CN109827568A - Pedestrian level location estimation method in tier building based on MEMS sensor - Google Patents
Pedestrian level location estimation method in tier building based on MEMS sensor Download PDFInfo
- Publication number
- CN109827568A CN109827568A CN201910086482.XA CN201910086482A CN109827568A CN 109827568 A CN109827568 A CN 109827568A CN 201910086482 A CN201910086482 A CN 201910086482A CN 109827568 A CN109827568 A CN 109827568A
- Authority
- CN
- China
- Prior art keywords
- height
- phase
- pedestrian
- data
- attitude
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Length Measuring Devices With Unspecified Measuring Means (AREA)
Abstract
本发明公开了基于MEMS传感器的多层建筑中行人高度位置估计方法，包括：将MEMS集成传感装置固定于行人下肢上；MEMS集成传感装置采集行人行走楼梯过程中的加速度数据、角速度数据及行人高度数据；采用多阈值检测方法确定出行人步态运动的姿态相和摆动相；对所述姿态相和摆动相的竖直方向加速度分别进行修正，且将姿态相的速度更新为零，对修正后的竖直方向加速度分别进行二次积分，得到行人姿态相和摆动相的估计高度；判断姿态相得到的估计高度数据和微气压计测量的目标高度数据之间是否存在偏差？如果是，则对摆动相得到的估计高度进行误差线性补偿，最终得到更新后的行人定位数据。本发明实现了基于一个传感器节点进行行人高度位置的精确估计。
The invention discloses a method for estimating the height position of a pedestrian in a multi-storey building based on a MEMS sensor, comprising: fixing a MEMS integrated sensing device on the lower limb of the pedestrian; Pedestrian height data; use multi-threshold detection method to determine the attitude phase and swing phase of pedestrian gait motion; correct the vertical acceleration of the attitude phase and swing phase respectively, and update the speed of the attitude phase to zero. The corrected vertical accelerations are respectively integrated twice to obtain the estimated heights of the pedestrian attitude phase and swing phase. Is there any deviation between the estimated height data obtained from the attitude phase and the target height data measured by the micro barometer? If it is, the error linear compensation is performed on the estimated height obtained by the swing phase, and finally the updated pedestrian positioning data is obtained. The present invention realizes accurate estimation of pedestrian height and position based on one sensor node.
Description
Technical Field
The invention relates to a pedestrian height position estimation method in a multi-story building based on an MEMS sensor, and belongs to the technical field of indoor navigation and positioning.
Background
With the development of micro-electro-mechanical systems (MEMS), small and low-cost sensors, such as accelerometers and gyroscopes, have been widely used in the fields of indoor positioning of pedestrians and limb tracking. An Inertial Measurement Unit (IMU) sensor for measuring the three-axis attitude angle (or angular rate) and acceleration of an object has inherent drift, so that a large error is generated when an accurate position is estimated, and the application of the MEMS sensor taking the sensor as a core technology in the aspects of pedestrian navigation, indoor pedestrian position estimation and the like is limited. Particularly when IMU sensors are used to estimate the height position of pedestrians in a multi-story building, it is difficult to achieve high-precision vertical height positioning.
There are two conventional methods for position estimation, one is Pedestrian Dead Reckoning (PDR) and the other is Inertial Navigation System (INS). In the PDR technique, the step count and step estimate are used to determine the location of a person. However, walking instability or inaccurate step size estimation can result in large errors. In INS, the pedestrian position is double-integrated estimated from the measured acceleration signal. To limit the increase in estimation error, many different techniques have been developed to assist the IMU sensor in error correction. Such as Radio Frequency Identification (RFID), map information, ultrasonic sensors, ground reaction sensors, optical sensors, ultra-wideband, pressure sensors, and the like. Although the above-described system may provide accurate position estimates, it involves cost and portability issues. In a position estimation system based on a MEMS sensor, methods such as Extended Kalman Filtering (EKF), robust adaptive kalman filtering (raff), zero velocity update (ZUPT) and the like are often used to reduce drift errors, and Heuristic Drift Reduction (HDR) and Zero Angular Rate Update (ZARU) methods are also proposed to reduce heading errors. These methods show high accuracy when tested on a horizontal straight path. In the vertical direction, the skilled person proposes methods based on multiple MEMS sensors to reduce the error of the height estimation, such as the Denavit-hartenberg (dh) method, which is built on the basis of wearing three sensor nodes on one lower limb, and the height difference information assisted barometer (HDIB) method, which is developed on the basis of wearing two sensor nodes on two different feet. Both methods utilize the relative motion of human joints to reduce the drift error of the IMU sensor, and the increase of the number of sensors leads to the deterioration of the portability of the detection system, so the research on the pedestrian height position estimation technology based on one sensor node is receiving general attention.
Disclosure of Invention
The invention aims to provide a pedestrian height position estimation method in a multi-story building based on an MEMS sensor, which can realize accurate estimation of the pedestrian height position based on one sensor node.
