (k) is a control amount of the current state;

4) the covariance P (k | k-1) of the current state is predicted, i.e.:

P(k|k-1)＝A·P(k-1|k-1)A^T+N (25)

wherein P (k-1| k-1) is the covariance for X (k-1| k-1); a. the^TIs the transposed matrix of A; n is the noise of the system;

5) estimate the reference state X (k | k), i.e.:

X(k|k)＝X(k|k-1)+k_g(k)·(Z(k)-H·X(k|k-1)) (26)

in the formula, k_g(k) Is the kalman gain at time k; h is a measurement parameter; z (k) is a status measurement value;

wherein, the Kalman gain k at the k moment_g(k) As follows:

k_g(k)＝P(k|k-1)·H^T·(H·P(k|k-1)·H^T+∑)^-1 (27)

wherein P (k | k-1) is the covariance at time k-1; Σ is the uncertainty of the observed distribution;

the covariance P (k | k) at time k is as follows:

P(k|k)＝(1-k_g(k)·H)·P(k|k-1) (28)

6) according to the position and the spatial angle of the kneeA spatial motion pose of the user is determined.

5. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 1, wherein: the system also comprises a communication module (4) for information interaction between the upper computer and the main control module (3).

6. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 5, wherein: the communication module (4) is a wireless communication module or a wired serial port communication module.

7. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 5, wherein: the power supply module (5) is used for supplying power to the sensing module (2), the main control module (3) and the communication module (4).

8. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 7, wherein: the power supply module (5) comprises a charging module, a voltage-stabilizing discharge power panel and a lithium battery;

the charging module is used for charging the lithium battery;

when the lithium battery discharges, the voltage is stabilized by the voltage stabilizing module and then supplies power to the sensing module (2), the main control module (3) and the communication module (4).

9. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 7, wherein: also comprises a connecting belt (6);

the connecting band (6) is including the signal line of connecting perception module (2) and host system (3), connects the signal line of host system (3) and communication module (4), connects perception module (2), host system (3), communication module (4), power supply module's power supply line.

10. The wearable device for tracking the motion of the lower limbs of the human body in real time according to claim 1, wherein: the wearable component with the sensing module (2) is a bandage for the lower limbs of the user to wear.

Technical Field

The invention relates to the field of intelligent wearing and human-computer interaction, in particular to wearing equipment for tracking the motion of lower limbs of a human body in real time.

Background

The human motion capture technology has great application value in the fields of automatic control, interactive entertainment, virtual reality, medical rehabilitation, motion training and the like.

The user motion trail tracking based on the wearable device at present has the problems of poor real-time performance, low accuracy and the like, can not provide good experience for a user, and even causes strong invasion for the user due to the wearing characteristic.

In addition, a human motion tracking method based on optical marks or images exists, although the identification accuracy is high, the method is limited by field illumination and the identification range of a camera system, the shielding problem still has no good solution, and the calculation and storage expenses generated by image processing are often large, so that the cost of the human motion tracking equipment is high.

Disclosure of Invention

The invention aims to provide wearable equipment for tracking the motion of lower limbs of a human body in real time, which comprises a wearable main body, a sensing module, a main control module and an upper computer, wherein the wearable main body is provided with a sensing module;

the wearable main body comprises n wearable components, wherein n-1 wearable components are provided with sensing modules, and 1 wearable component is provided with a main control module;

the sensing module monitors a motion signal of the wearable main body worn by the user in real time and transmits the motion signal to the main control module;

the master control module performs attitude calculation on the received motion signals to obtain attitude calculation results;

the master control module sends the attitude calculation result and the motion signal to an upper computer;

and the upper computer calculates the attitude calculation result and the motion signal to obtain the user space motion attitude.

Further, the sensing module comprises a nine-axis inertial sensor; the nine-axis inertial sensor comprises a three-axis gyroscope, a three-axis accelerometer and a three-axis magnetometer;

the nine-axis inertial sensor monitors a motion signal of a user; the motion signals comprise 3-axis linear acceleration sensing signals, 3-axis gyroscope sensing signals and 3-axis magnetometer sensing signals;

the nine-axis inertial sensor can move within the range of human lower limb joint kinematics; the six modules are detachable.

