A motion monitoring method, system, and wearable device

Abstract

This invention provides a motion monitoring method, system, and wearable device. It collects acceleration and angular velocity data during a patient's movement using an inertial measurement unit (IMU). The collected data is filtered and subjected to variational mode processing to extract gait information. Utilizing the waveform stability characteristics of the accelerometer and gyroscope at rest, the zero-velocity angular velocity point is identified. Based on the zero-velocity angular velocity point in the gyroscope data, the zero-velocity acceleration segment is also identified. Subsequently, the gyroscope's attitude is corrected using the acceleration data. The corrected gyroscope data generates a rotation matrix to transform the IMU coordinate system to the navigation coordinate system, and finally, trajectory reconstruction is performed. During trajectory reconstruction, data such as gait variability, step count, step frequency, stride length, movement distance, and movement trajectory are displayed on the wearable device or transmitted to a remote device for reference by doctors and patients.

Description

Technical Field

[0001] This invention relates to the field of inertial navigation monitoring, and more particularly to a motion monitoring method, system, and wearable device. Background Technology

[0002] Heart failure (HF) is a syndrome caused by impaired systolic and / or diastolic function of the heart, which fails to adequately pump venous blood back to the heart, leading to venous congestion and insufficient arterial perfusion. This results in circulatory disorders. HF patients experience a decline in quality of life due to impaired cardiac function, with significantly limited daily activities, typically manifested as a decrease in exercise capacity and intensity. Traditional treatment for HF primarily relies on medication. However, exercise rehabilitation for chronic HF began internationally in the 1970s, and substantial evidence-based medicine has demonstrated its safety and effectiveness. Exercise rehabilitation can reduce mortality, decrease the frequency of hospitalizations, improve exercise tolerance and quality of life, thereby justifying medical costs. In 2013, the American College of Exercise Foundation (ACCF) / AHA Heart Failure Management Guidelines listed exercise rehabilitation as a Class IA recommendation for patients with chronic stable heart failure.

[0003] During hospitalization, heart failure patients typically undergo an initial assessment of their heart failure level / severity, such as a 6-minute walk test or elbow flexion test. Doctors then develop a corresponding exercise rehabilitation plan based on the patient's heart failure level. The amount of exercise performed by heart failure patients during hospitalization is monitored by doctors. Once a patient has recovered to a certain extent, they can be discharged and continue rehabilitation at home. Post-discharge exercise rehabilitation progress is mostly reported to doctors through patient self-reports. Doctors cannot accurately understand the patient's actual daily exercise activity; therefore, other methods are needed to more precisely assess the patient's post-discharge exercise status.

[0004] Several methods for monitoring daily movement using monitoring devices have emerged in the existing technology, with inertial navigation being the most common. Inertial navigation is often used for indoor navigation, and it reconstructs the trajectory using accelerometers and gyroscopes in the IMU. The accuracy of trajectory reconstruction depends on the accuracy of the accelerometer's step counting. Therefore, the accuracy of zero-velocity detection by the accelerometer is particularly important. Existing zero-velocity detection technology is easily affected by noise and motion scene interference, which can lead to inaccurate zero-velocity detection and step counting errors. In addition, if the gyroscope's attitude shifts during exercise while wearing the monitoring device and is not corrected, there may also be cumulative errors in trajectory reconstruction. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a motion monitoring system and wearable device to provide doctors with detailed motion monitoring parameters for heart failure patients. At the same time, it provides a motion monitoring method that improves the accuracy of inertial navigation trajectory reconstruction and motion parameter monitoring through template matching zero-velocity detection, angular velocity and acceleration data fusion and attitude correction.

[0006] To solve the above-mentioned technical problems, the present invention specifically adopts the following technical solution:

[0007] A motion monitoring method includes the following steps:

[0008] Acquire data from the inertial measurement unit, including acceleration from the accelerometer and angular velocity from the gyroscope;

[0009] Kalman filtering was performed on both the angular velocity and acceleration data.