In order to solve the technical problems, the invention adopts the following technical scheme: the method for estimating the height position of the pedestrian in the multi-story building based on the MEMS sensor comprises the following steps of:
s1, fixing the MEMS integrated sensing device on the lower limb of the pedestrian; the MEMS integrated sensing device comprises: the three-axis accelerometer, the gyroscope and the barometer are respectively used for acquiring acceleration data, angular velocity data and pedestrian height data in the process of walking stairs of pedestrians;
s2, determining the posture phase and the swing phase of the gait motion of the pedestrian by adopting a multi-threshold detection method based on the acceleration data in the vertical direction;
s3, correcting the vertical direction acceleration of the attitude phase and the swing phase respectively, updating the speed of the attitude phase to zero, and performing quadratic integration on the corrected vertical direction acceleration respectively to obtain the estimated heights of the pedestrian attitude phase and the swing phase;
s4, it is determined whether there is a deviation between the estimated height data obtained from the attitude phase and the target height data measured by the barometer (the step height at each step measured by the barometer is defined as the target height)? And if so, carrying out error linear compensation on the estimated height obtained by the swing phase, and finally obtaining updated pedestrian positioning data.
Preferably, in step S2, the posture phase and swing phase of the gait motion of the pedestrian are determined by establishing the constraint conditions of three thresholds on the acceleration data in the gravity direction:
wherein,stands for verticalThe acceleration in the vertical direction, i is the number of sampling points, W is the size of a sliding window, g is the acceleration of gravity, and lambda1The variance threshold value represents the acceleration in the sliding window and is used for measuring the fluctuation degree of the acceleration data; x ═ a (τ, a)i+1-Ai) Y ═ t,0)), τ is the sampling period; lambda [ alpha ]2The threshold value is a normalized vertical acceleration change threshold value and is used for judging and determining a critical value of an attitude phase and a swing phase;represents the phase time of the current posture,representing the next attitude phase duration; lambda [ alpha ]3A threshold value representing the time length ratio between two adjacent attitude phases is used for screening wrong gait phases; determining the vertical direction acceleration data sequence which simultaneously meets the three threshold conditions as an attitude phase; the remaining data sequence is then the wobble phase.
The posture phase and the swing phase of the gait motion of the pedestrian are determined by the method, the motion state of the pedestrian is simplified, the accurate pedestrian height estimation is realized, and the motion data of the pedestrian can be corrected and analyzed conveniently under different running states. Some researchers use threshold detection methods based on inertial sensor multi-dimensional data and statistical analysis, as well as other gait phase detection techniques based on vision, etc. Compared with other prior art, the multi-threshold detection method only depends on vertical acceleration data, simplifies the motion state of the pedestrian, has high detection precision on the gait phase of upstairs going, walking and downstairs, and is simple and stable. And finally, high-precision correction of different stages of IMU data drifting is facilitated.
Preferably, the size of the sliding window W is 3 (data volatility can be better described); variance threshold lambda of acceleration in sliding window1Is 0.0096; normalized vertical acceleration change threshold λ2>0.04; threshold lambda of time length ratio between two adjacent attitude phases3Is 0.75. The values are set after a plurality of tests and comparisonsThe optimal parameters can more accurately detect the gait phases of the pedestrians going upstairs, walking and downstairs.
In the method for estimating the height position of the pedestrian in the multi-story building based on the MEMS sensor, the vertical direction accelerations of the attitude phase and the swing phase are respectively corrected in step S3, that is: updating the acceleration of the attitude phase in the vertical direction to zero; and expressing the motion direction of the swing phase by using a quaternion, carrying out data fusion on the acquired angular velocity and acceleration through a complementary filtering algorithm, and correcting the vertical acceleration with drift. By updating the acceleration to zero when the monopod is stationary, the attitude phase drift error can be minimized. In addition, the gyroscope has good dynamic characteristics, and the acceleration of the swing phase is corrected by utilizing a complementary filtering-based data fusion technology. Some researchers use data fusion technologies such as extended kalman filtering and particle filtering to realize the correction of the acceleration. The acceleration correction method adopted by the scheme starts from an actual physical model, combines gait phase detection and utilizes the complementary characteristics of data of a gyroscope and an accelerometer. A simpler and more reliable acceleration correction technology is provided for realizing accurate height estimation, so that the IMU data drift correction degree is higher.
Preferably, the target height data measured by the barometer in step S4 is: the method comprises the steps of respectively and sequentially carrying out preprocessing based on Lauda criterion, median filtering, sliding average filtering preprocessing and trend removing preprocessing on the original target height data of pedestrians measured by a barometer under an attitude phase in an accumulated mode, and then averaging the three processed attitude phase height data to obtain data. Through the preprocessing technology, noise and errors of the barometer height data caused by temperature, air pressure and system errors are removed, and accuracy of pedestrian height position estimation is finally improved.