Further, the step of the master control module performing attitude calculation on the received motion signal comprises:

1) establishing a user posture information equation, namely:

in the formula, parameter i²＝j²＝k²＝ijk＝-1；u＝[u_x，u_y，u_z]^TAn object rotating shaft under ground coordinates; theta is a rotation angle; (q) a₀，q₁，q₂，q₃) Is a quaternion;

2) establishing a rotation matrix R_GNamely:

3) utilizing a three-axis gyroscope to update the attitude of the quaternion to obtain an updated quaternion Q of the gyroscope_w，tNamely:

in the formula, Q_t-1The estimated quaternion which is calibrated at the last moment; Δ t is the time interval;

wherein the quaternion derivative of the gyroscopeAs follows:

in the formula, w_t＝[0，w_x，w_y，w_z]^T3-axis gyroscope sensing signals monitored by a gyroscope;

3) obtaining a gradient quaternion by using a gradient descent method, comprising the following steps of:

3.1) separately calculating the accelerometer estimatesAnd error function f of accelerometer_aNamely:

in the formula (I), the compound is shown in the specification,is the value of the accelerometer measurement after normalization; rotation matrix R_b＝(R_G)^-1；g_G＝[0，0，1]^TIs the acceleration of gravity;

3.2) calculating the error function f_aObtaining a Jacobian matrix J for partial derivatives of the quaternion Q_aNamely:

3.3) calculating the gradient of the error function of the accelerometerNamely:

3.4) setting the intermediate variable h_mNamely:

in the formula (I), the compound is shown in the specification,is the value of the normalized magnetometer measurement; h is_x，h_y，h_zAn element that is an intermediate variable;

3.5) calculating an estimate of the intermediate variableAnd error function f of the magnetometer_m；

3.6) calculating the error function f_mObtaining a Jacobian matrix J for partial derivatives of the quaternion Q_mNamely:

3.7) calculating the gradient of the error function of the magnetometerNamely:

3.9) calculating the overall gradientNamely:

in the formula, vector(Vector)

3.10) according to the overall gradientIteratively updating gradient quaternion in opposite direction of gradientObtaining:

wherein the optimal iteration step size mu_tAs follows:

wherein α is a gain; alpha is more than 1;

4) and fusing the quaternion of the gyroscope and the quaternion of the gradient, and updating the final quaternion.

In the formula, gamma_tIs a weight;

5) updating the quaternion through a formula (17) to obtain a user attitude calculation result, namely:

in the formula (I), the compound is shown in the specification,representing the spatial angle.

Further, the step of resolving the attitude resolving result and the motion signal by the upper computer comprises:

a) and outputting the postures of the leg joints of the user by using the D-H transformation matrix, namely:

loc_{ankle-to-knee}＝g(θ₁，θ₂，θ₃，θ₄) (20)

loc_ankle＝loc_knee+loc_{ankle-to-knee} (21)

in the formula, the parameter l_u＝||loc_knee||；loc_kneeIs the knee position obtained by D-H transformation; theta₁、θ₂、θ₃The thigh femur pitch angle, the roll angle and the yaw angle are respectively; theta₄Is the calf shank pitch angle; loc_{ankle-to-knee}Is the relative position of the ankle and knee; loc_ankleIs the absolute position of the ankle;

b) calculating the velocity v of the knee_knee(t), namely:

velocity v of knee movement_knee(t) bringing the formulas (20) to (21) to obtain:

in the formula: acecel_l，tIs the acceleration value measured by the inertial sensor at the lower leg; rot is a rotation matrix converted to a lower limb coordinate system; v. of_t-1The speed at the previous moment;

c) predict the current state X (k | k-1), i.e.:

X(k|k-1)＝A·X(k-1|k-1)+B·U(k) (24)

wherein A and B are parameter matrices; the states include knee position and knee velocity; x (k-1| k-1) is the optimal predicted value of the last state; u (k) is a control amount of the current state;

d) the covariance P (k | k-1) of the current state is predicted, i.e.:

P(k|k-1)＝A·P(k-1|k-1)A^T+N (25)

wherein P (k-1| k-1) is the covariance for X (k-1| k-1); a. the^TIs the transposed matrix of A; n is the noise of the system;

e) estimate the reference state X (k | k), i.e.:

X(k|k)＝X(k|k-1)+k_g(k)·(Z(k)-H·X(k|k-1)) (26)

in the formula, k_g(k) Is the kalman gain at time k; h is a measurement parameter; z (k) is a status measurement value;

wherein, the Kalman gain k at the k moment_g(k) As follows:

k_g(k)＝P(k|k-1)·H^T·(H·P(k|k-1)·H^T+∑)^-1 (27)

wherein P (k | k-1) is the covariance at time k-1;

the covariance P (k | k) at time k is as follows:

P(k|k)＝(1-k_g(k)·H)·P(k|k-1) (28)

where Σ is the uncertainty of the observed distribution;

f) according to the position and the spatial angle of the kneeA spatial motion pose of the user is determined.

Further, the system also comprises a communication module used for information interaction between the upper computer and the main control module.

Further, the communication module is a wireless communication module or a wired serial communication module.

And the power supply module supplies power to the sensing module, the main control module and the communication module.