[0010] Extract the gait component from the angular velocity data and determine the minimum point, output the angular velocity data VPP, and filter out the zero angular velocity points that meet the conditions by setting a threshold for VPP;

[0011] Filter out the corresponding zero-velocity segment based on the zero-velocity point of angular velocity;

[0012] Output the corresponding zero-velocity acceleration segment based on the zero-velocity segment of angular velocity;

[0013] Acceleration and angular velocity data are fused to calculate attitude, and then the coordinate system of the inertial measurement unit is rotated and transformed to coincide with the navigation coordinate system;

[0014] The speed in the navigation coordinate system is output based on the acceleration. The speed in the zero-speed segment of acceleration is set to zero, the speed drift rate of the non-zero speed segment is calculated, the drift speed of the non-zero speed segment is calculated based on the speed drift rate, and then the final speed is obtained.

[0015] Furthermore, the extraction of gait components from angular velocity data and the determination of minimum points specifically involves:

[0016] The result after Kalman filtering is input into the Butterworth low-pass filter for further low-pass filtering, and the minimum point is extracted from the low-pass filtered data.

[0017] Furthermore, variational mode decomposition is performed on the low-pass filtered data, specifically as follows:

[0018] Gait component extraction is performed using variational mode decomposition. First, the filtered angular velocity data is decomposed into n components, with the constraint that the sum of all modes is equal to the original data. The constraint variational expression is as follows:

[0019]

[0020] Where k is the number of modes to be decomposed, {u k} and {w k} correspond to the k-th modal component and the center frequency after decomposition, respectively, and δ(t) is the Dirac function;

[0021] Secondly, gait components are extracted by gaining the gait signal components. The augmented expression is as follows:

[0022]

[0023] Here, α is a quadratic penalty factor to reduce the interference of Gaussian noise.

[0024] As another parallel approach to extracting gait components, the extraction of gait components from angular velocity data and determination of minimum points can specifically include:

[0025] SSA decomposition was performed on the angular velocity data after Kalman filtering, and the first component was used for data reconstruction.

[0026] Furthermore, the step of selecting the corresponding zero-angular-velocity segment based on the zero-angular-velocity point specifically involves:

[0027] The zero-velocity point is mapped to the angular velocity data after Kalman filtering and determined as the zero-velocity reference point. A sample interval is taken with N sampling points to the left and right of the zero-velocity reference point as the center. Several matching segments are taken in the sample interval in a sequential search manner. The matching segments are matched with the preset zero-velocity angular velocity segment template to obtain a similarity score. The matching segment with the highest similarity score is taken as the best matching segment. This best matching segment is the zero-velocity angular velocity segment. The interval of the zero-velocity angular velocity segment is mapped to the acceleration data and determined as the zero-velocity acceleration segment.

[0028] Furthermore, the process of fusing acceleration and angular velocity data to perform attitude calculation, thereby rotating and transforming the inertial measurement unit coordinate system to coincide with the navigation coordinate system, specifically involves:

[0029] The gyroscope offset factor is output by taking the acceleration value on the inertial measurement unit and the gravity vector value on the accelerometer from the accelerometer output. The gyroscope attitude is corrected by the gyroscope offset factor, and a quaternion rotation matrix is ​​generated based on the corrected gyroscope measurement value to rotate and transform the inertial measurement coordinate system to coincide with the navigation coordinate system.

[0030] Furthermore, the calculation of the velocity drift rate of the non-zero speed segment, the calculation of the drift velocity of the non-zero speed segment based on the velocity drift rate, and the subsequent obtaining of the final step speed are as follows:

[0031] First, calculate the velocity in the navigation coordinate system;

[0032] Next, calculate the velocity drift rate in the non-zero velocity segment:

[0033]

[0034] Further calculation of drift velocity: v drift,i =t i *rate drift ,i∈[t start ,t end The final output is the final speed after drift elimination.

[0035] Furthermore, the trajectory is reconstructed based on the final velocity, and the movement distance, trajectory, step length, and movement time are output.

[0036] The steps, step frequency, and gait variability are output based on the Z-axis component of the final velocity in the navigation coordinate system.

[0037] On the other hand, the present invention also provides a motion monitoring system, comprising:

[0038] An inertial measurement unit, comprising an accelerometer and a gyroscope;

[0039] The filtering unit performs filtering processing on the data collected by the inertial measurement unit;

[0040] The zero-velocity detection unit detects the zero-velocity point of angular velocity and the zero-velocity segment of acceleration measured by the inertial measurement unit.