In the foregoing method for estimating the height position of a pedestrian in a multi-story building based on MEMS sensors, step S4 further includes: under the scene that the target height of each step of pedestrians in the multi-story building is the same and unchanged, correcting the target height data by the following method: calculating the target height increment of each step and the mean value and standard deviation of the target height increment; and when the absolute value of the difference value between the target height increment and the average value in a certain step is more than one standard deviation, updating the target height increment based on the weighted average of the target height increments in the adjacent three steps. The constant and accurate target height can be obtained under the scene that the target height is not changed (under the normal condition, the target heights measured by the micro-barometer are not necessarily the same due to the influence of various factors when the actual target heights are the same), so that the accuracy of the final pedestrian height position estimation is further improved.
In the foregoing method for estimating the height position of a pedestrian in a multi-story building based on MEMS sensors, step S4 further includes: under the scene that the target heights of the pedestrians at each step in the multi-story building are different, improving the target height increment of each step by using the following formula, and further improving the target height data of each step:
wherein,increment, h, representing target height of step j1Height constants for the first step of the attitude phase (e.g. the target height value for the first step can be taken in general)Represents the target height of the jth step, j being the number of steps.
By the method, the accurate target height which changes is obtained in the scene that the target height changes, so that the accuracy of the final pedestrian height position estimation is further improved.
Preferably, the error linear compensation is performed on the estimated height of the wobble phase in step S4, that is, the estimated height of the wobble phase is corrected by using a target error compensation algorithm, and the calculation formula is as follows:
hj＝hi-Δh(i/m)；
wherein, Δ h ═ ht-hs，htRespectively and sequentially carrying out preprocessing based on Lauda criterion, median filtering, sliding average filtering preprocessing and trend removing preprocessing on the original target height data of the pedestrian, which are measured by the barometer in an accumulated mode, and then averaging the three processed attitude phase height data to obtain data; h issThe estimated height of the attitude phase is obtained by performing quadratic integration according to the corrected vertical acceleration of the attitude phase; the period t of the swing phase is m tau, and m is the total number of sampling points in the swing phase; h isjIs the estimated height after error compensation, hiIs an estimated height of the swing phase obtained by quadratic integration based on the vertical acceleration of the swing phase, and Δ h (i/m) is a compensation height, where j ═ i ═ 1 … m. Based on the estimation of the scene target height, the estimated height of the swing phase is compensated by using the error between the estimated height and the target height, and the pedestrian height estimation is accurately realized.
Compared with the prior art, the invention has the following advantages:
1. determining a posture phase and a swing phase of pedestrian gait motion by adopting a multi-threshold detection method, respectively correcting vertical direction accelerations of the posture phase and the swing phase, updating the speed of the posture phase to zero, and respectively performing secondary integration on the corrected vertical direction accelerations to obtain estimated heights of the posture phase and the swing phase of the pedestrian; then, it is determined whether there is a deviation between the estimated height data obtained from the attitude phase and the target height data measured by the barometer (the step height at each step measured by the barometer is defined as the target height)? If so, carrying out error linear compensation on the estimated height obtained by the swing phase to finally obtain updated pedestrian positioning data, and particularly adopting a multi-threshold detection and error compensation algorithm to greatly reduce the problem of height estimation accumulated error caused by inherent drift of an IMU sensor, thereby realizing high-precision estimation of the height position of the pedestrian in the multi-story building based on one sensor node and solving the problem of low personnel height positioning precision in a complex environment; the method can accurately and reliably estimate the vertical height of the pedestrian in the multi-story building, and is generally suitable for human body tracking and indoor pedestrian navigation application.
2. The pedestrian height position estimation method only needs one sensor node when estimating the height position of the pedestrian, and has good portability;
3. compared with DH and HDIB methods, the invention has higher accuracy and reliability in height estimation, can realize that the height estimation error is less than 2cm at each step, and the accumulated error only occupies 2 percent of the total height of the stroke.
4. The whole scheme of the invention is very simple, mainly aims to realize pedestrian height estimation based on multi-threshold detection and complementary filtering algorithm of vertical acceleration, and compensates the estimated height obtained by integrating the acceleration by utilizing a target error compensation algorithm to effectively reduce height error. Compared with a DH and HDIB pedestrian height positioning method, the method does not depend on a mathematical physical model of lower limb movement, and is simple and easy to implement and good in precision.
The main difficulties of the invention are that: 1. the definition of a normalized vertical acceleration change threshold is put forward for the first time and is used for multi-threshold gait phase detection; the sliding variance threshold based on the vertical acceleration data is firstly used in the multi-threshold gait phase detection to measure the fluctuation degree of the acceleration data, and the multi-threshold phase detection method realizes the simple and accurate determination of the attitude phase and the swing phase of the pedestrian and lays a foundation for the accurate estimation of the final height position of the pedestrian. 2. A target error compensation algorithm is provided for the first time, and the estimated height of the swing phase is corrected, so that the high-precision estimation of the height position of the pedestrian in the multi-story building can be realized.
