Human motion state recognition system and method based on motion parameter fusion

Abstract

The invention discloses a human motion state recognition system and method based on motion parameter fusion, and belongs to the technical field of wearable exoskeletons. According to the method, angle and angular velocity data of inertial sensors at thighs are monitored in real time, and self-adaptive switching between a carrying power assisting mode and a walking power assisting mode is achieved through inertial sensor data change rules, motion monitoring parameter design and cross validation of the inertial sensors on the two legs. In the carrying and lifting mode, the assistance starting point assists human hip joint opening based on the human hip joint opening trend, and human body force is reduced; in the walking mode, the power assisting starting point is based on the time point when the legs of the human body touch the ground, so that the human body pedaling torque is increased, the human body walking speed is increased, and the human body walking force generating torque is reduced.

Description

Technical Field

[0001] The present invention relates to a human motion state recognition system and method based on motion parameter fusion, belonging to the technical field of wearable exoskeletons. Background Art

[0002] In the field of wearable exoskeletons, monitoring of human movement intentions is particularly important. Waist-lifting exoskeletons are often used in logistics warehouses, logistics warehousing and other scenarios. At present, mainstream waist-lifting exoskeletons can only provide lifting assistance for the human body. With the gradual maturity of waist-lifting exoskeleton technology, it will be difficult to meet the needs of the usage scenario by only realizing lifting actions. At the same time, during the walking process of the human body, if the exoskeleton provides assistance to the human body in the leg support state, its assistance method is similar to the assistance method when lifting goods. Based on this principle, walking assistance can be achieved by identifying the human body's movement state. In this context, how to accurately provide different assistance torques to the human body depends on the exoskeleton's ability to perceive various movement states in common scenarios.

[0003] At present, there are also motion state recognition methods based on sensors: for example, Chinese patent CN108836342A discloses a recognition method based on the Inception V3 neural network model, which slices the time data and converts it into images, inputs the neural network for training to realize motion state recognition; Chinese patent CN108596074A uses the SVM model to realize the recognition of multiple motion modes. However, these methods have certain limitations. First, the real-time performance is poor, and some algorithms require complete gait slices to achieve recognition. Secondly, the integration with the control algorithm is not high, and the recognition accuracy is not ideal. Summary of the invention

[0004] The technical problem solved by the present invention is: to overcome the shortcomings of the prior art, and to propose a human motion state recognition system and method based on motion parameter fusion, which collects motion data through inertial sensors, identifies the current motion state and assistance trigger point, and provides effective data support for scientific assistance.

[0005] The technical solution of the present invention is:

[0006] A human motion state recognition system based on motion parameter fusion, comprising a controller, a left thigh outer side inertial sensor, a right thigh outer side inertial sensor, and a cable;

[0007] The left thigh outer side inertial sensor and the right thigh outer side inertial sensor are respectively placed on the outer side of the corresponding thigh of the user, with the horizontal position coinciding with the coronal plane of the human body and the specific axis perpendicular to the sagittal plane of the human body, to collect the angle and angular velocity data of the position in real time and transmit them to the controller via a cable;

[0008] The controller is installed at the center of the user's lower back and is provided with an internal inertial sensor and a control unit; a specific axis of the internal inertial sensor is perpendicular to the coronal plane of the human body, and the angle and angular velocity data of the position are collected in real time; the control unit identifies the user's current motion state based on the angle and angular velocity data collected by the internal inertial sensor, the left thigh outer side inertial sensor, and the right thigh outer side inertial sensor, monitors whether the power-assistance trigger point is reached, and sends a power-assistance signal to the control system after reaching the point.

[0009] Furthermore, the left thigh outer side inertial sensor, the right thigh outer side inertial sensor and the internal inertial sensor are products of the same model and have the same accuracy.