[0041] The attitude calculation unit performs attitude calculation and correction on the gyroscope;

[0042] Coordinate system transformation unit, rotation transformation of the inertial measurement unit coordinate system;

[0043] A trajectory reconstruction unit reconstructs the motion trajectory of the motion monitoring system.

[0044] And an output module that outputs the motion distance, motion trajectory, step length, motion time, number of steps, step frequency and gait variability obtained from the trajectory reconstruction unit.

[0045] In another aspect, the present invention provides a computer-readable medium having stored thereon instructions executable by a processor, which, when executed, cause the processor to perform the motion monitoring method described above.

[0046] In another aspect, the present invention provides a wearable device equipped with the motion monitoring system described above, for implementing the motion monitoring method described above.

[0047] Furthermore, the wearable device is a wearable device worn on the feet of a person.

[0048] The beneficial effects of this invention are that, after filtering, the gyroscope angular velocity data extracts gait components and minimum points in the waveform, effectively removing repeated peak noise interference and enhancing the gait components. Subsequently, by setting a threshold for the VPP (velocity point deviation), waves below the VPP threshold are considered interference waves, further removing noise from the original gait signal and ensuring the accuracy of the subsequent angular velocity zero-velocity point selection. When selecting the zero-velocity acceleration segment, since the gyroscope waveform is more stable than the accelerometer waveform when stationary, the angular velocity data filtering waveform transformation and gait component enhancement are more efficient. The zero-velocity point of angular velocity is extracted, and the zero-velocity segment of angular velocity is filtered out based on the zero-velocity point of angular velocity and then mapped to acceleration, which makes the reliability of the zero-velocity segment of acceleration higher. Furthermore, when filtering the zero-velocity segment of angular velocity, a morphological matching method is adopted in a fixed interval using matching segments and preset templates. Compared with the existing technology of directly setting the zero-velocity segment threshold for acceleration data for filtering, it can avoid the situation where the zero-velocity segment of acceleration is too high or even exceeds the threshold due to the IMU being worn too loosely or the feet not being completely still when in contact with the ground, thus avoiding the situation where the zero-velocity segment is missed due to excessive interference or even exceeding the threshold. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A schematic diagram of human gait corresponding to the zero-speed segment and the non-zero-speed segment;

[0051] Figure 2 This is a waveform segment of angular velocity data after Kalman filtering.

[0052] Figure 3 This is a waveform segment of angular velocity data after being filtered by a Butterworth low-pass filter.

[0053] Figure 4 The waveform segment is a diagram of angular velocity data after variational mode decomposition and extraction of gait components.

[0054] Figure 5 This is a waveform segment diagram of angular velocity data after SSA decomposition and reconstruction of the first component data.

[0055] Figure 6 To select waveform segments of the zero-velocity acceleration segment based on the zero-velocity point of angular velocity;

[0056] Figure 7A diagram showing the zero-velocity segment of acceleration data;

[0057] Figure 8 for Figure 7 A larger image of a portion of the waveform segment from the zero-velocity segment marker diagram of the medium-accelerometer data;

[0058] Figure 9 Here is a flowchart of the gyroscope attitude correction process;

[0059] Figure 10 This is a segment of the velocity waveform of the inertial measurement unit in the navigation coordinate system;

[0060] Figure 11 This is a waveform segment of the displacement of the inertial measurement unit in the navigation coordinate system;

[0061] Figure 12 This is the trajectory diagram after the inertial measurement unit trajectory reconstruction;

[0062] Figure 13 This is a flowchart illustrating a motion monitoring method according to the present invention. Detailed Implementation

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. For example, terms such as “length,” “width,” “upper,” “lower,” “left,” “right,” “front,” “rear,” “vertical,” “horizontal,” “top,” “bottom,” “inner,” and “outer” indicate orientations or positions based on the orientations or positions shown in the accompanying drawings and are for ease of description only, and should not be construed as limiting the technical solution.

[0064] The terms "comprising" and "having," and any variations thereof, used in the specification, claims, and accompanying drawings of this invention are intended to cover non-exclusive inclusion; the terms "first," "second," etc., used in the specification, claims, and accompanying drawings are used to distinguish different objects, not to describe a particular order. "A plurality of" means two or more, unless otherwise explicitly specified.