Drawings
FIG. 1 is a schematic diagram of a method for estimating the height of a pedestrian in a multi-story building using MEMS integrated sensing devices in accordance with one embodiment of the present invention;
FIG. 2 is a schematic representation of a typical gait phase of a pedestrian according to an embodiment of the invention;
FIG. 3 is a diagram illustrating the results of the attitude phase and stance phase detection during the upstairs process of a pedestrian based on multi-threshold detection in accordance with an embodiment of the present invention;
FIG. 4 is a flow chart of a method for pedestrian height estimation in accordance with an embodiment of the present invention;
FIG. 5 is a flow chart of a complementary filtering method according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of an experimental scenario described in Experimental example 1 of the present invention;
fig. 7 is a diagram showing the result of estimating the pedestrian height upstairs in experimental example 1 of the present invention;
FIG. 8 is a graph showing the result of estimating the height of a pedestrian in a multi-story building for a long period of time according to the present invention in Experimental example 1;
FIG. 9 is a schematic diagram of an indoor self-constructed walking experiment platform according to experiment example 2 of the present invention;
fig. 10 is a diagram showing a result of estimating a pedestrian height of the indoor self-constructed walking experimental platform according to experimental example 2 of the present invention;
fig. 11 is a diagram showing error results of a pedestrian height estimation experiment based on the self-built walking experiment platform according to experimental example 2 of the present invention.
Detailed Description
The embodiment of the invention comprises the following steps: the method for estimating the height position of the pedestrian in the multi-story building based on the MEMS sensor, as shown in fig. 1 and 4, comprises the following steps:
s1, fixing the MEMS integrated sensing device on the lower limb of the pedestrian; the MEMS integrated sensing device comprises: the three-axis accelerometer, the gyroscope and the barometer (the specification parameters shown in the table 1 can be adopted in the specific implementation) are respectively used for acquiring acceleration data, angular velocity data and pedestrian height data in the pedestrian walking stair process; in specific acquisition, the raw data can be acquired and recorded at a sampling rate of 74hz, and then transmitted to the host computer through the Bluetooth module and stored.
In order to ensure that the coordinate system is consistent with the terrestrial coordinate system, the IMU can be preliminarily calibrated, namely, the raw data collected by the accelerometer and the gyroscope is subjected to Rayleigh criterion to remove abnormal values, and a median filter and a low-pass filter are adopted to remove data noise.
TABLE 1
| Sensing unit | Accelerometer | Gyroscope | Barometer |
| Dimension (d) of | 3axes | 3axes | 1axe |
| Dynamic range | ±2g | ±200dps/s | — |
| Sampling frequency | 37 | 37 | 37 |
| Degree of linearity | 0.5 | 0.1 | — |
And S2, determining the posture phase and the swing phase of the gait motion of the pedestrian by establishing constraint conditions of three threshold values to the acceleration data in the gravity direction on the basis of the acceleration data in the vertical direction:
wherein,representing the acceleration in the vertical direction, wherein i is the number of sampling points, W is the size of a sliding window, and in order to better describe the data volatility, the W is selected to be 3 through multiple tests and comparisons; g is the acceleration of gravity, λ1The variance threshold value representing the acceleration in the sliding window is used for measuring the fluctuation degree of the acceleration data, and the lambda can be set through multiple tests1＝0.0096；x＝(τ,Ai+1-Ai) Y ═ 0 (τ,0), τ is the sampling period; lambda [ alpha ]2In order to normalize the vertical acceleration change threshold value, the invention is used for judging and determining the critical values of the attitude phase and the swing phase, and the lambda can be selected through a plurality of experiments2>0.04；Represents the phase time of the current posture,representing the next attitude phase duration; lambda [ alpha ]3A threshold value representing the time-length ratio between two adjacent attitude phases,for screening wrong gait phase, the invention can select lambda30.75; determining the vertical direction acceleration data sequence which simultaneously meets the three threshold conditions as an attitude phase; the remaining data sequence is then the wobble phase. A typical schematic diagram of the pedestrian gait phase is shown in fig. 2. The attitude phase and the swing phase determined using the multi-threshold detection method are shown in fig. 3.
S3, correcting the vertical accelerations of the attitude phase and the swing phase, respectively, that is: in order to reduce drift error, when the single foot is static, the acceleration in the vertical direction of the attitude phase is corrected by updating the acceleration to zero; expressing the motion direction of the swing phase by using a quaternion, carrying out data fusion on the collected angular velocity and acceleration through a complementary filtering algorithm, and correcting the acceleration in the vertical direction with drift; updating the speed of the attitude phase to zero, and respectively performing secondary integration on the corrected acceleration in the vertical direction to obtain the estimated heights of the pedestrian attitude phase and the swing phase;
specifically, the calculation formula of the pedestrian height estimation is as follows:
wherein a isz(t),vz(t),hz(t) defined as vertical acceleration, vertical velocity and altitude, respectively;
because the accelerometer data has better static performance but poorer dynamic performance compared with the gyroscope data, the invention adopts a complementary filtering algorithm to perform data fusion on the acquired acceleration and angular velocity, thereby improving the dynamic performance of the accelerometer data. The complementary filtering algorithm is shown in fig. 5. Specifically, the calculation formula (3) is as follows:
wc＝wm-(Pe+Iea) (3)
wherein e is defined as the error between the measured acceleration and the corrected acceleration, and the accumulated error is defined as eaTo representIn the complementary filtering algorithm, the parameter P is 0.8, and I is 0.008. Angular velocity measured by gyroscopemAnd (4) showing. The corrected angular velocity is denoted wc＝[wxwywz]And is used to update the quaternion, thereby obtaining a corrected vertical direction acceleration. Using quaternionsDescribing the spatial motion direction of the foot, updating by using a calculation formula (4) and obtaining a motion rotation matrix of the footThe calculation formula is shown in (5).