[0010] Furthermore, the control unit identifies the current motion state of the user, which includes a walking mode and a lifting mode, and the identification method is:

[0011] Based on the angle data collected by the internal inertial sensor, the left thigh outer inertial sensor, and the right thigh outer inertial sensor, two motion monitoring parameters are calculated:

[0012] C=K c (sin(θ R )+sin(θ L ))

[0013] D=K d (sin(θ R )-sin(θ L ))

[0014] In the formula, θ R is the rotation angle of the right thigh outer inertial sensor around the Z axis, θ L K is the rotation angle of the left thigh outer inertial sensor around the Z axis, c , K d is the proportional magnification factor; the motion monitoring parameter D represents the difference between the two legs’ motions, and the motion monitoring parameter C represents the consistency between the two legs’ motions;

[0015] If the motion monitoring parameter C is greater than the set threshold T, and the motion monitoring parameter D is within the range of ±0.2, it is determined that the user has left the walking state and entered the lifting assistance mode; and the lifting main force trigger point is monitored.

[0016] If the motion monitoring parameter C is less than the set threshold value T, and the motion monitoring parameter D is not within the range of ±0.2, it is determined that the user has entered the walking assistance mode.

[0017] Furthermore, if the user's current motion state is the lifting mode, the motion monitoring parameter AVC is calculated:

[0018]

[0019] In the formula, is the angular velocity of the inertial sensor on the outside of the right thigh around the Z axis, It is the angular velocity of the inertial sensor on the outside of the left thigh rotating around the Z axis. AVC indicates the consistency of the angular velocities of both legs. When AVC is lower than the threshold, the human body shows an opening trend, that is, the user starts to lift. It is determined that the power-assistance trigger point is reached at the current moment, and the power-assistance signal is sent to the user control system.

[0020] Furthermore, if the user's current motion state is walking mode, the power assist trigger point is selected at the moment after the user touches the ground. Before determining the touchdown point, the human body posture is pre-determined and a pre-assistance interval is set. When entering the pre-assistance interval, the power assist trigger point is monitored.

[0021] Furthermore, after the user lifts his leg, it enters the pre-assist interval, that is, when the Z-axis angle of the thigh on one side enters the positive half zone, the Z-axis angular velocity is monitored, and when the Z-axis angular velocity enters the negative half zone, it enters the pre-assist interval; wherein, when the leg is lifted forward of the human body, the Z-axis angle is positive; when the leg is swung forward of the human body, the Z-axis angular velocity is positive.

[0022] Furthermore, after entering the pre-assistance interval, the assist trigger point is monitored, and the monitoring method is:

[0023] The sliding window method is used to screen the Z-axis angular velocity peak of the current side thigh outer inertial sensor that enters the pre-assistance interval. When the peak value is reached, if the Z-axis angle of the other side thigh outer inertial sensor enters the negative half zone, and at this moment the Z-axis angle of the current side thigh outer inertial sensor is greater than the minimum assist angle set by the system, this moment is identified as the assist trigger point, enters the assist interval, and sends an assist signal to the control system.

[0024] Furthermore, if the Z-axis angle of the inertial sensor on the outer side of the current thigh enters the negative half zone or the Z-axis angles of the inertial sensors on the outer sides of the thighs on both sides enter the positive half zone, the pre-assist interval is exited.

[0025] The recognition method of the human motion state recognition system based on motion parameter fusion includes:

[0026] S1: After the user puts on the human motion state recognition system, he / she powers on in an upright standing position. After the system is powered on, the control unit collects the first 10 sets of data from each inertial sensor, takes the average value, and uses the angle data of the internal inertial sensor to zero-calibrate the initial angles of the left thigh outer side inertial sensor and the right thigh outer side inertial sensor, and corrects the initial angles of the left thigh outer side inertial sensor and the right thigh outer side inertial sensor.

[0027] S2: The control unit collects the three-axis angles and angular velocities of the three inertial sensors in real time and performs smoothing processing on each data to identify the user's current motion state;

[0028] S3: In the current motion state, the angle and angular velocity of the inertial sensor on the outer side of the left thigh and the inertial sensor on the outer side of the right thigh are monitored. When the power-assistance touchdown point is reached, a power-assistance signal is sent to the control system. When it is recognized that the power-assistance has ended, return to S2.

[0029] Furthermore, the assisting end recognition method is:

[0030] If the current motion state is the lifting mode, the human body is fully opened as the standard, and when the average Z-axis angle difference of the three inertial sensors is less than the threshold, it is determined that the human body has reached an upright state and the assistance ends;

[0031] If the current motion state is walking mode, the motion assistance of the current leg is stopped when the Z-axis angular velocity of the inertial sensor on the outer side of one thigh reaches 0, or when the Z-axis angle of the inertial sensor on the outer side of the other thigh is greater than 0, the motion assistance of the current leg is stopped.