[0065] In the description and claims of this invention and the foregoing drawings, when an element is referred to as "fixed to," "mounted to," "disposed on," or "connected to" another element, it can be located directly or indirectly on that other element. For example, when an element is referred to as "connected to" another element, it can be directly or indirectly connected to that other element.

[0066] Furthermore, the reference to "embodiment" herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0067] Example

[0068] This embodiment describes a motion monitoring method for monitoring a person's walking state, path, etc., such as... Figure 13 As shown, the specific steps include:

[0069] Data is acquired from the inertial measurement unit, including acceleration from the accelerometer and angular velocity from the gyroscope. The inertial measurement unit is specifically a 6-axis or 9-axis IMU. For ease of illustration, only a portion of the data segments are shown in this embodiment and the accompanying drawings. The actual data segments acquired can be as long as several hours or as short as a few seconds, as in this embodiment.

[0070] Kalman filtering was performed on both the angular velocity and acceleration data. Specifically, for the angular velocity data, the magnitude of the angular velocity data was normalized. The normalized data is then subjected to Kalman filtering to remove glitches and noise from the waveform. The waveform after Kalman filtering is as follows: Figure 2 As shown, it can be seen that, with Figure 6 Comparison of zero velocity segment of medium acceleration Figure 2 The waveform stability of the zero-velocity segment is higher than that of acceleration, making it more suitable for screening the zero-velocity segment. This method requires defining a specific interval centered on the minimum point of angular velocity to screen the zero-velocity segment. Figure 2 There are still many disturbances in the zero-velocity segment of the angular velocity. In order to find the lowest point with the most reference value, the angular velocity data needs to be further processed to extract the lowest point of angular velocity.

[0071] As one optional embodiment, the angular velocity data is further filtered based on the Kalman filter in order to find the minimum point of the angular velocity data: the result of the Kalman filter is input into the Butterworth low-pass filter for further low-pass filtering, and the specific design of the filter is as follows:

[0072] The differential equation describing the LTI system is:

[0073] a n y n (t)+a n-1 yn-1 (t)+...+a1y'(t)+a0y(t)=b m x m (t)+b m-1 x m-1 (t)+...+b1x'(t)+b0x(t);

[0074] The complex frequency domain equation is:

[0075] [a n S n +a n-1 S n-1 +...+a1S+a0]Y zs =[b m S m +b m-1 S m-1 +...+b1S+b0]X(s);

[0076] The system function of the continuous LTI system is as follows:

[0077] Analog filters: Digital filters:

[0078] The principle of the Butterworth filter:

[0079]

[0080] Attenuation function: A(w) = -20log|H(jw)|

[0081] Maximum passband attenuation: A p =-20log|1-δp|

[0082] Minimum stopband attenuation: A c = -20log|δc|

[0083] Calculation of filter order:

[0084] Calculate the normalized cutoff frequency: w n =(2*w c ) / (f sample )

[0085] Among them, w p Indicates the passband cutoff frequency, w c δp represents the stopband cutoff frequency, δc represents the passband ripple, and δc represents the stopband ripple.

[0086] The angular velocity waveform after passing through the Butterworth low-pass filter is as follows: Figure 3 As shown;

[0087] Next, further noise removal is needed. Variational mode decomposition (VMD) is used to extract the gait component from the angular velocity data. The principle of VMD is to decompose the original signal into multiple components, giving gain to the desired components and suppressing or discarding the unwanted components. In this embodiment, the components requiring gain are the gait components. To ensure that the decomposed sequence consists of modal components with a finite bandwidth and a center frequency, and that the sum of the estimated bandwidths of each mode is minimized, the constraint condition is that the sum of all modes is equal to the original signal. Therefore, the corresponding constraint variational expression is:

[0088]

[0089]

[0090] Where K is the number of modes to be decomposed (a positive integer), {u k},{w k} correspond to the k-th modal component and center frequency after decomposition, respectively, and δ(t) is the Dirac function. Then, the Lagrange multiplication operator is introduced to transform the constrained variational problem into an unconstrained variational problem, resulting in the augmented Lagrange expression:

[0091]

[0092] The quadratic penalty factor is used to reduce the interference of Gaussian noise. After gait extraction by variational mode decomposition, most of the recurrent peaks in the angular velocity data have been suppressed, retaining the most basic gait components.

