Pedestrian dead reckoning system correction method based on real-time dynamic measurement technology

Abstract

The invention discloses a pedestrian dead reckoning system correction method based on a real-time dynamic measurement technology, and relates to the field of high-precision positioning, and the method comprises the steps: obtaining a PDR system position variable quantity and an RTK position variable quantity of a person wearing a PDR system device; according to the PDR system position variable quantity and the RTK position variable quantity, whether the PDR system is corrected or not is determined; if it is determined that the PDR system is corrected, determining a course angle offset by using the PDR system position variation and the RTK position variation; and carrying out pose correction on the PDR system based on the course angle offset. According to the method, the positioning accuracy and the system robustness of the PDR system in a long-time and long-span positioning task can be improved.

Description

Technical Field

[0001] The present invention relates to the field of high-precision positioning, and in particular to a correction method for a pedestrian dead reckoning system based on real-time dynamic measurement technology. Background Art

[0002] Pedestrian Dead Reckoning (PDR) uses data from multiple sensors such as accelerometers, magnetometers, and gyroscopes to estimate the movement of pedestrians, and combines the initial position information of pedestrians to calculate the real-time position of pedestrians through data fusion and filtering algorithms. The specific principles of traditional PDR technology involve four key steps: step detection, step length estimation, heading angle estimation, and position update. Specifically, after detecting the pedestrian's steps through the step detection algorithm, the step length (referred to as step length) is estimated based on the human kinematics and dynamic characteristics of the pedestrian during the step, and the pedestrian's current posture is estimated based on the gyroscope readings, and finally the real-time dead reckoning of the pedestrian's position is completed.

[0003] Although PDR is different from the traditional strapdown inertial navigation system (SINS), it avoids the direct accumulation of accelerometer errors into the system and causes system failure. However, due to the differences in human bodies and the high nonlinearity of movements, the lack of adaptability of heading angle deviation and step length estimation for different people is the main error of PDR and also the biggest challenge facing PDR.

[0004] In order to solve the problem of heading angle deviation caused by the accumulation of gyroscope errors, scholars have proposed the following two strategies: one strategy focuses on using other sensors, such as magnetometers, to assist in estimating the orientation of personnel, and to make the best estimate of the personnel posture calculated by Kalman filtering and gyroscope. The second strategy is a heading constraint algorithm based on heuristic assumptions for special occasions, such as the corridor heading constraint algorithm based on the assumption of indoor personnel turning curvature. This type of algorithm does not require the addition of new sensors, so it has also been widely studied. However, in the indoor working environment denied by the Global Navigation Satellite System (GNSS), the magnetometer is easily interfered by other equipment and needs to complete the more cumbersome ellipsoid fitting calibration; and the constraint algorithm based on heuristic assumptions, although it has good results in daily life, often cannot play a role in disaster relief scenarios with extremely complex terrain.

[0005] In order to solve the problem of estimation accuracy and personnel adaptability of step length estimation, scholars have established the following three types of step length estimation models: The first type of model is a step length prediction model established by the characteristic quantity related to the step length, and a statistical regression model that fits its parameters. Among this type, the model proposed by Shin et al. is the most classic. This model estimates the step length SL of a person based on the variance of the person's step frequency and waist acceleration. The expression is as follows: SL = a·f+b·σ 2 +c, where f is the pedestrian’s step frequency, σ 2 is the variance of the waist acceleration modulus, (a, b, c) are the parameters to be fitted, which need to be calibrated and set in advance. Under different motion states, the correlation between features such as step frequency and step length will change, which leads to certain differences in the parameters to be fitted. Therefore, the model has poor robustness and can often only be applied to a single motion state. The second type of model is a sports physiological model based on human motion characteristics. It is based on the human motion model and estimates the stride length of a person through the dynamic characteristics of various parts of the human body. For example, the Weinberg model and Kim model established based on the human inverted pendulum hypothesis are expressed as and In the equation, K1 and K2 are unknown coefficients related to the height and weight of the person, respectively, and need to be calibrated and set in advance. Because human motion is too complex, this type of model is often based on some heuristic assumptions, so it only performs well in individual motion states. The third type of model is a direct integration model based on the quadratic integral of the pedestrian's center of mass acceleration. The most representative model of this type is the Kose model based on the optimal filtering forward and reverse integration method proposed by Kose. However, this type of model must face the problem of accumulated accelerometer measurement errors. In addition to sensor fusion and Kalman filtering solutions, heuristic virtual observations are often used as corrections. Therefore, this type of model also has strong application scenario limitations.