Further, the power supply module comprises a charging module, a voltage-stabilizing discharge power panel and a lithium battery;

the charging module is used for charging the lithium battery;

when the lithium battery discharges, the voltage is stabilized by the voltage stabilizing module and then supplies power to the sensing module, the main control module and the communication module.

Further, the connecting belt is also included;

the connecting band is including the signal line of connecting perception module and host system, the signal line of connecting host system and communication module, the power supply line of connecting perception module, host system, communication module, power module.

Further, the wearable component with the perception module is a bandage for wearing on the lower limbs of the user.

The invention has the advantages that the invention is based on the micro-electro-mechanical system technology, can dynamically track the motion track of the lower limbs of the user in real time and identify the motion posture of the lower limbs, aims to solve the problems in the prior art, is not limited by a field, ensures the accuracy of real-time tracking of the motion of the lower limbs of the human body to be within an acceptable range with less equipment cost, and is accurately, efficiently and portably applied to the motion capture fields of control, interaction, rehabilitation and the like.

The invention provides a wired/wireless dual-mode wearing device for tracking the motion of lower limbs of a human body in real time, which is matched with the computational power of an upper computer, is combined with a situation perception algorithm with high recognition rate and a machine learning state prediction algorithm, effectively balances the problems of device cost and motion capture accuracy rate under the condition of considering wearing comfort, can sense the posture change of thighs and shanks of the human body in real time, tracks the motion track of the lower limbs and identifies the motion of the lower limbs, and has the advantages of flexibility, low time delay, simple structure, low manufacturing cost and the like.

Drawings

FIG. 1 is a block diagram of a wearable device for tracking human lower limb movement in real time;

FIG. 2 is a schematic view of a wearing structure of the wearable device for tracking the motion of the lower limbs of the human body in real time;

FIG. 3 is a flowchart of an algorithm for tracking human lower limb movement in real time;

FIG. 4 is a circuit diagram of a wearable device for tracking human lower limb movement in real time;

FIG. 5 is a schematic view of the lower limb fitness operation of the wearable device for tracking the human lower limb movement in real time according to embodiment 6; FIG. 5(a) is a schematic diagram of lower limb fitness activity I; FIG. 5(b) is a schematic view of lower limb fitness activity II; FIG. 5(c) is a schematic view of lower limb fitness activity III; FIG. 5(d) is a schematic view of lower limb fitness activity IV;

in the figure: wearable main part 1, perception module 2, host system 3, communication module 4, power module 5, connecting band 6.

Detailed Description

The present invention is further illustrated by the following examples, but it should not be construed that the scope of the above-described subject matter is limited to the following examples. Various substitutions and alterations can be made without departing from the technical idea of the invention and the scope of the invention is covered by the present invention according to the common technical knowledge and the conventional means in the field.

Example 1:

referring to fig. 1 to 4, a wearable device for tracking the motion of the lower limbs of a human body in real time comprises a wearable main body (1), a sensing module 2, a main control module 3, an upper computer, a communication module 4 for information interaction between the upper computer and the main control module 3, a power supply module 5 for supplying power to the sensing module 2, the main control module 3 and the communication module 4, and a connecting band 6;

the wearable main body 1 comprises 3 wearable components, wherein 2 wearable components are respectively provided with a sensing module 2, and 1 wearable component is provided with a main control module 3, a communication module 4 and a power supply module 5;

the wearable part with the sensing module 2 is a bandage for wearing on the lower limbs of the user.

The wearable part with the sensing module 2 comprises a bandage of a human shank and a human thigh. The wearable part with the main control module 3, the communication module 4 and the power supply module 5 comprises a human waist bandage.

The sensing module 2 monitors the motion signal of the wearable main body 1 worn by the user in real time and transmits the motion signal to the main control module 3;

the sensing module 2 comprises a nine-axis inertial sensor; the nine-axis inertial sensor comprises a three-axis gyroscope, a three-axis accelerometer and a three-axis magnetometer;

the nine-axis inertial sensor monitors a motion signal of a user; the motion signals comprise 3-axis linear acceleration sensing signals, 3-axis gyroscope sensing signals and 3-axis magnetometer sensing signals;

the nine-axis inertial sensor can move within the range of human lower limb joint kinematics; the six modules are detachable.

The main control module 3 performs attitude calculation on the received motion signal to obtain an attitude calculation result;

the main control module 3 sends the attitude calculation result and the motion signal to an upper computer;

the step of the master control module 3 performing attitude calculation on the received motion signal comprises:

1) establishing a user posture information equation, namely:

in the formula, parameter i²＝j²＝k²＝ijk＝-1；u＝[u_x，u_y，u_z]^TAn object rotating shaft under ground coordinates; theta is a rotation angle; the parameters i, j are imaginary numbers. q. q.s₀、q₁、q₃Respectively represent;

2) establishing a rotation matrix R_GNamely:

3) utilizing a three-axis gyroscope to update the attitude of the quaternion to obtain an updated quaternion Q of the gyroscope_w，tNamely:

in the formula, Q_t-1The estimated quaternion which is calibrated at the last moment; Δ t is the time interval;

wherein the quaternion derivative of the gyroscopeAs follows:

in the formula, w_t＝[0，w_x，w_y，w_z]^T3-axis gyroscope sensing signals monitored by a gyroscope; w is a_x，w_y，w_zGyroscope sensor signals on the x, y and z axes, respectively.