[0093] As an optional embodiment, the Kalman filtered result is decomposed using SSA and the first component is extracted for data reconstruction. This can also extract the minimum points of the angular velocity data. Specifically, in this embodiment, the angular velocity data is decomposed into 26 components in time series, each component having a feature value. The sum of all feature values ​​is the total feature value of the angular velocity data. Among these feature values, the component with the largest proportion of feature values ​​is identified as the first component, i.e., the principal component. This principal component is considered to be the gait component. Data reconstruction is performed using this component, and finally, all the minimum points in the reconstructed data are extracted. The reconstructed data is as follows: Figure 5 As shown.

[0094] When a person walks, the posture and amplitude of each step are similar and do not change significantly in the short term. However, some waveforms with significantly lower VPP (Vacuum Step Plexus) still appear in the angular velocity waveform after gait component extraction. These waveforms are likely not obtained from taking a single step, but may be noise waves caused by unsteady gait or IMU (Installation Unit) fluctuations. By setting a threshold for VPP, the zero-velocity points of angular velocity that meet the criteria are screened out to further eliminate noise interference. Waveforms with VPP above the threshold are retained as normal steps, while waveforms with VPP below the threshold are removed as noise. The threshold is set as an empirical value, and different thresholds should be set for different populations. For example, when a healthy young person walks, the height of the step and the degree of flexion are greater than those of an elderly person with heart failure. Therefore, for elderly people with heart failure, a lower threshold should be set for noise removal. All the troughs in the angular velocity data after noise removal are the zero-velocity points of angular velocity during walking. Figure 4 As shown, Figure 4 The trough points marked in the data are the zero velocity points. Since these trough points are the minimum points of the angular velocity data after noise interference removal and gait component extraction as described above, it is most meaningful to use these points as the center to determine the zero velocity segment matching interval for subsequent processing.

[0095] When an IMU is worn on the foot, it can clearly capture and reflect the movement characteristics of the foot, such as... Figure 1The right foot in the IMU is in the zero-velocity segment from when the right foot fully touches the ground until the next time the right foot is lifted. The non-zero-velocity segment is from when the right heel leaves the ground until the next time the right foot fully touches the ground. For the IMU, due to the synchronicity of kinematics and data acquisition, there is a corresponding relationship between the trough of angular velocity data and the zero-velocity segment of accelerometer data. That is, the point with the lowest angular velocity during each step is usually located at a point in the time between when the current foot fully touches the ground and before the next time the current foot is lifted, which is also a point in the zero-velocity segment of acceleration. Accelerometers are used to measure acceleration, and gyroscopes are used to measure angular velocity. Since the waveform of a gyroscope is more stable than that of an accelerometer when stationary, the reliability of the zero-velocity angular velocity segment is higher by using the zero-velocity angular velocity point as the standard to filter the zero-velocity angular velocity segment and then mapping it to the acceleration. Specifically, the zero-velocity angular velocity point is mapped to the angular velocity data after filtering in step S1, as described above, and determined as the zero-velocity reference point. This zero-velocity reference point must be any point within the zero-velocity angular velocity segment. Using this zero-velocity reference point as the center, a sample interval is taken with a sample length of 70 sampling points on both the left and right sides, and the length of this sample interval is 140 sampling points. Within the sample interval, several matching segments are taken in a sequential search manner. The matching segments are then morphologically matched with the preset zero-velocity angular velocity segment template, and a similarity score is output. The matching segment with the highest similarity score is selected as the best matching segment. The interval of the best matching segment is mapped to the acceleration data and determined as the zero-velocity segment of acceleration. The length of the specific sample interval can be determined by those skilled in the art based on experience. In principle, the length of the sample interval should be greater than or equal to the time of walking one step to ensure that a complete zero-velocity segment can be found within the sample interval. The length of the matching segment can be determined by those skilled in the art or it can be equal to the length of the preset template segment. Preferably, multiple templates can be preset, including templates with spur noise signals in the zero-velocity segment, to match each zero-velocity segment as accurately as possible. These templates are derived from the zero-velocity segments of typical angular velocity data collected in the past. The template shape is usually "U" shaped, that is, the zero-velocity segment of each step. In summary, using the most reliable angular velocity zero-velocity point, the optimal matching segment is selected through morphological similarity template matching to identify the acceleration zero-velocity segment. Compared to some existing technologies that directly set thresholds for filtering, this method can filter out more zero-velocity segments that are easily missed. For example, if the IMU is not worn tightly or the pace is too fast, causing large fluctuations in IMU data, some waveforms that should be in the zero-velocity segment may exceed the preset threshold. Threshold filtering would result in missed detections, while morphological detection can avoid such problems. Figure 6 The image shows a waveform segment used to filter out the zero-velocity acceleration segment from the zero-velocity angular velocity point. It can be seen that even if the waveform of acceleration (or angular velocity) shows drift, the zero-velocity acceleration segment can still be accurately identified. Figure 7 This is a 0-1 stationary plot of acceleration at zero velocity, where flag zupt=1 indicates zero speed range, flag zupt =0 represents the non-zero velocity range; an enlarged view of the 0-1 stationary diagram is shown below. Figure 8 As shown.