Wherein,the orientation of the sensor(s) relative to the earth (e) in a coordinate system is described, byObtaining a corrected vertical acceleration azThe calculation formula is shown as (6):
s4, in order to reduce height estimation errors, an IMU integrated barometer system is adopted to detect the height of the pedestrian in real time, the error between the estimated height and the target height is calculated, and target error compensation is carried out; i.e., whether there is a deviation between the estimated height data obtained from the attitude phase and the target height data measured by the barometer (the step height at each step measured by the barometer is defined as the target height)? If so, performing error linear compensation on the estimated height obtained by the swing phase (for reducing the error between the estimated height and the target height), and finally obtaining updated pedestrian positioning data.
Preferably, the target height data measured by the barometer is as follows: the method comprises the steps of respectively and sequentially carrying out preprocessing based on Lauda criterion, median filtering, sliding average filtering preprocessing and trend removing preprocessing on original target height data of pedestrians measured by a barometer under an attitude phase, and then averaging the three processed attitude phase height data to obtain data. The calculation formula is shown as (7):
whereinAnd i is a sampling point for the preprocessed barometer height data.
The error linear compensation is performed on the estimated height obtained by the swing phase in step S4, that is, the estimated height of the swing phase is corrected by using a target error compensation algorithm, and the calculation formula is as follows:
hj＝hi-Δh(i/m)； (8)
Δh＝ht-hs(9)
wherein h issThe estimated height of the attitude phase is obtained by performing quadratic integration according to the corrected vertical acceleration of the attitude phase; the period t of the swing phase is m tau, and m is the total number of sampling points in the swing phase; h isjIs the estimated height after error compensation, hiIs an estimated height of the swing phase obtained by quadratic integration based on the vertical acceleration of the swing phase, and Δ h (i/m) is a compensation height, where j ═ i ═ 1 … m; h istFor cumulative measurement of micro-barometersRespectively and sequentially carrying out preprocessing based on Lauda criterion, median filtering, sliding average filtering preprocessing and trend removing preprocessing on the original target height data of the pedestrian, and then averaging the three processed attitude phase height data to obtain data, or htThe data obtained after further processing were as follows:
under the scene that the target height of each step of pedestrians in the multi-story building is the same and unchanged, correcting the target height data by the following method: calculating the target height increment of each step and the mean value and standard deviation of the target height increment; and when the absolute value of the difference value between the target height increment and the average value in a certain step is more than one standard deviation, updating the target height increment based on the weighted average of the target height increments in the adjacent three steps.
Under the scene that the target heights of the pedestrians at each step in the multi-story building are different, improving the target height increment of each step by using the following formula, and further improving the target height data of each step:
wherein,increment, h, representing target height of step j1Height constants for the first step of the attitude phase (e.g. the target height value for the first step can be taken in general)Represents the target height of the jth step, j being the number of steps.
It should be clear that the above embodiments are only preferred embodiments of the present invention, and do not limit the technical solutions of the present invention, and any simple replacement or change based on the solutions falls within the protection scope of the present invention.
In order to verify the effect of the present invention, the inventors also performed the following experimental examples:
two young 21-year-old males of average height 176 cm were selected as subjects. In the experiment, these subjects walked at a usual speed and posture.
Experimental example 1
The subject walked from first to fourth floor for approximately 5 minutes, and a simplified experimental scenario is shown in fig. 6. The heights of the floors are respectively 4.31m, 3.46m and 3.46 m. The subject's activities in this multi-story building included three types of walking on level ground, going upstairs and going downstairs by elevator (to show that the 3 trials performed were consecutive). Based on the basic motion scene, the pedestrian height estimation method of the invention is used for estimating the height position of the pedestrian in the multi-story building, and the estimation result is shown in fig. 7. Where 1 represents a pedestrian altitude position estimate based on complementary filtering and target error compensation, the error between the estimated altitude and the actual stair altitude was analyzed in order to evaluate the accuracy of the altitude estimate, as shown in table 2. Where the true values are height data measured in the field. Table 2 illustrates the good accuracy of the results of the pedestrian height estimation method of the present invention.