[0032] The advantages of the present invention compared with the prior art are:

[0033] (1) The human motion sensing system and sensing method provided by the present invention require fewer inertial sensors, namely, a waist sensor and one inertial sensor each at the left and right thighs. Human posture recognition can be achieved by fusing and cross-validating the inertial sensor data.

[0034] (2) The present invention provides a human motion sensing system and sensing method, which generates motion parameters by fusing the monitored motion data. Dual-mode adaptive recognition can be achieved through parameter cross-validation, and two motion modes, walking and lifting, can be identified.

[0035] (3) A human motion sensing system and sensing method provided by the present invention. In walking mode, the judgment method mainly relies on the data of the leg sensor itself and the data of the opposite leg sensor for cross-verification, so as to realize the touchdown point judgment of the leg and provide assistance according to the touchdown point.

[0036] (4) The present invention provides a human motion sensing system and sensing method, which generates motion parameters by fusing the monitored motion data. In the lifting mode, the lifting mode can be entered by cross-validating the generated motion parameters, and the lifting trigger point can be identified.

[0037] (5) The human motion sensing system and sensing method provided by the present invention can be used for gaits in various terrains in walking mode through cross-validation of human motion data. It can identify common actions such as walking on flat ground, climbing stairs, and climbing slopes, and provide assistance; at the same time, it can screen out actions that do not require assistance, such as going down stairs and down slopes, and not provide assistance. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0039] Figure 1 This is a schematic diagram of the installation position of the human body motion state recognition system according to an embodiment of the present invention;

[0040] Figure 2 This is a flow chart of a method for identifying human motion status according to an embodiment of the present invention;

[0041] Figure 3 This is a flow chart of identifying the power-assistance triggering time point in the lifting state according to an embodiment of the present invention;

[0042] Figure 4 This is a flow chart of identifying the power assistance triggering time point in walking state according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0044] The present invention proposes a human motion state recognition system and recognition method based on motion parameter fusion, which can realize the determination of different motion modes and the recognition of power assistance trigger points for differentiated motion states according to the motion modes. At the same time, the walking assistance recognition method can also be used alone on a lower limb exoskeleton that mainly assists walking.

[0045] The installation location of the human motion state recognition system based on motion parameter fusion is as follows: Figure 1 As shown, the system includes a controller 1, a left thigh outer inertial sensor 3, a right thigh outer inertial sensor 4, a battery 5 and a cable 6. The controller is composed of a main control board, which includes a micro control unit, an internal inertial sensor 2 and a CAN communication module. The three inertial sensors are of the same model.

[0046] The controller 1 is installed at the center of the user's lower back, so that the specific axis of the inertial sensor 2 inside the controller is perpendicular to the human body's coronal plane, and the Y axis is selected to be perpendicular to the human body's coronal plane, so that the Z axis is perpendicular to the human body's sagittal plane, ensuring that the three sensors use the same precision axis of the same inertial sensor when rotating on the human body's sagittal plane.

[0047] The two inertial sensors on the outside of the left and right legs monitor the angle and angular velocity data of the inertial sensor at the thigh in real time. When the inertial sensor on the leg is installed, it is as close to the middle of the vastus lateralis muscle as possible in height, and the horizontal position coincides with the coronal plane of the human body. The specific axis of the inertial sensor is selected to be perpendicular to the sagittal plane of the human body. The axial selection is related to the accuracy of each axis of the inertial sensor and the data acquisition range of the inertial sensor. When the leg is raised forward to the human body, the angle is positive, and the angular velocity is positive when it swings forward to the front of the human leg. The data acquisition frequency is 100hz. Each set of data includes the three-axis angle, angular velocity and acceleration of the inertial sensor, which are transmitted through the CAN bus. The inertial sensor 3 on the outside of the left thigh and the inertial sensor 4 on the outside of the right thigh are connected to the controller through cables. The cable includes the power supply line and CAN transmission line of the inertial sensor.

[0048] In the motion state recognition system, the specific operation steps of the motion perception algorithm are as follows:

[0049] Step 1: After the human body wears the system, it is powered on in an upright standing position. After the system is powered on, the controller collects the first 10 sets of data from each inertial sensor, takes the average value, and uses the angle data of inertial sensor 2 to zero-calibrate the initial angles of inertial sensors 3 and 4 to correct the initial angles of inertial sensors 3 and 4.

[0050] Step 2: The controller will collect the three-axis angle, angular velocity and acceleration of the inertial sensor in real time, and smooth the data through sliding average filtering. After the human body starts to move, the motion mode is first identified through the upper-level control of the perception algorithm. The identified motion modes mainly include walking mode and lifting mode.

[0051] like Figure 2 The algorithm can be divided into two layers. The upper layer first determines the walking mode and the lifting mode. In the upper layer perception algorithm, a real-time data monitoring method is adopted, and two motion monitoring parameters are introduced:

[0052] C=K c (sin(θ R )+sin(θ L ))

[0053] D=K d (sin(θ R )-sin(θL ))

[0054] Among them, θ represents the angle of rotation of the inertial sensor around the Z axis, R represents the right side, and L represents the left side. K is the proportional magnification coefficient. According to the experimental results, the monitoring parameter D represents the degree of difference in the movement of the two legs, and the monitoring parameter C represents the degree of consistency of the movement state of the two legs. The monitoring parameter D is negatively correlated with the consistency of the movement of the two legs, and the monitoring parameter C is positively correlated with the consistency of the movement of the two legs. When the monitoring parameter C is greater than the threshold T and D is in the range of ±0.2, it is considered that the human body has left the walking state, and the perception system enters the lifting assistance mode and begins to monitor the main trigger point of the lifting. On the contrary, when the monitoring parameter C is less than the threshold T and D leaves the interval near 0, the perception system will enter the walking assistance mode.

[0055] Step 3: After completing the upper-level motion pattern recognition, the lower-level power-assist trigger point recognition will be entered. According to the motion pattern recognized by the upper-level perception algorithm, the specific recognition algorithm will be entered to identify the power-assist trigger time points of different motions. The sliding window data monitoring method will be used in the touchdown point recognition algorithm.

[0056] In lift mode, if Figure 3 As shown, new motion monitoring parameters are introduced:

[0057]

[0058] in Represents angular velocity. This method uses the Z-axis angular velocity. AVC represents the consistency of the angular velocity of both legs. When AVC is lower than the threshold, it can be considered that the human body is showing an opening trend, which means that the human body begins to lift. At this moment, the perception system determines that the power-assisting time point has been reached and sends the power-assisting signal to the control system.

[0059] In walking mode, the power assist trigger point is selected at the moment after the human body touches the ground. Figure 4As shown in the figure, in order to avoid other similar movements during human walking being mistaken for ground contact, this algorithm pre-determines the human posture before determining the touchdown point, sets the pre-assistance interval, and starts monitoring the power-assistance trigger point only after reaching the pre-assistance interval. The perception algorithm selects the human body to enter the pre-assistance interval after lifting the leg. The expression in the algorithm is that the angular velocity is monitored after the thigh Z-axis angle enters the positive half zone, and the pre-assistance interval is entered when the angular velocity enters the negative half zone. In the pre-assist interval, it is determined whether the assistance starting point has been reached by judging a variety of data. Taking the left leg as an example, the sliding window method is used to screen the peak values ​​of the data adjacent interval. When the Z-axis angular velocity of the inertial sensor 3 installed on the left leg reaches the peak value, if the Z-axis angle of the inertial sensor 4 installed on the right leg enters the negative half zone, and at this moment the Z-axis angle of the inertial sensor 3 is greater than the minimum assistance angle set by the system, it enters the assistance interval and sends an assistance signal to the control system; if the Z-axis angle of the inertial sensor 3 enters the negative half zone or the Z-axis angles of the inertial sensors 3 and 4 both enter the positive half zone, the pre-assist interval is exited and the pre-assist interval is re-determined. The judgment method for the right leg is opposite.

[0060] Step 5: After waiting for the control system to complete the corresponding assistance, the perception system determines the current state. If the assistance is terminated, it will return to the second step and enter the next motion recognition cycle. The perception system's recognition of the end of the movement is different in different motion modes, and it is also related to the control system. If the control system's assistance torque is based on the length of time, the control system will send a signal to the perception system after completing the assistance, which is used as a sign of the end of the motion assistance; the motion assistance can also be terminated by the perception system's monitoring of the human body posture. In the lifting mode, the human body is fully opened as the standard, and when the average value of the Z-axis angle difference between inertial sensor 2 and inertial sensors 3 and 4 is less than the threshold, it is considered that the human body has reached an upright state and the assistance interval is terminated, where the Z-axis angle difference refers to the Z-axis angle of the waist sensor minus the Z-axis angle of the leg sensor, and the angle difference represents the degree of bending of the human hip joint from the sagittal plane. In the walking mode, taking the left leg as an example, when the Z-axis angular velocity of the inertial sensor 3 reaches 0, the motion assistance of the left leg is stopped, or when the Z-axis angle of the inertial sensor 4 is greater than 0, the motion assistance of the left leg is stopped.

[0061] The present invention monitors the angle and angular velocity data of the inertial sensor at the thigh in real time, and realizes adaptive switching between lifting assistance and walking assistance modes through the change law of inertial sensor data, motion monitoring parameter design, specific data conditions and cross-validation of inertial sensors of both legs, and simultaneously realizes the positioning of the assistance starting point and assistance end point during lifting, walking, climbing stairs and climbing slopes. During the process of going down stairs and down slopes, these movements are screened out through a perception algorithm to achieve no assistance. In the lifting mode, the assistance starting point is based on the trend of the human hip joint opening, which assists the human hip joint to open and reduces the human body's force; in the walking mode, the assistance starting point is based on the time point when the human leg touches the ground, and the assistance process is to apply a torque to open the hip joint, so that the human body's pushing torque increases, thereby increasing the human body's walking speed and reducing the human body's own walking force torque.

[0062] The above-described embodiments are only preferred specific implementations of the present invention. Common changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included in the protection scope of the present invention.

Claims

1. A human motion state recognition system based on motion parameter fusion, characterized in that: Includes controller, left thigh outer inertial sensor, right thigh outer inertial sensor, and cables; The left thigh outer side inertial sensor and the right thigh outer side inertial sensor are respectively placed on the outer side of the corresponding thigh of the user, with the horizontal position coinciding with the coronal plane of the human body and the specific axis perpendicular to the sagittal plane of the human body, to collect the angle and angular velocity data of the position in real time and transmit them to the controller via a cable; The controller is installed at the center of the user's lower back and is provided with an internal inertial sensor and a control unit; a specific axis of the internal inertial sensor is perpendicular to the coronal plane of the human body, and the angle and angular velocity data of the position are collected in real time; the control unit identifies the user's current motion state based on the angle and angular velocity data collected by the internal inertial sensor, the left thigh outer side inertial sensor, and the right thigh outer side inertial sensor, monitors whether the power-assistance trigger point is reached, and sends a power-assistance signal to the control system after reaching the point.

2. The human motion state recognition system based on motion parameter fusion according to claim 1 is characterized in that: The left thigh outer inertial sensor, the right thigh outer inertial sensor and the internal inertial sensor are the same model products with the same accuracy.

3. The human motion state recognition system based on motion parameter fusion according to claim 1 is characterized in that: The control unit identifies the user's current motion state, which includes walking mode and lifting mode. The identification method is: Based on the angle data collected by the internal inertial sensor, the left thigh outer inertial sensor, and the right thigh outer inertial sensor, two motion monitoring parameters are calculated: C=K c (sin(θ) R )+sin(θ L )) D=K d (sin(θ) R )-sin(θ L )) In the formula, θ R is the rotation angle of the right thigh outer inertial sensor around the Z axis, θ L K is the rotation angle of the left thigh outer inertial sensor around the Z axis, c , K d is the proportional magnification factor; the motion monitoring parameter D represents the difference between the two legs’ motions, and the motion monitoring parameter C represents the consistency between the two legs’ motions; If the motion monitoring parameter C is greater than the set threshold T, and the motion monitoring parameter D is within the range of ±0.2, it is determined that the user has left the walking state and entered the lifting assistance mode; and the lifting main force trigger point is monitored. If the motion monitoring parameter C is less than the set threshold value T, and the motion monitoring parameter D is not within the range of ±0.2, it is determined that the user has entered the walking assistance mode.

4. The human motion state recognition system based on motion parameter fusion according to claim 1 is characterized in that: If the user's current motion state is lifting mode, calculate the motion monitoring parameter AVC: In the formula, is the angular velocity of the inertial sensor on the outside of the right thigh around the Z axis, It is the angular velocity of the inertial sensor on the outside of the left thigh rotating around the Z axis. AVC indicates the consistency of the angular velocities of both legs. When AVC is lower than the threshold, the human body shows an opening trend, that is, the user starts to lift. It is determined that the power-assistance trigger point is reached at the current moment, and the power-assistance signal is sent to the user control system.

5. The human motion state recognition system based on motion parameter fusion according to claim 1 is characterized in that: If the user's current motion state is walking mode, the power-assist trigger point is selected at the moment after the user touches the ground. Before determining the touchdown point, the human body posture is pre-determined and a pre-assist interval is set. When entering the pre-assist interval, the power-assist trigger point is monitored.

6. The human motion state recognition system based on motion parameter fusion according to claim 5 is characterized in that: After the user lifts his leg, it enters the pre-assist interval, that is, when the Z-axis angle of the thigh on one side enters the positive half zone, the Z-axis angular velocity is monitored. When the Z-axis angular velocity enters the negative half zone, it enters the pre-assist interval; when the leg is lifted forward, the Z-axis angle is positive; when the leg is swung forward, the Z-axis angular velocity is positive.

7. The human motion state recognition system based on motion parameter fusion according to claim 5 is characterized in that: After entering the pre-assistance range, the assist trigger point is monitored. The monitoring method is: The sliding window method is used to screen the Z-axis angular velocity peak of the current side thigh outer inertial sensor that enters the pre-assistance interval. When the peak value is reached, if the Z-axis angle of the other side thigh outer inertial sensor enters the negative half zone, and at this moment the Z-axis angle of the current side thigh outer inertial sensor is greater than the minimum assist angle set by the system, this moment is identified as the assist trigger point, enters the assist interval, and sends an assist signal to the control system.

8. The human motion state recognition system based on motion parameter fusion according to claim 7 is characterized in that: If the Z-axis angle of the inertial sensor on the outer side of the current thigh enters the negative half zone or the Z-axis angles of the inertial sensors on the outer side of the thighs on both sides enter the positive half zone, the pre-assist interval is exited.

9. The recognition method of the human motion state recognition system based on motion parameter fusion as claimed in claim 3 is characterized in that: include: S1: After the user puts on the human motion state recognition system, he / she powers on in an upright standing position. After the system is powered on, the control unit collects the first 10 sets of data from each inertial sensor, takes the average value, and uses the angle data of the internal inertial sensor to zero-calibrate the initial angles of the left thigh outer side inertial sensor and the right thigh outer side inertial sensor, and corrects the initial angles of the left thigh outer side inertial sensor and the right thigh outer side inertial sensor. S2: The control unit collects the three-axis angles and angular velocities of the three inertial sensors in real time and performs smoothing processing on each data to identify the user's current motion state; S3: In the current motion state, the angle and angular velocity of the inertial sensor on the outer side of the left thigh and the inertial sensor on the outer side of the right thigh are monitored. When the power-assistance touchdown point is reached, a power-assistance signal is sent to the control system. When it is recognized that the power-assistance has ended, return to S2.

10. The identification method according to claim 9, characterized in that: The assist end recognition method is: If the current motion state is the lifting mode, the human body is fully opened as the standard, and when the average Z-axis angle difference of the three inertial sensors is less than the threshold, it is determined that the human body has reached an upright state and the assistance ends; If the current motion state is walking mode, the motion assistance of the current leg is stopped when the Z-axis angular velocity of the inertial sensor on the outer side of one thigh reaches 0, or when the Z-axis angle of the inertial sensor on the outer side of the other thigh is greater than 0, the motion assistance of the current leg is stopped.

Images