3) Obtaining a gradient quaternion by using a gradient descent method, comprising the following steps of:

3.1) separately calculating the accelerometer estimatesAnd error function f of accelerometer_aNamely:

in the formula (I), the compound is shown in the specification,is the value of the accelerometer measurement after normalization; rotation matrix R_b＝(R_G)^-1；g_G＝[0，0，1]^TIs the acceleration of gravity; a is_x，a_y，a_zThe values measured by the accelerometers on the x, y, and z axes, respectively.

3.2) calculating the error function f_aObtaining a Jacobian matrix J for partial derivatives of the quaternion Q_aNamely:

in the formula (I), the compound is shown in the specification,respectively representing error functions f_aThe partial derivatives in the x, y, z axes.

3.3) calculating the gradient of the error function of the accelerometerNamely:

3.4) setting the intermediate variable h_mNamely:

in the formula (I), the compound is shown in the specification,is the value of the normalized magnetometer measurement; h is_x，h_y，h_zAn element that is an intermediate variable;

3.5) calculating an estimate of the intermediate variableAnd error function f of the magnetometer_m；

3.6) calculating the error function f_mObtaining a Jacobian matrix J for partial derivatives of the quaternion Q_mNamely:

in the formula (I), the compound is shown in the specification,respectively representing error functions f_mThe partial derivatives in the x, y, z axes.

3.7) calculating the gradient of the error function of the magnetometerNamely:

3.9) calculating the overall gradientNamely:

in the formula, vector(Vector)

3.10) according to the overall gradientIteratively updating gradient quaternion in opposite direction of gradientObtaining:

wherein the optimal iteration step size mu_tAs follows:

wherein α is a gain; alpha is more than 1;

4) and fusing the quaternion of the gyroscope and the quaternion of the gradient, and updating the final quaternion.

In the formula, gamma_tIs a weight;

5) updating the quaternion through a formula (17) to obtain a user attitude calculation result, namely:

in the formula (I), the compound is shown in the specification,representing the spatial angle.

And the upper computer calculates the attitude calculation result and the motion signal to obtain the user space motion attitude.

The step of resolving the attitude resolving result and the motion signal by the upper computer comprises the following steps:

a) and outputting the postures of the leg joints of the user by using the D-H transformation matrix, namely:

loc_{ankle-to-knee}＝g(θ₁，θ₂，θ₃，θ₄) (20)

loc_ankle＝loc_knee+loc_{ankle-to-knee} (21)

in the formula, the parameter l_u＝||loc_knee||；loc_kneeIs the knee position obtained by D-H transformation; theta₁、θ₂、θ₃The thigh femur pitch angle, the roll angle and the yaw angle are respectively; theta₄Is the calf shank pitch angle; loc_{ankle-to-knee}Is the relative position of the ankle and knee; loc_ankleIs the absolute position of the ankle;

b) calculating the velocity v of the knee_knee(t), namely:

velocity v of knee movement_knee(t) bringing the formulas (20) to (21) to obtain:

in the formula: acecel_l，tIs the acceleration value measured by the inertial sensor at the lower leg; rot is a rotation matrix converted to a lower limb coordinate system; v. of_t-1The speed at the previous moment;

c) predict the current state X (k | k-1), i.e.:

X(k|k-1)＝A·X(k-1|k-1)+B·U(k) (24)

wherein A and B are parameter matrices; the states include knee position and knee velocity; x (k-1| k-1) is the optimal predicted value of the last state; u (k) is a control amount of the current state;

d) the covariance P (k | k-1) of the current state is predicted, i.e.:

P(k|k-1)＝A·P(k-1|k-1)A^T+N (25)

wherein P (k-1| k-1) is the covariance for X (k-1| k-1); a. the^TIs the transposed matrix of A; n is the noise of the system;

e) estimate the reference state X (k | k), i.e.:

X(k|k)＝X(k|k-1)+k_g(k)·(Z(k)-H·X(k|k-1)) (26)

in the formula, k_g(k) Is the kalman gain at time k; h is a measurement parameter; z (k) is a status measurement value;

wherein, the Kalman gain k at the k moment_g(k) As follows:

k_g(k)＝P(k|k-1)·H^T·(H·P(k|k-1)·H^T+∑)^-1 (27)

wherein P (k | k-1) is the covariance at time k-1; Σ is the uncertainty of the observed distribution;

the covariance P (k | k) at time k is as follows:

P(k|k)＝(1-k_g(k)·H)·P(k|k-1) (28)

f) according to the position and the spatial angle of the kneeA spatial motion pose of the user is determined.