TABLE 2
To demonstrate that the method of the present invention can achieve accurate pedestrian height position estimation for a long period of time in a multi-story building, the inventors performed 3 consecutive height estimation experiments within 12 hours, each lasting 20 minutes. Throughout the experiment, subjects had 3 walking periods (T1, T2, T3) and 2 rest periods. During walking, the testee goes from first floor to fourth floor (F1-F4), then goes downstairs by elevator, and the sensors continuously acquire and send data during walking and downstairs by elevator. After the subject makes 3 rounds of going up and down stairs (S1, S2, S3), the sensor is removed and placed on a table for rest. During the rest period, the sensor enters a standby mode after the sensor cannot detect the walking movement, so that the energy is saved. When the wearing sensor walks again, the sensor is awakened. The height estimation results obtained after 12 hours of continuous experiments are shown in fig. 8. The results of the error analysis between the pedestrian height position and the actual measurement are shown in table 3, and it can be seen that the height estimation error accounts for about 2% of the total height of the trip. Since the drift error of the MEMS sensor is cumulative throughout the test, it can also be derived from table 3: the scheme of the invention can solve the problem of accumulative error of the MEMS sensor.
TABLE 3
Experimental example 2
An indoor walking step is built, the step model schematic diagram is as shown in fig. 9, and the following description is made according to the walking route direction of a pedestrian, namely, the pedestrian goes up the step from the left side and goes down the step from the right side: the height of the front 4 single steps on the left side is 0.12m, the height of the last step is 0.28m, the height of the first step on the right side is 0.22m, and the heights of the rest three steps are 0.18 m. Because the sensor device is worn on the right calf of a pedestrian and the left leg and the right leg alternately step on steps, a subject walks on the first step by the right foot, walks on the second step by the left foot, walks on the third step by the right foot, walks on the fourth step by the left foot, walks on the highest position by the fifth step by the right foot, walks on the middle platform by the left foot and the right foot alternately in the sixth step, walks on the first step on the right side under the right foot in the ninth step, walks on the second step under the left foot, walks on the third step under the right foot in the eleventh step by the right foot, walks on the left foot down on the terrace in the. Therefore, when the pedestrian walks on the landing, the target height (based on the right foot) increment of each step is respectively 0.12m, 0.24m, 0.40m, 0.00m, -0.22m, -0.36m and-0.18 m. Fig. 10 is a schematic diagram of the height estimation result. Wherein 1 represents the pedestrian height position calculated by the complementary filtering and target error compensation algorithm of the invention based on the acceleration data collected by the accelerometer, 2 represents the original pedestrian height measured by the barometer, and 3 represents the pedestrian height measured by the filtered barometer. It can be seen that the estimation 1 is close to the true target height (as noted in step) and the barometric measurement height 2. In order to further verify the accuracy and robustness of the pedestrian height position estimation method, the experiment is continuously repeated for 5 times, and the experimental result is evaluated and analyzed. The pedestrian height calculated based on the acceleration data, using the complementary filtering and the target error compensation algorithm, the pedestrian height calculated by using the complementary filtering algorithm alone, and the pedestrian height data measured based on the barometer are compared with the actual stair height, and the obtained results are shown as curves 1,2 and 3 in fig. 11. The pedestrian height estimation method based on the complementary filtering and the target error compensation has the height estimation error within 2cm, which is obviously lower than the height estimation error of a barometer and a complementary filter. Experiments show that in a stair scene with uneven step heights, the pedestrian height estimation method based on the complementary filtering and target error compensation algorithm can achieve more accurate height position estimation.
The invention accurately estimates the height position of the pedestrian in the stairs of the multi-storey building. To reduce the drift error of the IMU, the attitude addition velocity is corrected by updating the attitude acceleration to zero, and the roll addition velocity is corrected by a complementary filter. And estimating the target height of each step according to the height of the barometer, and performing error compensation to reduce estimation errors. The experimental result shows that the method can accurately estimate the height position of the testee when going up and down stairs in the building, and the detection device can stably and continuously work for more than 12 hours. Meanwhile, the method of the invention has good performance on self-built pedestrian terraces with different step heights, and the height estimation error of each step is within 2 cm. When the subject travels on the stairs with the same step height, the height estimation error of each step is 0.5cm, and the accumulated height error of all the steps in the process of going upstairs is 1.7 percent of the total height. In 3 pedestrian height estimation trials (20 minutes each) completed in 12 consecutive hours, the cumulative height error was about 2% of the total height of travel (11.23 meters).