[0096] Trajectory reconstruction using an IMU involves accumulating the displacement of the inertial measurement unit in the navigation coordinate system. Since the IMU's coordinate system does not coincide with the navigation coordinate system, this embodiment requires transforming the IMU's coordinate system to coincide with the navigation coordinate system before trajectory reconstruction. In this embodiment, acceleration and angular velocity data are first fused for attitude calculation to correct gyroscope drift errors. The accelerometer in the IMU measures acceleration, and the gyroscope measures angular velocity. Due to their kinematic characteristics, the accelerometer is highly stable when static (less prone to drift), but its data is relatively unreliable during motion. The gyroscope, on the other hand, has better dynamic stability, but its data is relatively unreliable when stationary (prone to drift). Therefore, the output of the accelerometer when stationary can be used to correct the gyroscope drift error.

[0097] Specifically, the gravity vector in the IMU coordinate system is first calculated using a quaternion rotation matrix. Then, the gyroscope offset factor is calculated based on the gravity vector value in the IMU coordinate system and the accelerometer value. Specifically, let the standard gravitational acceleration in the navigation coordinate system n be g, defined as g... n =[0,0,1] T The quaternion q = [q] represents rotation. w Using Hamiltonian convention, the rotation matrix [q1, q2, q3] is (from local to global):

[0098]

[0099] Since the rotation matrix is ​​an orthogonal matrix, therefore:

[0100]

[0101] Therefore, when at rest, the value of the gravity vector in the IMU coordinate system b is:

[0102]

[0103] Furthermore, the gyroscope's offset factor:

[0104]

[0105] Where θ is the offset angle between the gyroscope measurement and the projection of the gravity vector in the IMU coordinate system b. As long as this offset angle is not 0, the gyroscope measurement is corrected. The aforementioned rotation matrix is ​​then generated using the corrected gyroscope measurements. The gyroscope can ensure accurate attitude during coordinate system transformation. The attitude correction flowchart is as follows: Figure 9 As shown.

[0106] The inertial measurement unit coordinate system is then rotated and transformed to coincide with the navigation coordinate system:

[0107]

[0108] After the coordinate system transformation is complete, output the velocity in the navigation coordinate system based on the acceleration: Since trajectory reconstruction is performed by integrating velocity to obtain displacement, noise signals in the zero-velocity segment will cause cumulative errors in the displacement. Therefore, before trajectory reconstruction, the velocities in the zero-velocity segment are first set to zero according to the 0-1 stationary diagram mentioned above. For non-zero-velocity segments, drift errors still exist, and these drift errors will accumulate. To eliminate drift errors in non-zero-velocity segments, the velocity drift rate of the non-zero-velocity segment is calculated. Where V end Let t be the velocity before the end of each step. start and t end Given the start and end times of each step, the drift rate is obtained, and the drift speed for each step is output: v. drift,i =t i *rate drift ,i∈[t start ,t end The final output is the final velocity after drift elimination; for example... Figure 10 The figure shows the velocity waveform after drift elimination. It can be seen that the subject's velocity has larger components in the x and y directions in the navigation coordinate system and no reverse direction, while the component in the z direction is smaller and the integral result at each step approaches 0. This is because the subject is walking on a plane, and in reality, displacement only occurs in the x and y planes. The displacement result obtained after velocity integration is as follows... Figure 11 The subject's path map after accumulating displacement and reconstructing the trajectory is as follows: Figure 12 .

[0109] Those skilled in the art should note that the above-described steps of this invention are not limited in their specific order. For example, this invention is not limited to performing zero-velocity segment detection first, followed by IMU coordinate system transformation, and the order is not important. That is, this invention may also first transform the IMU to coincide with the navigation coordinate system, and then perform zero-velocity segment detection of acceleration.

[0110] Furthermore, the present invention can extract the required motion parameters during the inertial navigation trajectory reconstruction process in the above embodiments. For example, based on the trajectory reconstruction results, the subject's stride length, movement distance, movement trajectory, and movement scenario (going upstairs, going downstairs, walking on flat ground) can be output. Based on the Z component of the velocity, the number of steps, step speed, step frequency, and gait variability can be output. The gait variability mentioned here refers to the change in gait over a period of time. For example, the degree of change in the subject's step speed and step length over 30 minutes. Elderly people with heart failure may experience a decrease in walking speed, difficulty lifting their legs, and a decrease in step length after walking for a period of time due to a decline in exercise capacity.

[0111] On the other hand, as an optional embodiment, the present invention provides a motion monitoring system, which includes an inertial measurement unit, comprising an accelerometer and a gyroscope. In addition, the motion monitoring system further includes:

[0112] The filtering unit performs filtering processing on the data collected by the inertial measurement unit;

[0113] The zero-velocity detection unit detects the zero-velocity segment of acceleration measured by the accelerometer.

[0114] The attitude calculation unit performs attitude calculation and correction on the gyroscope;

[0115] Coordinate system transformation unit, rotation transformation of the inertial measurement unit coordinate system;

[0116] A trajectory reconstruction unit reconstructs the motion trajectory of the motion monitoring system.

[0117] And an output module that outputs the motion distance, motion trajectory, step length, motion time, number of steps, step frequency and gait variability obtained from the trajectory reconstruction unit.

[0118] This motion monitoring system is primarily used to implement the motion monitoring methods described above.

[0119] On the other hand, as an alternative embodiment, the present invention provides a computer-readable medium having instructions stored thereon that can be executed by a processor, wherein when the instructions are executed, the processor performs the motion monitoring method described above.

[0120] On the other hand, as an optional embodiment, the present invention provides a wearable device equipped with the motion monitoring system described in the above embodiments to implement the motion monitoring method described in the above embodiments. In this embodiment, due to the physical characteristics of the inertial measurement unit, the wearable device is preferably a wearable device worn on the foot. It is also important to note that the data collected by the inertial measurement unit in this invention does not necessarily need to be processed and displayed as described above on the wearable device. The collected signals can also be sent to a remote or cloud-based server for processing and display as described above. In this case, the wearable device can be made thinner and lighter, and easier to wear. For example, the wearable device can send the collected acceleration and angular velocity data to a remote or cloud-based server, where zero-velocity detection, attitude calculation, coordinate system transformation, and trajectory reconstruction are performed as described above. The output motion distance, motion trajectory, step length, motion time, number of steps, step frequency, and gait variability are then displayed on a remote mobile app, web interface, or mini-program for patients and doctors to refer to. Doctors can also use the obtained data to assess and monitor the exercise capacity of heart failure patients or prescribe corresponding exercise prescriptions for them.

[0121] The above embodiments are preferred implementations of the present invention. In addition, the present invention can be implemented in other ways. Any obvious substitutions without departing from the concept of the present technical solution are within the protection scope of the present invention.

Claims

1. A motion monitoring method, characterized in that: Includes the following steps: Acquire data from the inertial measurement unit, including acceleration from the accelerometer and angular velocity from the gyroscope; Kalman filtering was applied to both the angular velocity and acceleration data. Extracting gait components from angular velocity data and determining local minima: Inputting the Kalman filtered results into a Butterworth low-pass filter for further low-pass filtering, and extracting local minima from the low-pass filtered data. Output angular velocity data VPP, and filter out the zero velocity points of angular velocity that meet the conditions by setting a threshold for VPP; Based on the zero angular velocity point, the corresponding zero angular velocity segment is selected: the zero angular velocity point is mapped to the angular velocity data after Kalman filtering, and a zero velocity reference point is determined. With the zero velocity reference point as the center, a sample interval is taken with N sampling points on the left and right as the sample length. Several matching segments are taken in the sample interval in a sequential search manner. The matching segments are matched with the preset zero angular velocity segment template to obtain a similarity score. The matching segment with the highest similarity score is taken as the best matching segment. This best matching segment is the zero angular velocity segment. Output the corresponding zero-velocity acceleration segment based on the zero-velocity segment of angular velocity; Acceleration and angular velocity data are fused to calculate attitude, and then the coordinate system of the inertial measurement unit is rotated and transformed to coincide with the navigation coordinate system; The speed in the navigation coordinate system is output based on the acceleration. The speed in the zero-speed segment of acceleration is set to zero, the speed drift rate of the non-zero speed segment is calculated, the drift speed of the non-zero speed segment is calculated based on the speed drift rate, and then the final speed is obtained.

2. The motion monitoring method according to claim 1, characterized in that: The extraction of gait components from angular velocity data and determination of local minima are specifically as follows: First, the filtered angular velocity data is decomposed into multiple components, with the constraint that the sum of all modes is equal to the original data. The constraint variational expression is as follows: Where k is the number of modes to be decomposed, {u k } and {w k } correspond to the k-th modal component and the center frequency after decomposition, respectively, and δ(t) is the Dirac function; Secondly, gait components are extracted by gaining the gait signal components. The augmented expression is as follows: Where α is the second-order penalty factor.

3. The motion monitoring method according to claim 1, characterized in that: Extracting the gait component from the angular velocity data and determining the minimum point is specifically as follows: SSA decomposition is performed on the angular velocity data after Kalman filtering. The first component is taken for data reconstruction, and the minimum point of the reconstructed data is extracted.

4. The motion monitoring method according to claim 1, characterized in that: The process of fusing acceleration and angular velocity data to perform attitude calculation, thereby rotating and transforming the inertial measurement unit coordinate system to coincide with the navigation coordinate system, specifically involves: The gyroscope offset factor is output by taking the acceleration value and gravity vector value in the inertial measurement unit coordinate system from the accelerometer output. The gyroscope attitude is corrected by the gyroscope offset factor, and a quaternion rotation matrix is ​​generated based on the corrected gyroscope measurement value to rotate and transform the inertial measurement coordinate system to coincide with the navigation coordinate system.

5. The motion monitoring method according to claim 1, characterized in that: The calculation of the velocity drift rate in the non-zero speed segment, the calculation of the drift velocity in the non-zero speed segment based on the velocity drift rate, and the subsequent obtaining of the final step speed are as follows: First, calculate the velocity in the navigation coordinate system; Next, calculate the velocity drift rate in the non-zero velocity segment: Further calculation of drift velocity: v drift,i =t i *rate drift ,i∈[t start ,t end ], The final output is the final velocity after drift elimination.

6. The motion monitoring method according to claim 1, characterized in that: The trajectory is reconstructed based on the final velocity, and the movement distance, trajectory, step length, and movement time are output. Based on the final velocity, output the number of steps, step frequency, and gait variability for each axis component in the navigation coordinate system.

7. A motion monitoring system, comprising an inertial measurement unit, wherein the inertial measurement unit includes an accelerometer and a gyroscope, characterized in that: The motion monitoring system also includes: The filtering unit performs filtering processing on the data collected by the inertial measurement unit; The zero-velocity detection unit detects the zero-velocity point of angular velocity and the zero-velocity segment of acceleration measured by the inertial measurement unit. The attitude calculation unit performs attitude calculation and correction on the gyroscope; Coordinate system transformation unit, rotation transformation of the inertial measurement unit coordinate system; A trajectory reconstruction unit reconstructs the motion trajectory of the motion monitoring system. And an output unit that outputs the motion distance, motion trajectory, step length, motion time, number of steps, step frequency and gait variability obtained from the trajectory reconstruction unit; The motion monitoring system is used to perform the motion monitoring method according to any one of claims 1-6.

8. A wearable device, characterized in that: The wearable device is equipped with the motion monitoring system as described in claim 7, and is used to implement the motion monitoring method described in any one of claims 1-6.

9. A computer-readable medium having stored thereon instructions executable by a processor, said instructions, when executed, causing the processor to perform the motion monitoring method as described in any one of claims 1-6.

Images