[0006] In summary, the defects of existing PDR systems are as follows: (1) Regarding the problem of pedestrian heading angle deviation, the pedestrian heading angle deviation makes the PDR system unable to cope with long-term and large-span positioning tasks, and the existing heading correction methods often have low estimation accuracy and robustness, making it difficult to put them into practical application scenarios. (2) Regarding the problem of pedestrian step length estimation method's adaptability to people, the existing pedestrian step length estimation model often has parameters related to people's height, weight, and health status that need to be manually calibrated due to its inherent assumptions. Before use, it needs to be calibrated according to the wearer's height, weight, etc. Therefore, the use process is cumbersome and the adaptability to different wearers is poor, making it difficult to put them into practical application scenarios. Summary of the invention

[0007] The present invention provides a pedestrian dead reckoning system correction method based on real-time dynamic measurement technology, which at least solves the problem that the PDR system cannot cope with long-term and large-span positioning tasks.

[0008] The present invention provides a pedestrian dead reckoning system correction method based on real-time dynamic measurement technology, the method comprising: obtaining the PDR system position change and RTK position change of a person wearing a PDR system device; determining whether to correct the PDR system according to the PDR system position change and the RTK position change; if it is determined that the PDR system is to be corrected, determining the heading angle offset using the PDR system position change and the RTK position change; and correcting the posture of the PDR system based on the heading angle offset. The present invention estimates the heading angle offset of the PDR system in combination with the RTK technology to complete the PDR system posture correction task, thereby improving the positioning accuracy and system robustness of the PDR system in long-term and long-span positioning tasks.

[0009] In addition, the method further includes: determining whether to calibrate the step length estimation model parameters of the PDR system according to the position change of the PDR system and the RTK position change; if it is determined to calibrate the step length estimation model parameters of the PDR system, determining the calibration values ​​of the parameters affected by the person's height and weight in the step length estimation model according to the step length of each step of the person determined based on the RTK measurement value and the characteristic value of the motion characteristic quantity of the person determined based on the IMU; calibrating the parameters of the step length estimation model affected by the person's height and weight using the calibration values ​​of the parameters affected by the person's height and weight. The present invention completes the automatic real-time calibration task of key parameters in the PDR system by combining the real-time position information of the person measured by the RTK technology, significantly simplifies the use process, and improves the adaptability of the PDR system to different wearers in practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 It is a posture correction flow chart of the PDR system correction method based on real-time dynamic measurement technology provided by the present invention;

[0011] Figure 2 It is a typical system flow diagram of the PDR system correction method based on the real-time dynamic measurement technology provided by the present invention;

[0012] Figure 3 It is a schematic diagram of a sliding data window;

[0013] Figure 4 It is a parameter calibration flow chart of the PDR system correction method based on real-time dynamic measurement technology provided by the present invention;

[0014] Figure 5 This is a typical system flow diagram of PDR system calibration based on real-time dynamic measurement technology;

[0015] Figure 6 It is a flow chart of the step detection algorithm based on continuous wavelet transform analysis;

[0016] Figure 7 is a schematic diagram of the analysis results of the step detection algorithm;

[0017] Figure 8 It is a schematic diagram of the positioning results of the traditional PDR system and the GPDR system using the PDR system calibration and correction method based on real-time dynamic measurement technology in an actual application environment. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0019] The invention discloses a correction method for a pedestrian dead reckoning system based on real-time kinematic measurement technology. On the one hand, the real-time kinematic (RTK) technology is used to measure the real-time position information of personnel, and on this basis, the heading angle deviation, which is one of the main error sources of the PDR system, is estimated and constrained, and the heading angle deviation rate estimation and PDR posture correction tasks are completed. At the cost of a relatively low increase in system complexity, the positioning accuracy and system robustness of the PDR system, especially the waist-hanging PDR system, in long-term and large-span positioning tasks are effectively improved. On the other hand, the PDR system calibration based on the real-time kinematic carrier phase difference technology uses the RTK technology to measure the real-time position information of personnel, and combines the step detection algorithm based on continuous wavelet transform analysis to calculate the actual step size of the personnel (the length of each step), and completes the automatic real-time calibration task of key parameters in the PDR system that are highly related to the height, weight, and health status of the personnel, which significantly simplifies the use process of the PDR system and effectively improves the adaptability of the PDR system to different wearers in practical applications.

[0020] The above two aspects are described in detail below in conjunction with the accompanying drawings.

[0021] According to a first aspect of the present invention, see Figure 1 The posture correction involved in the PDR system correction method based on real-time dynamic measurement technology of the present invention may include the following steps:

[0022] Step S1: obtaining the PDR system position change and the real-time dynamic carrier phase difference RTK position change of a person wearing a PDR system device.

[0023] The step of obtaining the PDR system position change of the person wearing the PDR system device in step S1 may include: updating the posture of the person according to the measurement value of the inertial sensor; determining the step length of the person by performing step detection and step length estimation on the person; determining the PDR system position of the person according to the posture and step length of the person, and determining the PDR system position change of the person by comparing with the PDR system position of the person determined in the previous posture update cycle.

[0024] The step of obtaining the RTK position change of the person wearing the PDR system device in step S1 may include: obtaining the real-time dynamic RTK position of the person, and converting the RTK position of the person into the northeastern sky position of the person; determining the RTK position change of the person according to the northeastern sky position of the person. Further, the step of obtaining the RTK position change of the person wearing the PDR system device may also include: after obtaining the RTK position of the person, obtaining the RTK status, the horizontal dilution of precision (HDOP) value and the vertical dilution of precision (VDOP) value; determining whether to convert the RTK position of the person into the northeastern sky position of the person according to the RTK status, the HDOP value and the VDOP value.

[0025] Step S2: Determine whether to correct the PDR system based on the position change of the PDR system and the RTK position change.

[0026] The step S2 may include: when the absolute value of the position change of the PDR system is greater than or equal to a first preset change and the absolute value of the position change of the RTK is greater than or equal to a first preset change, determining to correct the PDR system. The first preset change may be set to 5 m.

[0027] Step S3: If it is determined to correct the PDR system, the heading angle offset is determined using the position change of the PDR system and the RTK position change.

[0028] The step S3 may specifically include: respectively determining the cross product result of the PDR system position change and the RTK position change, and the dot product result of the PDR system position change and the RTK position change; determining the ratio of the cross product result to the dot product result, and using the ratio as the tangent value; determining the angle corresponding to the tangent value, and using the angle as the heading angle offset of the PDR system.

[0029] Step S4: Based on the heading angle offset, the PDR system is corrected in position and posture.

[0030] The step S4 may include: adding the heading angle offset to the existing heading angle to obtain a corrected heading angle.

[0031] When applying the above method to a PDR system, the typical system flow is shown in Figure 2 .

[0032] First, the real-time attitude update based on the inertial sensor is completed. Specifically, it is determined whether it is initialized. If so, the attitude update is completed based on the equivalent rotation vector method. If not, the dual-vector attitude determination method of the stationary platform is used to use the data collected by the gyroscope and accelerometer to determine the initial attitude of the PDR system module, and then the attitude update is completed based on the equivalent rotation vector method.

[0033] Among them, the coordinate transformation matrix from the carrier coordinate system (b system) to the navigation coordinate system (n system) is The attitude matrix of the vehicle is called the attitude matrix. "Attitude update" refers to the real-time update of the attitude matrix based on the output of the inertial sensor. matrix, matrix.

[0034] Furthermore, the waist-mounted PDR has a large body shaking and a high real-time requirement, so the present invention adopts an equivalent rotation vector algorithm for attitude update. Taking into account the attitude update frequency and the compensation of non-commutative errors, the present invention will adopt an equivalent rotation vector algorithm of a single sample plus the previous cycle. The details are as follows:

[0035] The carrier attitude can be expressed by attitude quaternion Q, Q and attitude matrix There is a corresponding relationship, which can also be represented by The yaw, pitch and roll angles of the carrier are calculated. The attitude quaternion Q is related to the angular velocity of the carrier relative to the geographic coordinate system. Satisfying the quaternion differential equation, the attitude differential equation is as follows:

[0036]

[0037] To complete the attitude update of strapdown inertial navigation, it is necessary to solve the differential equation of attitude quaternion. Using the equivalent rotation vector method, the rotation vector is defined as:

[0038] Φ=Φu i (Φ is the magnitude of the vector, u i is the direction of the vector)

[0039] After the rotation vector Φ is calculated by the gyroscope in the IMU, the attitude change quaternion q(Δk) can be calculated from the rotation vector. The attitude change quaternion q(Δk) can be further calculated to obtain the attitude quaternion Q(t k+1 ), the calculation formula is as follows:

[0040]

[0041] Among them, p(Δk) is the quaternion of the navigation coordinate system attitude change from time k to time k+1, Q(t k ) is the attitude quaternion from the carrier coordinate system to the navigation coordinate system at time k, and q(Δk) is the attitude change quaternion from the carrier coordinate system at time k to time k+1. The attitude change quaternion q(Δk) can be expressed as:

[0042]

[0043] The differential equation for the rotation vector is:

[0044]

[0045] The approximate equation is:

[0046]

[0047] In engineering, the above formula is usually expanded by Taylor. (i represents the inertial coordinate system, b represents the carrier coordinate system, n represents the navigation coordinate system, and w represents the angular velocity; the superscript indicates the coordinate system in which the current w is projected, so The projection of the rotation angular velocity of the b coordinate system relative to the n coordinate system in the b coordinate system) is generally fitted using multiple points. The error caused by conical motion can also be compensated using a straight line fitting. When , the equivalent rotation vector single sample plus the angular increment of the previous cycle is expressed as follows:

[0048]

[0049] Among them, in one attitude update cycle, Δθ1 is the angular increment of the gyro output in this cycle, and Δθ0 is the angular increment of the gyro output in the previous cycle. Based on the equivalent rotation vector Φ(Δk) obtained from the above formula, the attitude change quaternion q(Δk) can be calculated. Finally, the updated Euler attitude angle att = [roll angle roll, pitch angle pitch, heading angle yaw] is obtained by solving.

[0050] Secondly, in order to solve the heading deviation problem caused by gyroscope error, this method measures the real-time position information of personnel based on real-time dynamic carrier phase difference technology, estimates the heading angle error, and uses Kalman filtering to complete the heading angle correction. The details are as follows:

[0051] The pedestrian dead reckoning system based on real-time dynamic measurement technology measures and calculates the real-time longitude and latitude information of pedestrians and records it as [longitude L latitude λ height h]. Figure 3 The length of the sliding data window shown is 10 seconds, and the overlap rate is 30, which is used to estimate the track angle error of the PDR system. First, the relative position change of the personnel in the data window is calculated. At this time, the longitude L, latitude λ and altitude h need to be converted to the east, north and sky positions. The conversion relationship is as follows:

[0052]

[0053] Among them, R N , R M are the principal curvature radii of the zodiac circle and the meridian circle of the earth (ellipsoid model); P E , P N , P U They refer to the relative position coordinates (in meters) converted from longitude and latitude (L, λ, h) (radians) to the east, north, and sky coordinate systems (with a certain longitude and latitude as the origin and calculated based on the change in longitude and latitude), referred to as east, north, and sky position coordinates.

[0054] According to the recorded PDR system position change Perform conditional judgment. If satisfied Then calculate the heading angle offset:

[0055]

[0056] On this basis, the RTK measurement position is used to correct the current position of the PDR. At the same time, the heading angle offset δθ calculated based on the RTK and PDR results is added to the existing heading angle yaw as the repaired heading angle to repair the YAW angle offset caused by the gyroscope error. The formula is as follows:

[0057]

[0058] yaw=yaw+δθ

[0059] Ultimately, the position and heading of the PDR system are corrected, providing higher accuracy and stronger system robustness for long-term positioning tasks.

[0060] According to the second aspect of the present invention, see Figure 4 The PDR system calibration method based on real-time dynamic measurement technology uses real-time dynamic carrier phase difference technology to measure the real-time position information of personnel to complete the automatic calibration task of the PDR system, which may include the following steps:

[0061] Step S10: Determine whether to calibrate the step length estimation model parameters of the PDR system according to the position change of the PDR system and the RTK position change.

[0062] The step 10 may be specifically as follows: when the absolute value of the position change of the PDR system is greater than or equal to a second preset change and the absolute value of the position change of the RTK is greater than or equal to a second preset change, determining to calibrate the step length estimation model parameters of the PDR system. The second preset change may be set to 5m.

[0063] Step S20: If it is determined to calibrate the step length estimation model parameters of the PDR system, the calibration values ​​of the parameters affected by the person's height and weight in the step length estimation model are determined according to the step length of each step of the person determined based on the RTK measurement value and the characteristic value of the motion characteristic quantity of the person determined based on the IMU.

[0064] The step 20 may be specifically as follows: constructing a linear expression of the step length estimation model according to the step length estimation model involved in parameter calibration; constructing a step length matrix of the person according to the step length of each step of the person determined based on the RTK measurement value; constructing a motion feature matrix of the person according to the eigenvalues ​​of the characteristic quantities applied to the step length estimation model within each step of the person's motion interval determined based on the IMU; and solving the matrix constructed by parameters affected by the person's height and weight in the linear expression of the step length estimation model by using the step length matrix of the person and the motion feature matrix of the person.

[0065] Step S30: calibrating the parameters of the step length estimation model affected by the person's height and weight using the calibration values ​​of the parameters affected by the person's height and weight.

[0066] When the above parameter calibration method is applied to the PDR system, the typical system flow is shown in Figure 5 .

[0067] First, in order to complete the calibration of the step length estimation module in the PDR system, the step detection task must be completed.

[0068] Specifically, a step detection algorithm is constructed based on the continuous wavelet analysis method. The algorithm flow is as follows: Figure 6As shown in the figure, it includes: ① To avoid the influence of the IMU installation method, the L2 norm of the acceleration measured by the accelerometer is first calculated; ② A high-pass filter with a cutoff frequency of 0.5Hz is used to filter the low-frequency error caused by the hardware; ③ Daubechies wavelet is used as the mother wave, and the filtered signal is subjected to continuous wavelet transform (or continuous wavelet analysis) and continuous wavelet parameter analysis; ④ Finally, the key action time is detected to determine the step boundary.

[0069] Among them, wavelet is a short waveform, generally named And with the neighborhood of t=0 as the center, it satisfies:

[0070]

[0071] Assume ψ is a real wavelet, then the wavelet transform corresponding to the function f can be written as:

[0072]

[0073] The variable u represents the time conversion variable, which controls the wavelet to scan the entire time window; the variable s represents the scale of the extended wavelet to analyze the different frequency characteristics of the signal at a specific time. This operation obtains the time-frequency characteristics of the function f(t) by analyzing the convolution of the wavelet of scale s and the function f(t) at time u. The function f(t) represents a filtered version of the acceleration modulus. The higher the correlation between the function and the appropriately extended wavelet, the higher the convolution value obtained.

[0074] Among them, the acceleration modulus, the coefficients of the relevant frequency band after continuous wavelet transformation, and their corresponding actions are as follows Figure 7 As shown in the figure. The moment when the step action occurs (i.e. when the acceleration modulus reaches the local maximum value) is considered to be the moment when the correlation between the L2 norm of the acceleration and the wavelet is the highest. After analyzing the time-frequency characteristics of the acceleration modulus and obtaining the step time node, the original data is segmented according to the step time node and the step features are extracted to provide a data interface for the subsequent algorithm.

[0075] Secondly, complete the calibration task of the step size model.

[0076] Specifically, ① after obtaining the RTK measurement value, make a preliminary assessment of the RTK solution status and positioning error; ② if the RKT is in a fixed solution state and the HDOP and VDOP values ​​are low, convert the longitude and latitude measured by RTK into relative coordinates and store them together with the PDR predicted position in the above data window; ③ process the data window and calculate the change in the personnel position and ④If satisfied Then enter the parameter calibration routine of the step size estimation model.

[0077] Taking five step length estimation models in the PDR field as examples, the step length estimation model calibration routine is explained. The step length estimation models involved in the calibration include:

[0078] (1) The Weinberg model is widely used in the field of pedestrian navigation due to its accurate dynamic step length estimation capability and small amount of calculation. Its expression can be expressed as:

[0079]

[0080] Among them, A max With A min are the maximum and minimum values ​​of the L2 norm of acceleration in each step respectively, and K1 is the parameter that needs to be calibrated.

[0081] (2) The Mikov model takes into account the duration of each step on the basis of the Weinberg model and adds it to the parameters to be calibrated in the Weinberg expression. The expression is as follows:

[0082]

[0083] Among them, A max With A min are the maximum and minimum values ​​of the L2 norm of acceleration in each step, t step is the duration of each step, and K2 is the parameter that needs to be calibrated.

[0084] (3) The step length estimation of Kim's model is based only on the average value of the acceleration modulus during the step.

[0085]

[0086] Among them, |a i | is the acceleration modulus of the ith step, and K3 is the parameter that needs to be calibrated.

[0087] (4) The Zhu model replaces the parameters to be calibrated in the Weinberg model with a linear regression model of cadence and acceleration variance;

[0088]

[0089] Among them, f, v are the variance of step frequency and acceleration, r lf With r lv They are the Pearson correlation coefficients between stride length and stride frequency, and between stride length and acceleration variance.

[0090] (5) The Shin model is a priori statistical regression model that assumes that the step length is linearly related to the pedestrian's step frequency and the variance of the acceleration signal, which can be expressed as follows:

[0091]

[0092] Among them, f and v are the variances of the step frequency and acceleration modulus respectively, and K5 = [abc], which contains the three coefficients that need to be calibrated for the Shin model.

[0093] The real-time and accurate position information of personnel can be obtained based on the RTK measurement value of personnel position, so that the step length L (l1, l2, ..., l k ,…l N ), as well as the personnel motion feature extracted by IMU, the linear representation of the above step length estimation model is uniformly written as:

[0094] HK=5L

[0095]

[0096] K=[K1 K2 K3 K4 K5] T

[0097] L=[l1 l2 … l u … l N ] T

[0098] h u,g represents the calculated value of the feature quantity used by the g-th model in the u-th step action interval. K represents the parameters contained in each model, which needs to be determined by solving a linear equation system. u Represents the true value of the step length of the u-th step of the person.

[0099] Finally, the least squares solution of the above equation is:

[0100] K=(H T H) -1 H T L

[0101] In fact, the models involved in the automatic calibration of parameters can be less than or more than 5. In the application, the PDR system will select different models according to different personnel states, such as Shin for walking and Weinberg for running, to obtain a system with better accuracy and robustness. It should be pointed out that personnel state recognition does not belong to the present invention and is therefore not described here.

[0102] In summary, the automatic calibration task of the PDR system is completed based on the real-time dynamic carrier phase difference technology, which significantly simplifies the use process of the PDR system and improves the personnel adaptation capability of the PDR.

[0103] In a 23-minute actual scenario test, the conventional PDR effect and the GPDR positioning effect of the pedestrian dead reckoning system calibration and correction method based on real-time dynamic measurement technology of the present invention are shown as follows: Figure 8 As shown, the positioning trajectory of the present invention is basically consistent with the RTK trajectory.

[0104] In summary, the present invention can achieve high-precision positioning of the PDR system, and its advantages are mainly reflected in the following two aspects:

[0105] 1. The present invention measures the real-time position information of personnel based on real-time dynamic carrier phase differential technology, estimates the pedestrian heading angle error in the PDR system, and completes fusion and correction based on Kalman filtering. This method effectively constrains the heading angle deviation problem caused by gyroscope error in traditional PDR, and improves the positioning accuracy and system robustness of the PDR system in long-term tasks;

[0106] 2. The present invention measures the real-time position information of personnel based on real-time dynamic carrier phase differential technology, and automatically calibrates the PDR system in an outdoor environment. This method simplifies the use process of the traditional PDR system and improves the adaptability and positioning accuracy of the PDR system to different wearers.

[0107] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the present invention is not limited thereby. Any modification, equivalent substitution and improvement made by those skilled in the art without departing from the scope and essence of the present invention shall be within the scope of the present invention.

Claims

1. A pedestrian dead reckoning system correction method based on real-time dynamic measurement technology, characterized in that: The method comprises: Obtain the PDR system position change and real-time dynamic carrier phase difference RTK position change of the person wearing the PDR system device; Determining whether to correct the PDR system according to the position change of the PDR system and the RTK position change; If it is determined to correct the PDR system, the heading angle offset is determined using the position change of the PDR system and the RTK position change; Based on the heading angle offset, the PDR system is corrected in position and posture.

2. The method according to claim 1, characterized in that The method further comprises: Determining whether to perform step size estimation model parameter calibration on the PDR system according to the position change of the PDR system and the RTK position change; If it is determined to calibrate the step length estimation model parameters of the PDR system, the calibration values ​​of the parameters affected by the person's height and weight in the step length estimation model are determined according to the step length of each step of the person determined based on the RTK measurement value and the characteristic value of the motion characteristic quantity of the person determined based on the IMU; The parameters of the step length estimation model affected by the person's height and weight are calibrated using the calibration values ​​of the parameters affected by the person's height and weight.

3. The method according to claim 1 or 2, characterized in that: The step of obtaining the PDR system position change of the person wearing the PDR system device comprises: updating the posture of the person according to the measurement value of the inertial sensor; Determining the stride length of the person by performing stride detection and stride length estimation on the person; The PDR system position of the person is determined according to the posture and step length of the person, and the change in the PDR system position of the person is determined by comparing the PDR system positions of adjacent posture update cycles.

4. The method according to claim 1 or 2, characterized in that: The step of obtaining the RTK position change of the person wearing the PDR system device includes: Acquire the real-time dynamic RTK position of the person, and convert the RTK position of the person into the northeast sky position of the person; The RTK position change of the person is determined according to the northeast celestial position of the person.

5. The method according to claim 4, characterized in that The method further comprises: After obtaining the real-time dynamic RTK position of the person, obtaining the RTK status, HDOP value and VDOP value; According to the RTK state, the HDOP value and the VDOP value, it is determined whether to convert the RTK position of the person into the northeastern sky position of the person.

6. The method according to claim 1 or 2, characterized in that: The determining whether to correct the PDR system according to the position change of the PDR system and the RTK position change includes: When the absolute value of the position change of the PDR system is greater than or equal to a first preset change and the absolute value of the position change of the RTK is greater than or equal to the first preset change, it is determined that the PDR system is to be corrected.

7. The method according to claim 1 or 2, characterized in that: Determining the heading angle offset by using the PDR system position change and the RTK position change comprises: Determine the cross product result and the dot product result of the PDR system position change and the RTK position change respectively; Determine a ratio of the cross product result to the dot product result, and use the ratio as a tangent value; Determine the angle corresponding to the tangent value, and use the angle as the heading angle offset of the PDR system.

8. The method according to claim 1 or 2, characterized in that: The performing posture correction on the PDR system according to the heading angle offset comprises: The heading angle offset is added to the existing heading angle to obtain a corrected heading angle.

9. The method according to claim 2, characterized in that: The determining whether to perform step length estimation model parameter calibration on the PDR system according to the position change of the PDR system and the RTK position change comprises: When the absolute value of the position change of the PDR system is greater than or equal to the second preset change and the absolute value of the position change of the RTK is greater than or equal to the second preset change, it is determined to calibrate the step size estimation model parameters of the PDR system.

10. The method according to claim 2, characterized in that Determining the calibration values ​​of parameters affected by the person's height and weight in the step length estimation model according to the step length of each step of the person determined based on the RTK measurement value and the characteristic value of the motion characteristic quantity of the person determined based on the IMU includes: According to the step length estimation model involved in parameter calibration, a linear expression of the step length estimation model is constructed; Constructing a step length matrix of the person according to the step length of each step of the person determined based on the RTK measurement value; Constructing a motion feature matrix of the person according to the feature values ​​of the feature quantities applied to the step length estimation model in each step of the person's motion interval determined based on the IMU; The stride length matrix of the person and the motion feature matrix of the person are used to solve the matrix constructed by parameters affected by the person including height and weight in the linear expression of the stride length estimation model.

Images