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910086482.XA CN109827568B (en) | 2019-01-29 | 2019-01-29 | Estimation method of pedestrian height position in multi-storey building based on MEMS sensor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910086482.XA CN109827568B (en) | 2019-01-29 | 2019-01-29 | Estimation method of pedestrian height position in multi-storey building based on MEMS sensor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109827568A true CN109827568A (en) | 2019-05-31 |
| CN109827568B CN109827568B (en) | 2020-12-15 |
Family
ID=66862845
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910086482.XA Active CN109827568B (en) | 2019-01-29 | 2019-01-29 | Estimation method of pedestrian height position in multi-storey building based on MEMS sensor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109827568B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110986874A (en) * | 2019-12-11 | 2020-04-10 | 金华航大北斗应用技术有限公司 | Height measurement system and method based on barometer and micro-electromechanical inertia measurement element |
| CN112747747A (en) * | 2021-01-20 | 2021-05-04 | 重庆邮电大学 | Improved UWB/IMU fusion indoor pedestrian positioning method |
| CN113008268A (en) * | 2021-02-05 | 2021-06-22 | 中国科学院空天信息创新研究院 | Wearable pedestrian positioning elevation correction method based on step detection |
| CN113483763A (en) * | 2021-06-30 | 2021-10-08 | 北京航空航天大学 | Indoor personnel elevation estimation method with autonomy |
| CN113609440A (en) * | 2021-07-26 | 2021-11-05 | 华智生物技术有限公司 | Sensor data fusion method, device and equipment |
| CN114018217A (en) * | 2021-09-29 | 2022-02-08 | 河南省联睿智能科技研究院有限公司 | Robust difference absolute height estimation method based on secondary calibration |
| CN115655217A (en) * | 2022-09-21 | 2023-01-31 | 深圳泰辑物联网科技有限公司 | A method, device, electronic equipment and storage medium for calculating cumulative climb height |
| CN116108873A (en) * | 2022-12-12 | 2023-05-12 | 天津大学 | Motion posture assessment system based on RFID/IMU fusion |
| CN117545983A (en) * | 2021-06-22 | 2024-02-09 | 三菱电机楼宇解决方案株式会社 | Position measuring system for operator |
| CN117731276A (en) * | 2024-02-19 | 2024-03-22 | 天津大学 | Information processing method and wearable device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102175244A (en) * | 2011-03-16 | 2011-09-07 | 公安部沈阳消防研究所 | In-building person positioning system and positioning method thereof based on inertial sensor |
| CN104501796A (en) * | 2014-12-16 | 2015-04-08 | 重庆邮电大学 | Indoor WLAN/MEMS fusion cross-stair three-dimensional positioning method |
| CN104613965A (en) * | 2015-03-02 | 2015-05-13 | 大连理工大学 | Stepping type pedestrian navigation method based on bidirectional filtering smoothing technology |
| CN105241454A (en) * | 2015-10-23 | 2016-01-13 | 中国兵器工业集团第二一四研究所苏州研发中心 | System and method for pedestrian navigation based on multiple sensors |
| US9832617B1 (en) * | 2015-09-30 | 2017-11-28 | Apple Inc. | Using periodic magnetometer signals for indoor navigation |
| CN107990895A (en) * | 2017-11-08 | 2018-05-04 | 北京工商大学 | A kind of building floor gap pedestrian track tracking and system based on wearable IMU |
| CN108537101A (en) * | 2018-01-05 | 2018-09-14 | 浙江大学 | A kind of pedestrian's localization method based on state recognition |
| CN108875668A (en) * | 2018-06-28 | 2018-11-23 | 东北大学 | A kind of pedestrian detection method based on multiple features and cascade classifier |
-
2019
- 2019-01-29 CN CN201910086482.XA patent/CN109827568B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102175244A (en) * | 2011-03-16 | 2011-09-07 | 公安部沈阳消防研究所 | In-building person positioning system and positioning method thereof based on inertial sensor |
| CN104501796A (en) * | 2014-12-16 | 2015-04-08 | 重庆邮电大学 | Indoor WLAN/MEMS fusion cross-stair three-dimensional positioning method |
| CN104613965A (en) * | 2015-03-02 | 2015-05-13 | 大连理工大学 | Stepping type pedestrian navigation method based on bidirectional filtering smoothing technology |
| US9832617B1 (en) * | 2015-09-30 | 2017-11-28 | Apple Inc. | Using periodic magnetometer signals for indoor navigation |
| CN105241454A (en) * | 2015-10-23 | 2016-01-13 | 中国兵器工业集团第二一四研究所苏州研发中心 | System and method for pedestrian navigation based on multiple sensors |
| CN107990895A (en) * | 2017-11-08 | 2018-05-04 | 北京工商大学 | A kind of building floor gap pedestrian track tracking and system based on wearable IMU |
| CN108537101A (en) * | 2018-01-05 | 2018-09-14 | 浙江大学 | A kind of pedestrian's localization method based on state recognition |
| CN108875668A (en) * | 2018-06-28 | 2018-11-23 | 东北大学 | A kind of pedestrian detection method based on multiple features and cascade classifier |
Non-Patent Citations (2)
| Title |
|---|
| YUN-KI KIM; SEUNG-HWAN CHOI; HYUN-WOO KIM; JANG-MYUNG LEE: "Performance improvement and height estimation of pedestrian dead-reckoning system using a low cost MEMS sensor", 《2012 12TH INTERNATIONAL CONFERENCE ON CONTROL, AUTOMATION AND SYSTEMS》 * |
