Pedestrian navigation method based on adaptive search interval zero-speed detection

Abstract

The invention discloses a pedestrian navigation method based on self-adaptive search interval zero-speed detection. The pedestrian navigation method comprises the following steps: S1, periodically collecting acceleration data and angular velocity data output by an inertial sensor; s2, judging whether the navigation system is initialized or not, if the navigation system is not initialized, initializing the navigation system, and entering S1 to recollect acceleration data and angular velocity data; if so, entering the step S3; s3, performing adaptive interval zero-speed point search based on the acceleration data and the angular velocity data, obtaining an inertia factor and a zero-speed factor based on an adaptive search interval, and constructing an optimization equation to predict inertia information according to the inertia factor and the zero-speed factor; and S4, predicting the pose change from the previous zero speed point to the current moment based on the inertial information, and outputting carrier navigation information. According to the method, accumulative error divergence can be effectively restrained by combining a self-adaptive zero-speed point detection technology with a graph optimization framework, and pedestrian positioning information is obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of pedestrian navigation, and in particular relates to a pedestrian navigation method based on adaptive search interval zero-speed detection. Background Art

[0002] Pedestrian navigation systems play a vital role in firefighting, emergency rescue, and other fields. However, their positioning accuracy is limited in satellite-denied environments, such as indoors and underground, resulting in reduced mission efficiency or even failure. To address this issue, pedestrian positioning methods based on inertial navigation have become a research hotspot due to their independence from external signals and strong autonomy.

[0003] Inertial navigation systems integrate the output of inertial sensors (such as accelerometers and gyroscopes) to calculate the wearer's position changes and achieve continuous positioning through recursion. However, due to the large noise and drift errors in the output of micro-inertial devices, inertial navigation systems will accumulate errors over long periods of operation, causing positioning results to quickly diverge. Therefore, how to effectively correct inertial positioning errors is one of the core issues of pedestrian navigation systems.

[0004] Zero-speed correction is a commonly used error correction method in inertial navigation systems. Its basic principle is to use the velocity error information at the time the wearer is detected to be stationary (zero-speed state) to correct the inertial navigation system. The accuracy of zero-speed correction is highly dependent on the accuracy of zero-speed detection. Traditional zero-speed detection methods typically use a fixed threshold to determine the zero-speed state. When the detector output falls below the threshold, the wearer is considered stationary. However, this method has significant limitations: the gait characteristics of different wearers vary significantly, and the dynamic characteristics of the same wearer in different motion states also vary. Fixed thresholds are difficult to adapt to complex and changing real-world scenarios. Furthermore, during large dynamic movements (such as running and jumping), the zero-speed interval is short or even difficult to capture, rendering zero-speed correction ineffective. Therefore, a parameter-adaptive maximum probability zero-speed detection method for complex gaits is needed to ensure accurate zero-speed detection in different gaits and wearers, thereby achieving high-precision pedestrian positioning in complex gaits. Summary of the Invention

[0005] To solve the above technical problems, the present invention proposes a pedestrian navigation method based on adaptive search interval zero-speed detection, which can effectively constrain the cumulative error divergence by utilizing adaptive zero-speed point detection technology combined with a graph optimization framework to obtain pedestrian positioning information.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides a pedestrian navigation method based on adaptive search interval zero-speed detection, comprising:

[0008] S1, periodically collect acceleration data and angular velocity data output by the inertial sensor;

[0009] S2. Determine whether the navigation system's posture, accelerometer, and gyroscope zero bias are initialized. If not, initialize the navigation system and proceed to S1 to recollect acceleration data and angular velocity data. If already initialized, proceed to step S3.

[0010] S3, performing an adaptive interval zero-speed point search based on the acceleration data and the angular velocity data, obtaining an inertia factor and a zero-speed factor in the adaptive search interval, and predicting inertia information based on the inertia factor and the zero-speed factor;

[0011] S4. Predict the posture change from the last zero-speed point to the current moment based on the inertial information, output the carrier navigation information, and return to step 1.

[0012] Optionally, performing adaptive interval zero-speed point search based on the acceleration data and angular velocity data includes:

[0013] Based on the acceleration data and angular velocity data, using a likelihood ratio method, obtaining a detection value within a preset time window;

[0014] According to the detection value, determine whether the number of variable nodes in the sliding window reaches a preset value, and obtain a determination result;

[0015] According to the judgment result, the maximum probability zero speed point is obtained.

[0016] Optionally, according to the judgment result, obtaining the maximum probability zero speed point includes:

[0017] When the number of variable nodes in the sliding window does not reach a preset value, a coarse search is performed based on a minimum value search method in a fixed search interval to obtain candidate zero-speed points; and a maximum probability zero-speed point is obtained based on the candidate zero-speed points;

[0018] When the number of variable nodes in the sliding window reaches a preset value, an adaptive search interval fitting is performed to obtain an interval search matrix, and the maximum probability zero-speed point is obtained according to the interval search matrix.

[0019] Optionally, obtaining the maximum probability zero-speed point according to the interval search matrix includes:

[0020] Obtaining interval credibility according to the interval search matrix;

[0021] Obtaining an optimal search interval according to the interval credibility;

[0022] Based on the optimal search interval, evaluating the confidence of the minimum value in each interval;

[0023] The maximum probability zero-speed point is obtained according to the confidence level.

[0024] Optionally, obtaining interval credibility according to the interval search matrix includes:

[0025] updating the interval search matrix, and obtaining the interval credibility according to the updated interval search matrix;

[0026] The method for updating the interval search matrix is:

[0027]

[0028] in, is the value of the search matrix in the i-th row and j-th column, i represents the size of the search interval, ranging from 1 to (3 / f imu );f imu represents the sampling period of the inertial sensor, j represents the sampling sequence number in the current sliding window, T j Indicates t j GLRT detection value at time T j-1 t j-i GLRT detection value at time T j+1 t j+i GLRT detection value at the moment,.

[0029] Optionally, according to the interval credibility, a method for obtaining the optimal search interval is:

[0030] evaluating the interval credibility, and recording the search interval with the maximum credibility as the optimal search interval;

[0031] The method for evaluating the credibility of the interval is:

[0032]

[0033] in, is the interval credibility when the search interval is i, W s is the total number of columns of the interval search matrix, that is, the total number of variable nodes in the current sliding window. When the maximum value is obtained, the optimal search interval W can be obtained. max =i.

[0034] Optionally, based on the optimal search interval, a method for evaluating the confidence level of the minimum value in each interval is:

[0035]

[0036] in, is the minimum confidence when the search index is j, W max is the optimal search interval, the minimum confidence The corresponding index is the maximum probability zero velocity point.

[0037] Optionally, obtaining the inertia factor of the adaptive search interval includes:

[0038] The inertia factor is calculated based on the inertia data from the time corresponding to the previous maximum probability zero-speed point to the current time.

[0039] Optionally, obtaining the zero-speed factor of the adaptive search interval includes:

[0040] Construct foot state quantity;

[0041] Calculating a zero-speed factor residual according to the foot state quantity;

[0042] Calculate the residual confidence level based on the likelihood ratio test results;

[0043] The zero-speed factor is obtained according to the zero-speed factor residual and the residual confidence level.

[0044] Optionally, predicting inertia information according to the inertia factor and the zero-speed factor includes:

[0045] Obtaining inertia constraint and zero speed constraint according to the inertia factor and zero speed factor;

[0046] Inertia information is predicted based on the inertia constraint and the zero-speed constraint in combination with an objective function.

[0047] Compared with the prior art, the present invention has the following advantages and technical effects:

[0048] The present invention discloses zero-speed point detection for adaptive interval search based on inertial sensor measurement values, and performs zero-speed constraint correction optimization based on a factor graph, thereby eliminating the cumulative error of pedestrian navigation under different wearers and motion conditions, and can effectively reduce the error caused by inaccurate zero-speed detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0050] Figure 1 This is a flow chart of a pedestrian navigation method based on adaptive search interval zero-speed detection according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0052] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0053] This embodiment proposes a pedestrian navigation method based on adaptive search interval zero speed detection, such as Figure 1 As shown, the specific steps include:

[0054] S1, periodically collect acceleration data and angular velocity data output by the inertial sensor;

[0055] S2. Determine whether the navigation system's posture, accelerometer, and gyroscope zero bias are initialized. If not, initialize the navigation system and proceed to S1 to recollect acceleration data and angular velocity data. If already initialized, proceed to step S3.

[0056] S3, performing a zero-speed point search in an adaptive interval based on the acceleration data and the angular velocity data, obtaining an inertia factor and a zero-speed factor in the adaptive search interval, and predicting inertia information based on the inertia factor and the zero-speed factor;

[0057] S4. Predict the posture change from the last zero-speed point to the current moment based on the inertial information, output the carrier navigation information, and return to step 1.

[0058] Specifically, step 1. periodically collect acceleration data and angular velocity data output by the inertial sensor at time k;

[0059] Step 2. Determine whether the navigation system is initialized. If not, initialize it, estimate the zero bias and initial attitude of the inertial sensor, and jump to step 1. If it has been initialized, go to step 3.

[0060] Step 3. Perform adaptive interval zero-speed point search based on acceleration data and angular velocity data, construct adaptive search intervals, and nest and compare the confidence of each search interval. The optimal search interval is obtained, and the zero-speed point in the optimal interval is the maximum probability zero-speed point Z j , and construct inertia and zero speed factors at the zero speed point, and optimize the zero speed point posture by combining inertia and zero speed constraints;

[0061] Step 4. Predict the pose change from the previous zero-speed point to the current moment based on the inertial information, output the carrier navigation information, and jump to step 1.

[0062] In this example, a mixed walking and running gait test was conducted around the teaching building to verify the improvement of positioning accuracy achieved by the zero-speed detection algorithm.

[0063] Furthermore, performing adaptive interval zero-speed point search based on acceleration data and angular velocity data includes:

[0064] Based on the acceleration data and angular velocity data, the likelihood ratio method is used to obtain the detection value within the preset time window;

[0065] According to the detection value, determine whether the number of variable nodes in the sliding window reaches the preset value and obtain the judgment result;

[0066] According to the judgment result, the maximum probability zero speed point is obtained.

[0067] Furthermore, according to the judgment result, obtaining the maximum probability zero speed point includes:

[0068] When the number of variable nodes in the sliding window does not reach the preset value, a coarse search is performed based on the minimum value search method in a fixed search interval to obtain candidate zero-speed points; based on the zero-speed point candidates, the maximum probability zero-speed point is obtained;

[0069] When the number of variable nodes in the sliding window reaches a preset value, an adaptive search interval fitting is performed to obtain an interval search matrix, and the maximum probability zero-speed point is obtained according to the interval search matrix.

[0070] Furthermore, according to the interval search matrix, obtaining the maximum probability zero-speed point includes:

[0071] Obtain interval credibility based on interval search matrix;

[0072] According to the interval credibility, the optimal search interval is obtained;

[0073] Based on the optimal search interval, evaluate the confidence of the minimum value in each interval;

[0074] According to the confidence level, the maximum probability zero-speed point is obtained.

[0075] Specifically, in S3, the adaptive interval zero speed point search process is as follows:

[0076] For the inertial information output by MIMU, first, based on the generalized likelihood ratio test (GLRT), the acceleration and angular velocity output are considered simultaneously. Taking the acceleration and angular velocity information into consideration, t is calculated according to the following formula: k Detection value within the time window;

[0077]

[0078] Among them, T k is the detection value of the GLRT zero-speed detector at time k, W is the sliding window size, are the noise variances of the gyroscope and accelerometer respectively, g is the local gravitational acceleration modulus, is the mean value of the accelerometer output within the window; k represents the sampling time of the MIMU; if the number of variable nodes in the sliding window does not reach the sliding window value, a coarse search is performed based on the minimum value search method of the fixed search interval; during human motion, a complete gait cycle is 0.7s to 1.3s, and the search interval is set to 1s to ensure the search quality. The minimum value of the current interval is extracted every other search interval to obtain the zero-speed point candidate i is the candidate point number; non-minimum value suppression with a window size of 0.3s is performed on the candidate zero-speed point found. If the point is the minimum value within the surrounding 0.3s, it is the maximum probability zero-speed point Z i ;

[0079] When the number of variable nodes in the sliding window reaches the preset value, an adaptive search interval fitting is performed based on the periodic characteristics of the GLRT detection value itself. To ensure that no detection points are lost, the upper limit of the detection period is set to 3s, and then the initial value is 0 and the dimension is (3 / f imu )×W s The interval search matrix M s , the interval search matrix is ​​updated as follows:

[0080]

[0081] Where i represents the search interval size, ranging from 1 to (3 / f imu ); j represents the sampling sequence number in the current sliding window, T j Indicates t j The GLRT detection value at the moment; the matrix update process is a nested loop process. First, a loop is performed starting from i, and a nested loop with j as the index is performed inside the loop. The starting size of j is i+1 and ends at W s -i, to avoid out-of-bounds search; in the nested loop process, the interval search matrix M is realized s Updates;

[0082] M s The row vector records the minimum search results corresponding to the search interval size i, and the column vector records the situation where the jth data is the minimum value in different search intervals; the number of minimum value points searched in different sliding window sizes is used as the interval credibility to evaluate the credibility of each interval Such as:

[0083]

[0084] in, is the interval credibility when the search interval is i, W s is the total number of columns of the interval search matrix, that is, the total number of variable nodes in the current sliding window. When the maximum value is obtained, the optimal search interval W can be obtained. max =i.

[0085] Records provide credibility The maximum search interval W max =i as the optimal search interval; based on the optimal search interval, evaluate the confidence of the minimum value in each interval, the result Such as:

[0086]

[0087] in, is the minimum confidence when the search index is j, W max is the optimal search interval, the minimum confidence The corresponding index is the maximum probability zero velocity point.

[0088] After completing the loop, look for With W max The index j is the same, and the maximum probability zero speed point Z detected can be obtained. j .

[0089] Furthermore, obtaining the inertia factor and the zero speed factor of the adaptive search interval includes:

[0090] Construct foot state quantity;

[0091] Calculate the factor residual according to the foot state quantity;

[0092] According to the factor residual, obtain the zero-speed factor residual;

[0093] Obtain the zero-speed factor based on the zero-speed factor residual.

[0094] Specifically, the inertia and zero-speed factors are constructed as follows:

[0095] In the factor graph optimization sliding window, the foot MIMU state x is constructed as follows: i , the state quantity to be optimized i-1:

[0096] X={x0,x1,x2,...,x i ,...,x n}

[0097]

[0098] Where Δp i-1,i Indicates the sliding window size, i-1 represents the state vector at the i-th maximum probability zero velocity point, including quaternion Position i, speed Accelerometer bias Gyroscope bias After constructing the variable node and adding the sliding window, the factor residual is calculated as the edge constraint of the corresponding variable node.

[0099] The inertia pre-integration factor residual is as follows:

[0100]

[0101] in, is a quaternion, i is the i-th position, For speed, is the accelerometer bias, is the gyroscope bias, is the navigation information measurement value obtained by recursion from inertial data i to j, is the position residual corresponding to the inertial data moment from i to j, is the attitude residual corresponding to the inertial data moment from i to j, is the velocity residual corresponding to the inertial data moment from i to j, Δb a is the acceleration bias residual corresponding to the inertial data moment from i to j, Δb ω is the angular velocity bias residual corresponding to the inertial data moment from i to j, is the attitude matrix in the world system corresponding to the i-th inertial data moment, is the world position corresponding to the j-th inertial data moment, is the world position corresponding to the i-th inertial data moment, is the world system velocity corresponding to the i-th inertial data moment, Δt ij is the time change from inertial data i to j, g w is the projection of gravity in the world system, is the position measurement value corresponding to the inertial data moment from i to j, is the attitude measurement value corresponding to the inertial data moment from i to j, is the attitude quaternion in the world system corresponding to the j-th inertial data moment, is the world-system velocity corresponding to the j-th inertial data moment, is the velocity measurement value corresponding to the inertial data moment from i to j, and X is the state quantity when constructing the current residual.

[0102] In terms of zero-speed factor construction, the foot speed is close to zero during the stance phase. At the maximum probability zero-speed point, a zero-speed factor can be constructed and added to the factor graph optimization framework to correct the inertial positioning error of individual soldiers. The residual of the i-th zero-speed factor is defined as follows:

[0103]

[0104] Residuals treated as optimization variables The derivative is shown as follows:

[0105]

[0106] Therefore, the zero-speed factor residual is about the variable to be optimized The Jacobian matrix is ​​shown as follows:

[0107]

[0108] To address the problem of degraded zero-speed correction performance under large dynamic conditions, an adaptive zero-speed factor covariance adjustment method based on motion evaluation is proposed. The confidence of the zero-speed factor correction is expressed based on the covariance. In individual soldier positioning, the initial moment is static. The GLRT detection value obtained at this time can be used as a relative reference closest to zero speed. By comparing it with the GLRT detection value at the maximum probability zero-speed point detected at subsequent moments, the confidence level of the zero-speed correction can be evaluated. Zero-speed factor covariance It can be described as follows:

[0109]

[0110] Among them, κ zv It is a coefficient related to the performance of the MIMU device, generally 0.3, T initial is the initial value of the GLRT detector at the initial time, k(i-1) is t i The GLRT detection value at the moment. By constructing the zero-speed residual and adaptive covariance, the adaptive zero-speed factor under different motion dynamics is constructed to ensure accuracy and robustness under complex gaits.

[0111] Furthermore, based on the inertia factor and the zero-speed factor, the predicted inertia information includes:

[0112] Obtain inertia constraint and zero speed constraint according to inertia factor and zero speed factor;

[0113] According to the inertia constraint and zero speed constraint, the inertia information is predicted in combination with the objective function.

[0114] Specifically, the method for optimizing the zero-speed point pose by combining inertia and zero-speed constraints is as follows:

[0115] In the factor graph optimization sliding window, the foot MIMU state Δv is constructed as follows: i-1,i , the state quantity to be optimized i-1:

[0116] X={x0,x1,x2,...,x i ,...,x n}

[0117]

[0118] Where Δp i-1,i Indicates the sliding window size, i-1 represents the state vector at the i-th maximum probability zero velocity point, including quaternion Position i, speed Accelerometer bias Gyroscope bias After constructing the variable node and adding the sliding window, the factor residual is calculated as the edge constraint of the corresponding variable node.

[0119] Combining prior information, pre-integration factor, and zero-speed factor, the optimization objective function of the distributed inertial positioning method for mixed gaits is shown as follows:

[0120]

[0121] Among them, ||r p -H p X|| 2 represents the prior information from marginalization, represents the inertia pre-integration residual between the i-th to j-th variable nodes, Represents the zero-speed residual of the i-th variable node. imu , I zv , I sl It is the set of all pre-integration, zero-speed, and step-size measurements corresponding to the variable nodes within the sliding window. is the covariance of the inertial pre-integration factor, which is related to the performance of the inertial device. is the covariance needle of the step size factor, which is determined by the confidence of the neural network output. For factors that exceed the sliding window range, marginalization is performed according to the dynamic situation.

[0122] The Levenberg-Marquardt method is used to perform state iterative optimization based on the ceres-solver solver to achieve batch optimization estimation of pedestrian positioning states.

[0123] The final experimental results are shown in Table 1 below:

[0124] Table 1

[0125] Method Name RMSE / m Accuracy Head and tail error / m Accuracy SHOE 3.66 1.76％ 5.74 2.77％ ADA-SHOE 2.52 1.22％ 4.13 1.99％ Proposed method 2.29 1.10％ 3.51 1.69％

[0126] This embodiment discloses a pedestrian navigation method based on zero-speed detection in an adaptive search interval. Inertial sensor measurement values ​​are used to detect the zero-speed point in an adaptive interval search, and zero-speed constraint correction optimization is performed based on a factor graph. This eliminates the cumulative error of pedestrian navigation under different wearers and motion conditions, and can effectively reduce the error caused by inaccurate zero-speed detection.

[0127] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A pedestrian navigation method based on adaptive search interval zero speed detection, characterized in that: include: S1, periodically collect acceleration data and angular velocity data output by the inertial sensor; S2. Determine whether the navigation system's posture, accelerometer, and gyroscope zero bias are initialized. If not, initialize the navigation system and proceed to S1 to recollect acceleration data and angular velocity data. If already initialized, proceed to step S3. S3, performing an adaptive interval zero-speed point search based on the acceleration data and the angular velocity data, obtaining an inertia factor and a zero-speed factor in the adaptive search interval, and predicting inertia information based on the inertia factor and the zero-speed factor; S4. Predict the posture change from the last zero-speed point to the current moment based on the inertial information, output the carrier navigation information, and return to step 1.

2. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 1, characterized in that: Performing adaptive interval zero-speed point search based on the acceleration data and angular velocity data includes: Based on the acceleration data and angular velocity data, using a likelihood ratio method, obtaining a detection value within a preset time window; According to the detection value, determine whether the number of variable nodes in the sliding window reaches a preset value, and obtain a determination result; According to the judgment result, the maximum probability zero speed point is obtained.

3. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 2, characterized in that: According to the judgment result, obtaining the maximum probability zero speed point includes: When the number of variable nodes in the sliding window does not reach a preset value, a coarse search is performed based on a minimum value search method in a fixed search interval to obtain candidate zero-speed points; and a maximum probability zero-speed point is obtained based on the candidate zero-speed points; When the number of variable nodes in the sliding window reaches a preset value, an adaptive search interval fitting is performed to obtain an interval search matrix, and the maximum probability zero-speed point is obtained according to the interval search matrix.

4. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 3, characterized in that: Obtaining the maximum probability zero-speed point according to the interval search matrix includes: Obtaining interval credibility according to the interval search matrix; Obtaining an optimal search interval according to the interval credibility; Based on the optimal search interval, evaluating the confidence of the minimum value in each interval; The maximum probability zero-speed point is obtained according to the confidence level.

5. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 4, characterized in that: Obtaining interval credibility according to the interval search matrix includes: updating the interval search matrix, and obtaining the interval credibility according to the updated interval search matrix; The method for updating the interval search matrix is: in, is the value of the search matrix in the i-th row and j-th column, i represents the size of the search interval, ranging from 1 to (3 / f imu );f imu represents the sampling period of the inertial sensor, j represents the sampling sequence number in the current sliding window, T j Indicates t j GLRT detection value at time T j-i t j-i GLRT detection value at time T j+1 t j+i GLRT detection value at the moment,.

6. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 5, characterized in that: According to the interval credibility, the method for obtaining the optimal search interval is: evaluating the interval credibility, and recording the search interval with the maximum credibility as the optimal search interval; The method for evaluating the credibility of the interval is: in, is the interval credibility when the search interval is i, W s is the total number of columns of the interval search matrix, that is, the total number of variable nodes in the current sliding window. When the maximum value is obtained, the optimal search interval W can be obtained. max =i.

7. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 6, characterized in that: Based on the optimal search interval, the method for evaluating the confidence of the minimum value in each interval is: in, is the minimum confidence when the search index is j, W max is the optimal search interval, the minimum confidence The corresponding index is the maximum probability zero velocity point.

8. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 1, characterized in that: Obtaining the inertia factor of the adaptive search interval includes: The inertia factor is calculated based on the inertia data from the time corresponding to the previous maximum probability zero-speed point to the current time.

9. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 1, characterized in that: Obtaining the zero-speed factor of the adaptive search interval includes: Construct foot state quantity; Calculating a zero-speed factor residual according to the foot state quantity; Calculate the residual confidence level based on the likelihood ratio test results; The zero-speed factor is obtained according to the zero-speed factor residual and the residual confidence level.

10. The pedestrian navigation method based on adaptive search interval zero speed detection according to claim 1, characterized in that: According to the inertia factor and the zero speed factor, the predicted inertia information includes: Obtaining inertia constraint and zero speed constraint according to the inertia factor and zero speed factor; Inertia information is predicted based on the inertia constraint and the zero-speed constraint in combination with an objective function.

Images