The communication module 4 is a wireless communication module or a wired serial communication module.

The power supply module 5 comprises a charging module, a voltage-stabilizing discharge power panel and a lithium battery;

the charging module is used for charging the lithium battery;

when the lithium battery discharges, the voltage is stabilized by the voltage stabilizing module and then supplies power to the sensing module 2, the main control module 3 and the communication module 4.

The connecting band 6 is including the signal line of connecting perception module 2 and host system 3, the signal line of connecting host system 3 and communication module 4, the power supply line of connecting perception module 2, host system 3, communication module 4, power module.

Example 2:

referring to fig. 1 and 2, a wearable device for dynamically tracking the motion of lower limbs of a human body in real time mainly comprises a wearable main body 1, a sensing module 2, a main control module 3, a communication module 4, a power supply module 5, a connecting band 6 and an upper computer.

The wearable main body 1 comprises human shank, thigh and waist bandage;

further, the position of the waist bandage of the wearable main body 1 is provided with a main switch and a wireless communication switch.

The wearable main body 1 is provided with a perception module 2 on the crus and thigh bandage, and the waist bandage is provided with a main control module 3, a communication module 4 and a power supply module 5.

Six modules can be dismantled, and the bandage is adjustable, consequently, the wearing equipment can adjust to different users.

The sensing module 2 monitors and collects the posture change of the wearable main body 1 in real time;

the sensing module 2 mainly comprises nine-axis inertial sensors, namely a three-axis gyroscope, a three-axis accelerometer and a three-axis magnetometer;

nine inertial sensor are set up in wearable main part 1 shank, thigh bandage department, and it is portable in human low limbs joint accords with the kinematics within range, monitors the motion state of human low limbs, obtains the motion signal of human thigh and shank: the three-axis gyroscope comprises two groups of 3-axis linear acceleration sensing signals, two groups of 3-axis gyroscope sensing signals and two groups of 3-axis magnetometer sensing signals, and the two groups of 3-axis linear acceleration sensing signals are transmitted to the main control module 3 through a connecting belt 6.

Referring to the flow chart of fig. 3, the main control module 3 first performs elliptic curve calibration on the magnetometer and performs static calibration on the accelerometer and the gyroscope. After the calibration is completed, the posture of the leg movement signals acquired in parallel is resolved to obtain a resolving result, the original data and the posture information are transmitted to the upper computer through the communication module 4, the upper computer performs constraint processing on the posture information according to the constructed mathematical model, the lower limb movement posture of the wearer is predicted and tracked by combining the original data, and the lower limb movement track drawing is completed.

The communication module 4 is a wireless communication module or a wired serial communication module.

The communication module 4 has an indicator light, and the wireless communication module is not limited to bluetooth, Zigbee, Wi-fi and other wireless protocols.

Under wireless mode, need open 1 waist master switch of wearable body and wireless communication switch, carry out the bluetooth and pair, connect the pilot lamp on the successful back communication module and become normally bright by the scintillation.

Under wired mode, open 1 waist master switch of wearable body, close wireless communication switch, pass through the data line with wearable body and host computer and be connected and can use.

The connecting band 6 is connected with the shank, the thigh and the waist, is communicated with the signal lines of the sensing module 2 and the main control module 3, and also comprises a power supply line communicated with the power supply module 5.

The power supply module 5 supplies power to the sensing module 2, the main control module 3 and the communication module 4 through a connecting belt 6;

the power supply module 5 comprises a charging and voltage-stabilizing discharging module and a lithium battery;

the voltage of the lithium battery is stabilized by the voltage stabilizing module and then supplies power to the sensing module 2, the main control module 3 and the communication module 4 through the connecting band 6.

Example 3:

the utility model provides a wearing equipment of human low limbs motion is tracked in real time developments, mainly includes wearable main part 1, perception module 2, host system 3, communication module 4, power module 5, connecting band 6 and host computer.

The wearable main body 1 comprises human shank, thigh and waist bandage;

the waist bandage position of the wearable main body 1 is provided with a main switch and a wireless communication switch.

The wearable main body 1 is provided with a perception module 2 on the crus and thigh bandage, and the waist bandage is provided with a main control module 3, a communication module 4 and a power supply module 5.

The sensing module 2 monitors and collects the posture change of the wearable main body 1 in real time;

the sensing module 2 mainly comprises nine-axis inertial sensors, namely a three-axis gyroscope, a three-axis accelerometer and a three-axis magnetometer;

the nine-axis inertial sensor is arranged at the position of a shank and a thigh bandage of the wearable main body 1, can move within a human lower limb joint motion conforming range, and monitors the motion state of the human lower limb.

The main control module 3 firstly performs elliptic curve calibration on the magnetometer, and performs static calibration on the accelerometer and the gyroscope. Because the magnetometer is sensitive to the environmental magnetic field, when the wearable device is in a complex magnetic field scene, the calibration of the elliptic curve fails, when the attitude is resolved, the gradient descent algorithm adjusts the confidence weight of the geomagnetic signals collected by the magnetometer downwards or sets the magnetic error function to 0, and transmits an instruction signal that the geomagnetic data is not credible to the upper computer.

The communication module 4 is a wireless communication module or a wired serial communication module.

The connecting band 6 is connected with the shank, the thigh and the waist, is communicated with the signal lines of the sensing module 2 and the main control module 3, and also comprises a power supply line communicated with the power supply module 5.

The power supply module 5 supplies power to the sensing module 2, the main control module 3 and the communication module 4 through a connecting belt 6.

Example 4:

the utility model provides a wearing equipment of human low limbs motion is tracked in real time developments, mainly includes wearable main part 1, perception module 2, host system 3, communication module 4, power module 5, connecting band 6 and host computer.

The wearable main body 1 comprises human shank, thigh and waist bandage;

the waist bandage position of the wearable main body 1 is provided with a main switch and a wireless communication switch.

The sensing module 2 monitors and collects the posture change of the wearable main body 1 in real time;

the sensing module 2 mainly comprises a nine-axis inertial sensor;

the nine-axis inertial sensor is arranged at the position of a shank and a thigh bandage of the wearable main body 1, can move within a human lower limb joint motion conforming range, and monitors the motion state of the human lower limb.

The main control module 3 performs attitude calculation on the received acceleration, magnetic force and angular speed data by using a gradient descent method, sends the result and the original data to an upper computer, and the upper computer performs further state prediction to draw a tracking track of the lower limb movement. The upper computer monitors tracking delay in real time, only completes constraint processing on a calculation result when the delay exceeds a threshold value, does not make further state prediction, and directly outputs a lower limb movement track.

The communication module 4 is a wireless communication module or a wired serial communication module.

The connecting band 6 is connected with the shank, the thigh and the waist, is communicated with the signal lines of the sensing module 2 and the main control module 3, and also comprises a power supply line communicated with the power supply module 5.

The power supply module 5 supplies power to the sensing module 2, the main control module 3 and the communication module 4 through a connecting belt 6.

Example 5:

referring to fig. 4, a self-made wearing device circuit design diagram for tracking the motion of the lower limbs of the human body in real time mainly comprises the following steps:

waist bandage department installation have Arduino Nano development board, bluetooth module, power module. Wherein, the Arduino Nano that this embodiment used is the miniature version of Arduino USB interface, and it is convenient, small in size to develop, and the computing power is sufficient in this example. The Bluetooth module that the schematic diagram shows carries out wireless communication with the host computer. The lithium battery of the power module provides electric power for tracking the lower limb movement wearable equipment through the voltage boosting and stabilizing module.

In this embodiment, contain two nine inertia measurement unit MPU9250, MPU9250 is a space motion sensor chip, built-in accelerometer, gyroscope and magnetometer, parallelly independent collection single leg thigh, the motion information of shank, contain triaxial linear acceleration, triaxial angular velocity and triaxial magnetic force measurement signal, the collection frequency satisfies human activity completely, its characteristics are that its is small easily to carry, can gather nine inertia data, the price is low, the precision is high etc. communicate through I2C bus or SPI protocol with Arduino Nano.

It is noted that the MPU9250 is mounted on the leg strap with the front side facing up and aligned with the XYZ axes of fig. 2 for subsequent calculations.

Example 6:

the original sensor data is subjected to posture calculation and state prediction, the processed data has characteristic information of lower limb movement, the method is not limited to the method of using simple condition constraint and a machine learning classifier to classify and recognize the special actions of the lower limbs, the action completion degree can be evaluated, and the method is applied to the fields of human activity recognition, exercise rehabilitation and the like, and the lower limb body-building action schematic diagram of the wearable device for tracking the human lower limb movement in real time is shown in fig. 5.

Example 7:

a wearing device for tracking the motion of lower limbs of a human body in real time mainly comprises a wearable main body 1, a sensing module 2, a main control module 3, a communication module 4, a power supply module 5, a connecting belt 6 and an upper computer;

the wearable main body 1 comprises human shank, thigh and waist bandage;

the wearable main body 1 is provided with sensing modules 2 on crus and thigh bandages, and the waist bandage of the wearable main body 1 is provided with a main control module 3, a communication module 4 and a power supply module 5;

the sensing module 2 monitors and collects posture change and motion information of the wearable main body 1 in real time;

the sensing module 2 mainly comprises nine-axis inertial sensors, namely a three-axis gyroscope, a three-axis accelerometer and a three-axis magnetometer;

nine inertial sensor are set up in wearable main part 1 shank, thigh bandage department, monitor the motion state of human low limbs, obtain the motion signal of human thigh and shank: the three-axis gyroscope comprises two groups of 3-axis linear acceleration sensing signals, two groups of 3-axis gyroscope sensing signals and two groups of 3-axis magnetometer sensing signals, and the two groups of 3-axis linear acceleration sensing signals are transmitted to the main control module 3 through a connecting belt 6;

the connecting band 6 is connected with the shank, the thigh and the waist, is used for connecting signal lines of the sensing module 2 and the main control module 3, and comprises a power supply line connected with the power supply module 5.

After the calibration of the inertial sensor is completed, the main control module 3 performs attitude calculation on the received sensor information of the crus and the thighs, then sends the obtained attitude calculation result and the original signal to an upper computer through the communication module 4, and takes a gradient descent method for attitude calculation in consideration of real-time performance and the calculation force of the main control module 3, and the main steps are as follows:

1) and expressing the posture information of the lower limbs of the human body by using quaternions. The quaternion representation method can avoid the problem of dead lock of the universal joint caused by using Euler angles and can reduce the complexity of the calculation process. The quaternion form is:

in the formula i²＝j²＝k²＝ijk＝-1，u＝[u_x，u_y，u_z]^TIs the rotation axis of the object under the ground coordinates, and theta is the rotation angle. Given a unit quaternion, the rotation matrix can be constructed as follows:

2) and updating the attitude of the quaternion by using the gyroscope. Solving a quaternion differential equation by using a Longge Kutta method, wherein the updating process is as follows:

wherein, the quaternion Q of the gyroscope_w，tIs an estimated quaternion Q calibrated at the last moment_t-1Derivative with gyroscope quaternionTo complete the update, w_t＝[0，w_x，w_y，w_z]^TIs the angular velocity obtained directly from the gyroscope.

3) A minimization problem is constructed from the measurements of the accelerometer and magnetometer and another set of quaternions is obtained using a gradient descent method.

In the first step, the gradient of the acceleration error function is found. The data measured by the accelerometer is approximately the value of gravity, so the acceleration of gravity g_G＝[0，0，1]^TThrough the rotation matrix R_b＝(R_G)^-1Obtaining an accelerometer estimate:

in the formula (6), the reaction mixture is,the method is characterized in that the measured value of the accelerometer is standardized, the difference between the measured value and the measured value is used for obtaining the error of the accelerometer, the effect of the accelerometer on quaternion updating is reflected, and a function f is further solved_aObtaining a Jacobian matrix J for partial derivatives of the quaternion Q_a：

The gradient of the accelerometer error function is obtained by combining equations (6) and (7):

and secondly, solving the gradient of the magnetic force error function. The direction of the north pole of the earth magnet is taken as the positive direction of the X axis of the reference coordinate system, so that the magnetometer can quickly and automatically converge to the current attitude, and the data of the Y axis is 0. The process of solving the gradient of the magnetic error function is similar to the first step, and for the convenience of description, an intermediate variable h is introduced_m：

Thirdly, combining the error functions of the accelerometer and the magnetometer obtained in the first two steps and the Jacobian matrix thereof, obtaining the overall gradient by the formulas (8) and (13):

the minimization problem is to make the error function f of the accelerometer and the magnetometer_a，mMinimum value, according to overall gradientIteratively updating gradient quaternion in opposite direction of gradientTo complete the correction:

in the formula: mu.s_tFor the iteration step, the optimal values are as follows:

wherein alpha is gain and is adjusted by prior knowledge, and alpha is more than 1.

4) And fusing the quaternion of the gyroscope and the quaternion of the gradient, and updating the final quaternion.

When the convergence rate of the gyro quaternion and the convergence rate of the gradient quaternion are the same as the weights in front of them, i.e., (1- γ)_t)β＝γ_tμ_t/Δt，γ_tThe value of beta is related to the random error of the gyroscope, and because the gyroscope has high dynamic characteristic, the problems of zero drift and low-frequency interference exist during temperature change and movement.

5) And (2) updating the quaternion through an equation (17), wherein the quaternion represents real-time attitude information of the object, the unit quaternion can represent three-dimensional transformation, and the Euler angle of the motion of the object can be obtained from the quaternion. According to formula (1):

therefore, the lower limb wearing equipment calculates the spatial angle information of the lower limb in 6 degrees of freedom in real time, and can quickly estimate the motion posture of the lower limb.

The power supply module 5 supplies power to the sensing module 2, the main control module 3 and the communication module 4.

The upper computer is a computer host, and the main control module 3 of the wearable device has limited computing power, so that the lower limb movement tracking needs the upper computer to provide computing power assistance so as to obtain more accurate lower limb movement posture information.

The human knee joint movement obeys a hinge model, the hip joint belongs to a spherical joint, the movement state of the lower limb can be defined by 4 degrees of freedom, and the sensing module 2 arranged near the thigh femur and the shank tibia parallelly acquires the movement information of 6 axes. For ease of description, the following discussion will first be made from the case of a single leg of a human body, based on which such a mathematical model is constructed: taking the joint of the femur and the hip bone as the origin of a coordinate system of the lower limb and the length of the thigh and the shank as a system parameter l_uAnd l_lEstimating the position of the knee in a three-dimensional rectangular coordinate system to finish the tracking of the lower limb movement posture, and the method mainly comprises the following steps:

1) medically, the range of angles of motion of the lower limb joint is limited. The upper computer firstly carries out primary screening on the motion calculation result transmitted by the wearable equipment, deletes abnormal data and carries out linear interpolation according to normal data of front and back time points. Because the knee joint movement follows a hinge model, the yaw angle and the roll angle acquired by the inertial sensors at the positions of the thigh and the shank are averaged, and 4 degrees of freedom of the lower limb movement are expressed as a space angle: { theta ]₁，θ₂，θ₃，θ₄}，θ₁、θ₂、θ₃Respectively thigh femur pitch angle, roll angle, yaw angle theta₄Is the calf shank pitch angle.

2) The posture of the whole leg joint is output using the D-H transformation matrix.

In the formula: | loc_knee||＝l_u，loc_kneeIs the knee position obtained by D-H transformation. Similarly, the relative position of the ankle with respect to the knee can also be calculated, and further, the vector sum of the knee position and the ankle relative position is the absolute position of the ankle:

loc_{ankle-to-knee}＝g(θ₁，θ₂，θ₃，θ₄) (20)

loc_ankle＝loc_knee+loc_{ankle-to-knee} (21)

likewise, the formula (20) satisfies | | | loc_{ankle-to-knee}||＝f_l。

3) Velocity is the first derivative of the displacement of the object's motion, and acceleration information at the knee is calculated.

The compounds of formulas (20) and (21) are carried:

in the formula: acecel_l，tIs the acceleration value measured by the inertial sensor at the lower leg, Rot is the rotation matrix converted to the lower limb coordinate system.

4) And predicting the position of the knee by using an extended Kalman algorithm, wherein the state comprises the position of the knee and the speed of the knee. The Kalman filtering predicts the current state through the optimal value of the last state, namely:

X(k|k-1)＝A·X(k-1|k-1)+B·U(k) (24)

wherein A and B are system parameters. X (k | k-1), and U (k) is the control variable of the current state, and if not, may be 0, using the result of the last state optimal prediction value X (k-1| k-1). The system results are updated by equation (1), and the covariance corresponding to X also needs to be predicted:

P(k|k-1)＝A·P(k-1|k-1)A^T+N (25)

where P (k | k-1) is the covariance corresponding to X (k | k-1) and represents the uncertainty of the state estimate, P (k-1| k-1) is the covariance corresponding to X (k-1| k-1), A^TIs the transpose of a and N is the noise of the system.

After the system state quantity and the covariance are predicted, the estimation of a reference measurement value is carried out:

X(k|k)＝X(k|k-1)+k_g(k)·(Z(k)-H·X(k|k-1)) (26)

in the formula k_g(k) Is the kalman gain at time k, H is a parameter of the measurement system, and z (k) is a state measurement. To implement recursion, k at a time_gAll are updated in real time:

k_g(k)＝P(k|k-1)·H^T·(H·P(k|k-1)·H^T+∑)^-1 (27)

P(k|k)＝(1-k_g(k)·H)·P(k|k-1) (28)

each time P (k | k) and k_g(k) The value of the previous moment is needed to update, and sigma is the uncertainty of the observation distribution and combines the historical optimal state quantity, the uncertainty and the current measurement state.

Therefore, the knee position is predicted, and the spatial motion posture of the leg is accurately tracked by combining the spatial angle.

The communication module 4 is a wireless communication module or a wired serial communication module and is used for exchanging information with an upper computer.

The power supply module 5 comprises a charging and voltage-stabilizing discharging power panel and a lithium battery;

the voltage of the lithium battery is stabilized by the voltage stabilizing module and then supplies power to the sensing module 2, the main control module 3 and the communication module 4 through the connecting band 6.

The nine-axis inertial sensor can move within the range of human lower limb joint kinematics; the six modules are detachable.