| 朱金鑫等: "基于MEMS惯性传感器的行人室内高度估计方法", 《工业控制计算机》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110986874A (en) * | 2019-12-11 | 2020-04-10 | 金华航大北斗应用技术有限公司 | Height measurement system and method based on barometer and micro-electromechanical inertia measurement element |
| CN112747747A (en) * | 2021-01-20 | 2021-05-04 | 重庆邮电大学 | Improved UWB/IMU fusion indoor pedestrian positioning method |
| CN112747747B (en) * | 2021-01-20 | 2022-10-11 | 重庆邮电大学 | An Improved UWB/IMU Fusion Indoor Pedestrian Localization Method |
| CN113008268B (en) * | 2021-02-05 | 2023-02-17 | 中国科学院空天信息创新研究院 | An elevation correction method for wearable pedestrian positioning based on step detection |
| CN113008268A (en) * | 2021-02-05 | 2021-06-22 | 中国科学院空天信息创新研究院 | Wearable pedestrian positioning elevation correction method based on step detection |
| CN117545983A (en) * | 2021-06-22 | 2024-02-09 | 三菱电机楼宇解决方案株式会社 | Position measuring system for operator |
| CN113483763A (en) * | 2021-06-30 | 2021-10-08 | 北京航空航天大学 | Indoor personnel elevation estimation method with autonomy |
| CN113483763B (en) * | 2021-06-30 | 2022-11-18 | 北京航空航天大学 | An Autonomous Altitude Estimation Method for Indoor Personnel |
| CN113609440A (en) * | 2021-07-26 | 2021-11-05 | 华智生物技术有限公司 | Sensor data fusion method, device and equipment |
| CN114018217B (en) * | 2021-09-29 | 2023-09-05 | 河南省联睿智能科技研究院有限公司 | Robust differential absolute height estimation method based on secondary calibration |
| CN114018217A (en) * | 2021-09-29 | 2022-02-08 | 河南省联睿智能科技研究院有限公司 | Robust difference absolute height estimation method based on secondary calibration |
| CN115655217A (en) * | 2022-09-21 | 2023-01-31 | 深圳泰辑物联网科技有限公司 | A method, device, electronic equipment and storage medium for calculating cumulative climb height |
| CN116108873A (en) * | 2022-12-12 | 2023-05-12 | 天津大学 | Motion posture assessment system based on RFID/IMU fusion |
| CN116108873B (en) * | 2022-12-12 | 2024-04-19 | 天津大学 | Sports posture assessment system based on RFID/IMU fusion |
| CN117731276A (en) * | 2024-02-19 | 2024-03-22 | 天津大学 | Information processing method and wearable device |
| CN117731276B (en) * | 2024-02-19 | 2024-04-19 | 天津大学 | Information processing method and wearable device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109827568B (en) | 2020-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109827568B (en) | Estimation method of pedestrian height position in multi-storey building based on MEMS sensor | |
| CN109827577B (en) | High-precision inertial navigation and positioning algorithm based on motion state detection | |
| CN104296750B (en) | A zero-speed detection method and device, and a pedestrian navigation method and system | |
| KR101250215B1 (en) | Pedestrian dead-reckoning system using kalman filter and walking state estimation algorithm and method for height estimation thereof | |
| US10352959B2 (en) | Method and system for estimating a path of a mobile element or body | |
| CN110398245B (en) | Indoor pedestrian navigation attitude estimation method based on foot-worn inertial measurement unit | |
| US10267646B2 (en) | Method and system for varying step length estimation using nonlinear system identification | |
| CN106017461B (en) | Pedestrian navigation system three-dimensional fix method based on human body/environmental constraints | |
| Kappi et al. | MEMS-IMU based pedestrian navigator for handheld devices | |
| JP7023234B2 (en) | How to estimate pedestrian movement | |
| CN107990895B (en) | A method and system for pedestrian trajectory tracking between floors in buildings based on wearable IMU | |
| CN111024126A (en) | An Adaptive Zero Speed Correction Method in Pedestrian Navigation and Positioning | |
| CN112362057A (en) | Inertial pedestrian navigation algorithm based on zero-speed correction and attitude self-observation | |
| CN104897158B (en) | Double orientation method and system in a kind of pedestrian room | |
| CN109099913B (en) | Wearable navigation device and method based on MEMS inertial device | |
| CN107490378B (en) | Indoor positioning and navigation method based on MPU6050 and smart phone | |
| Brahms et al. | Stride length determination during overground running using a single foot-mounted inertial measurement unit | |
| CN104613965B (en) | A kind of step-by-step movement pedestrian navigation method based on bidirectional filtering smoothing technique | |
| Li et al. | A robust pedestrian navigation algorithm with low cost IMU | |
| CN111649742B (en) | Elevation estimation method based on ANFIS assistance | |
| CN118149823B (en) | Foot binding type pedestrian three-dimensional indoor positioning method assisted by ranging information | |
| Kim et al. | Performance improvement and height estimation of pedestrian dead-reckoning system using a low cost MEMS sensor | |
| KR101642286B1 (en) | Heading Orientation Estimation Method Using Pedestrian Characteristics in Indoor Environment | |
| KR101301462B1 (en) | Pedestiran dead-reckoning apparatus using low cost inertial measurement unit and method thereof | |
| CN109612464B (en) | Multi-algorithm enhanced indoor navigation system and method based on IEZ framework |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |