CN116448111A - Pedestrian indoor navigation method, device and medium based on multi-source information fusion - Google Patents

Abstract

The invention discloses a pedestrian indoor navigation method, a device and a medium based on multi-source information fusion, wherein the method comprises the following steps: map information of the current indoor environment is built, and a Wi-Fi fingerprint library is built; acquiring multi-source information; constructing a position model of the pedestrian to obtain position information of the pedestrian; obtaining the movement distance of the pedestrian through gait detection and step estimation; constructing a fusion model based on extended Kalman filtering, and carrying out fusion processing on multi-source information and pedestrian position information to obtain fused position information; and further correcting and updating the position information of the pedestrians according to the map information and the particle filtering algorithm by utilizing the fused position information. According to the invention, a multisource information fusion technology is introduced, an extended Kalman filtering algorithm and a particle filtering algorithm are introduced, wi-Fi fingerprint matching and map information are added, a proper model is created, errors in an inertial navigation system are corrected, and safer, more reliable and more accurate position information is provided for the system.

Description

A pedestrian indoor navigation method, device and medium based on multi-source information fusion

technical field

The present invention relates to the technical field of navigation, in particular to a pedestrian indoor navigation method, device and medium based on multi-source information fusion.

Background technique

With the development and progress of society, people's living standards are improving day by day, and the demand for reliable and real-time location-based services is also getting higher and higher. However, under indoor conditions, due to the complex and changeable indoor environment, there are many obstacles and interference sources, etc., which make the positioning performance of satellite navigation signals deteriorate, and the effect in indoor positioning is poor. When satellite navigation fails, common navigation methods are:

(1) Inertial navigation. Indoor positioning technology based on inertial devices is based on inertial sensors. It mainly includes pedestrian inertial navigation system PINS and pedestrian dead reckoning PDR. The former often requires high sensor accuracy, and needs to ensure long-term positioning reliability by suppressing error drift; the latter often requires lower sensor accuracy and depends on the characteristics of people walking. Updating the pedestrian position in the way of accumulating steps. Compared with PINS, PDR can often achieve better positioning effect and obtain higher cost-effectiveness under the condition of limited cost. However, because low-cost MEMS inertial sensors tend to have large errors, and the errors will gradually accumulate over time, which has an extremely adverse impact on the positioning results.

(2) Wi-Fi positioning. Due to the popularization of Wi-Fi technology in recent years, most indoors have already been covered by Wi-Fi signals. The positioning technology based on Wi-Fi signals is supported by a large number of existing infrastructures. Moreover, the signal strength of Wi-Fi can be obtained directly through smartphones, which greatly promotes the development of Wi-Fi positioning technology. However, because Wi-Fi signals are susceptible to environmental interference, positioning accuracy is usually difficult to guarantee.

(3) Multi-source information fusion indoor positioning technology, multi-source information fusion technology uses information obtained by multiple sensors, and classifies, processes, fuses and uses the information according to specific rules, so as to realize the complementary advantages of information between sensors . Complementing the advantages of inertial navigation, it can effectively suppress the cumulative error of inertial navigation and further improve the positioning accuracy.

With the increasing demand for indoor high-precision positioning, the requirements for positioning accuracy also increase, but there is still a lack of a technical solution for indoor high-precision positioning.

Contents of the invention

In order to solve one of the technical problems in the prior art at least to a certain extent, the object of the present invention is to provide a pedestrian indoor navigation method, device and medium based on multi-source information fusion.

The technical scheme adopted in the present invention is:

A pedestrian indoor navigation method based on multi-source information fusion, comprising the following steps:

Construct the map information of the current indoor environment, select reference points on the map, and establish a Wi-Fi fingerprint database;

Use the sensor on the smart terminal to obtain multi-source information, the multi-source information includes Wi-Fi signal strength, angular velocity, velocity, acceleration and attitude information;

According to described Wi-Fi signal strength and Wi-Fi fingerprint storehouse, build the position model of pedestrian, obtain the position information of pedestrian;

Judging the instantaneous attitude of the intelligent terminal through the attitude information, transforming the acceleration from the carrier coordinate system to the navigation coordinate system, and obtaining the movement distance of the pedestrian through gait detection and step length estimation;

According to the multi-source information and the movement distance, the pedestrian movement model is constructed to obtain the pedestrian movement information, the movement information includes: position, speed; according to the extended Kalman filter algorithm, the fusion model based on the extended Kalman filter is constructed, and the multi-source information and the location information of pedestrians are fused, and the fused location information is obtained, and the state vector update and the optimal estimation of the error value in the extended Kalman filter are completed;

Using the fused position information, the pedestrian's position information is further corrected and updated according to the map information and the particle filter algorithm, and the pedestrian indoor navigation with multi-source information fusion is completed.

Further, the expression of the position model of the pedestrian is:

in, Represents the estimated position, x(t _k ), y(t _k ) represent the position coordinates corresponding to the first K position points, ω _j (t _k ) represents the weight coefficient of the corresponding data point, and the weight coefficient based on signal strength can be expressed/ > RSSI is signal strength.

Further, the acceleration is converted from the carrier coordinate system to the navigation coordinate system in the following way:

in, is the direction cosine matrix from the navigation system to the carrier system, C _r is the rotation matrix around the Y axis, C _p is the rotation matrix around the Z axis, _Cy is the rotation matrix around the X axis, γ is the rotation angle around the Y axis, p is the rotation angle around the X axis, and y is the rotation angle around the Z axis.

Further, the step of described gait detection comprises:

Compute the overall acceleration:

In the formula, a _x , a _y , a _z are the collected three-axis acceleration values, and the influence of the sensor attitude is reduced by calculating the overall acceleration a;

According to the overall acceleration, the potential peak value is obtained in the sliding window, and the initial judgment is made using the preset acceleration threshold;

Calculate the time difference between the potential peak and the previous peak, and use the preset walking step time threshold range to make a second judgment;

Compare the potential peak with the acceleration of the front and rear neighbors, and make three judgments to remove the false peak. If the potential peak point is the maximum value, the algorithm will count one step, otherwise, no step counting process will be performed.

Further, the step size is estimated as:

Where SL is the step length, a _max and a _min are the maximum and minimum values of the pedestrian's longitudinal acceleration, respectively.

Further, the extended Kalman filter algorithm includes:

X _k+1 ＝f(X _k )+w _k

Among them, X _k+1 is the state variable of the k+1th step, X _k is the state variable of the kth step, w _k is the process noise; f() is the nonlinear state function in the extended Kalman filter;

Methods to achieve linearization of the system matrix include:

Among them, Z _k is the observation variable, v _k is the observation noise, h() is the measurement function in the extended Kalman filter;

Methods to achieve linearization of the observation matrix include:

Among them, H _k is the Jacobian matrix of the partial derivative of h(X) to X, is the state estimate of the kth step.

Further, the steps of updating the state vector in the expanded Kalman filter include:

One-step state prediction update:

In the formula, is the optimal value of the previous state, /> is the one-step forecast value of the current state;

One-step forecast estimation error covariance matrix update:

In the formula, P _k+1|k is Corresponding covariance one-step forecast value, φ _k+1|k is the state transition matrix, P _k|k is /> Corresponding covariance, /> is the transpose of φ _k+1|k , Q _k is the process noise covariance matrix;

Compute the extended Kalman filter gain:

In the formula, K _k+1 is the extended Kalman filter gain matrix, H _k+1 is the observation matrix, is the transposition of the observation matrix, and R _k+1 is the measurement noise covariance matrix;

Compute innovations from observation vectors and update state estimates:

In the formula, is the optimal estimated value of the current state, Z _k+1 is the current observed value, /> is the predicted value for the current observation; update the estimated error covariance:

P _k+1 ＝[IK _k+1 H _k+1 ]P _k+1|k

In the formula, P _k+1 is The corresponding covariance, I is the identity matrix.

Further, the further correction and updating of the pedestrian's position information according to the map information and the particle filter algorithm includes:

The state transition equation and observation equation of the particle filter are as follows:

x(t)=f(x(t-1),u(t),w(t))

y(t)=f(x(t),e(t))

In the formula, x(t) is the state at time t, u(t) is the control quantity, w(t) and e(t) are the state noise and observation noise respectively, and the particle filter is obtained from the observation y(t) and the state at the last time x(t-1), u(t), w(t) filter out the state x(t) at time t;

The update method in particle filter mainly includes the following steps:

A1. Initialization: take the initial value x ⁱ (k) of the observed value as the initial value of the probability density function x(0), by right Sampling is carried out N times, and the initial w ⁱ (k) is set to />

A2. One-step prediction: For each particle x ⁱ (k), obtain a new particle by converting the formula p(x(k+1)| ^xi (k));

A3. Importance sampling: For any particle x ⁱ (k+1), solve their weight w ⁱ (k+1)=p(z(k+1)| ^xi (k+1));

A4. Normalization: Normalize the weights:

A5. Resampling: Resampling of particles according to the weight size, particles with large weights have a higher probability of repetition, and particles with smaller weights have a lower probability of repetition;

A6. The particle update also adds map information:

Due to the limitation of the indoor walking area, map matching is performed according to the map information. When the one-step predicted particle crosses the passable area (such as crossing the wall), its weight value is assigned to 0:

A7. The basic particle filter algorithm is used for calculation, and the indoor positioning and navigation of pedestrians are completed according to the calculation results.

Another technical scheme adopted in the present invention is:

A pedestrian indoor navigation device based on multi-source information fusion, comprising:

at least one processor;

at least one memory for storing at least one program;

When the at least one program is executed by the at least one processor, the at least one processor implements the above method.

Another technical scheme adopted in the present invention is:

A computer-readable storage medium stores a processor-executable program therein, and the processor-executable program is used to perform the above method when executed by a processor.

The beneficial effects of the present invention are: the present invention introduces multi-source information fusion technology, introduces extended Kalman filter algorithm and particle filter algorithm, adds Wi-Fi fingerprint matching and map information at the same time, creates a suitable model, and corrects errors in the inertial navigation system , to provide more secure, reliable and accurate location information for the system.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following describes the accompanying drawings of the embodiments of the present invention or the related technical solutions in the prior art. It should be understood that the accompanying drawings in the following introduction are only In order to clearly describe some embodiments of the technical solutions of the present invention, those skilled in the art can also obtain other drawings based on these drawings without creative work.

FIG. 1 is a schematic diagram of an overall flow of a pedestrian indoor navigation method based on multi-source information fusion in an embodiment of the present invention;

Fig. 2 is a schematic flow chart of building an extended Kalman filter fusion model in an embodiment of the present invention;

Fig. 3 is a schematic diagram of the construction process of the particle filter model in the embodiment of the present invention.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention. For the step numbers in the following embodiments, it is only set for the convenience of illustration and description, and the order between the steps is not limited in any way. The execution order of each step in the embodiments can be adapted according to the understanding of those skilled in the art sexual adjustment.

In the description of the present invention, it should be understood that the orientation descriptions, such as up, down, front, back, left, right, etc. indicated orientations or positional relationships are based on the orientations or positional relationships shown in the drawings, and are only In order to facilitate the description of the present invention and simplify the description, it does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention.

In the description of the present invention, several means one or more, and multiple means more than two. Greater than, less than, exceeding, etc. are understood as not including the original number, and above, below, within, etc. are understood as including the original number. If the description of the first and second is only for the purpose of distinguishing the technical features, it cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features relation.

In the description of the present invention, unless otherwise clearly defined, words such as setting, installation, and connection should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meanings of the above words in the present invention in combination with the specific content of the technical solution.

The present invention introduces multi-source information fusion technology, introduces extended Kalman filter algorithm and particle filter algorithm, adds Wi-Fi fingerprint matching and map information at the same time, creates a suitable model, corrects errors in the inertial navigation system, and provides a safer and more reliable solution for the system. and precise location information. Therefore, an indoor pedestrian navigation method based on multi-source information fusion is established to provide an effective method for precise indoor positioning of pedestrians.

Referring to Figure 1, this embodiment provides a pedestrian indoor navigation method based on multi-source information fusion, including the following steps:

Step 1: Construct the map information of the current indoor environment, select reference points on the map, and establish a Wi-Fi fingerprint library; use the built-in sensor of the mobile phone to obtain multi-source information, and the multi-source information includes: Wi-Fi signal strength, angular velocity, speed, Magnetic field strength and attitude information.

Step 2: Process the Wi-Fi signal collected in step 1 to construct a pedestrian location model; obtain pedestrian location information.

In this embodiment, when the pedestrian position model is constructed, the pedestrian position model is improved, and the k-proximity model is replaced by a weighted k-proximity model.

Specifically, building a pedestrian position model includes:

in Represents the estimated position, x(t _k ), y(t _k ) represent the position coordinates corresponding to the first K position points, ω _j (t _k ) represents the weight coefficient of the corresponding data point, and the weight coefficient based on signal strength can be expressed/ > RSSI is signal strength.

Step 3: Determine the instantaneous attitude of the mobile phone through the direction sensor in step 1, and convert the acceleration from the carrier coordinate system to the navigation coordinate system through a certain method. Then the pedestrian's movement distance is obtained through gait detection and step length estimation.

Among them, the method of transforming the acceleration coordinate system is:

In the formula, the navigation system n is defined to rotate around the Z axis (positive axis upward) by an angle of y (yaw angle, yaw), then rotate around the X axis by an angle of p (pitch angle, pitch), and finally rotate around the Y axis by an angle of γ (roll angle, roll) to get the cosine matrix of the direction from the navigation system to the vehicle system And the direction cosine matrix from the carrier system to the navigation system />

Gait detection includes:

1) Calculate the overall acceleration, and reduce the influence of the sensor attitude by calculating the overall acceleration a. In the formula, a _x , a _y , a _z are the collected three-axis acceleration values.

2) Obtain the potential peak in the sliding window, and use the acceleration threshold [1.2g, 3g] for initial judgment.

3) Calculate the time difference between the potential peak and the previous peak, and use the time threshold range of one step [0.4s, 1s] for secondary judgment.

4) Compare the potential peak with the acceleration of the front and back neighbors, and perform three judgments to remove the false peak. If the potential peak point is the maximum value, the algorithm will count one step, otherwise, no step counting process will be performed.

The step size is estimated as:

Where SL is the step length, a _max and a _min are the maximum and minimum values of the pedestrian's longitudinal acceleration, respectively.

Step 4: Process the multi-source information collected in step 1 to construct a pedestrian motion model; obtain pedestrian motion information, including: position, speed; according to the extended Kalman filter algorithm, construct a fusion model based on the extended Kalman filter; The multi-source information obtained in step 1 and the position information of pedestrians obtained in step 2 are fused to obtain the fused position information; the update of the state vector and the optimal estimation of the error value in the extended Kalman filter are completed.

Extended Kalman filter algorithms include:

The state equation is:

X(k+1)=φX(k)+ΓW(k)

Where X(k)=[N(k)E(k)v _n (k)v _e (k)] ^T , N(k) and E(k) are respectively the north and east positions at time k State value, v _n (k) and v _e (k) are the speed state values in the north direction and east direction at time k respectively, W(k) is the system noise at time k, φ and Γ are respectively State transition matrix and system noise coefficient matrix at time 1.

The observation equation is:

Z(k+1)=H(k+1)X(k+1)+V

Where Z(k)=[N _wifi E _wifi s] ^T , N _wifi and E _wifi are the north and east coordinates obtained through Wi-Fi, respectively, and s is the plane within two seconds obtained through gait detection displacement value. H(k+1) is the measurement matrix, which can be derived from the formula and obtained after linearization. V is the observation noise, and it is assumed that the noise obeys a Gaussian distribution.

The method for updating the state vector in the extended Kalman filter includes the following steps:

Step 4-1, one-step state prediction update:

in, is the optimal value of the previous state, /> is the one-step forecast value of the current state;

Step 4-2, one-step forecast estimation error covariance matrix update:

Among them, P _k+1|k is Corresponding covariance one-step forecast value, φ _k+1|k is the state transition matrix, P _k|k is /> Corresponding covariance, /> is the transpose of φ _k+1|k , Q _k is the process noise covariance matrix;

Step 4-3, calculate the extended Kalman filter gain:

Among them, K _k+1 is the extended Kalman filter gain matrix, H _k+1 is the observation matrix, is the transposition of the observation matrix, and R _k+1 is the measurement noise covariance matrix;

Step 4-4, calculate the innovation from the observation vector, and update the state estimate:

in, is the optimal estimated value of the current state, Z _k+1 is the current observed value, /> is the predicted value for the current observation. Steps 4-5, update the estimated error covariance:

P _k+1 ＝[IK _k+1 H _k+1 ]P _k+1|k

Among them, P _k+1 is The corresponding covariance, I is the identity matrix.

Step 5: Use the location information based on extended Kalman filter obtained in step 4, according to the particle filter algorithm, and add map matching at the same time, further correct and update the location information of pedestrians, and complete the indoor navigation of pedestrians with multi-source information fusion.

Particle filter algorithms include:

The particle filter uses random samples to directly estimate the approximate state posterior probability density function:

Among them, w ⁱ (k) is the weight value of particle x ⁱ _k , the sum of all values is 1, and δ(.) is the Dirac function.

The update method in particle filtering consists of the following steps:

Step 5-1, initialization.

Taking the initial value x ⁱ (k) of the observed value as the initial value of the probability density function x(0), by right /> Sampling is carried out N times, and the initial w ⁱ (k) is set to />

Step 5-2, one-step prediction.

For each particle ^xi (k), a new particle ^xi (k+1) is obtained by converting the formula p(x(k+1)| ^xi (k)).

Step 5-3, importance sampling.

For any particle x ⁱ (k+1), solve their weight value w ⁱ (k+1)=p(z(k+1)| ^xi (k+1)).

Step 5-4, normalization.

Normalize the weights.

Step 5-5, resampling.

Particles are resampled according to the weight size, the probability of repetition of particles with large weight is higher, and the probability of repetition of particles with smaller weight is small.

Step 5-6, particle update.

make and />

In steps 5-7, the basic particle filter algorithm is used for calculation.

Use the above calculation results to complete pedestrian indoor navigation.

The above method will be explained in detail below in conjunction with the accompanying drawings and specific embodiments.

As shown in Figure 1, this embodiment provides a pedestrian indoor navigation method based on multi-source information fusion. The specific technical points of the method are as follows:

S1. Provide a Wi-Fi positioning location model building function.

It can obtain all the sensor information in the navigation system through the built-in sensor of the mobile phone, and build a pedestrian position model by selecting the position and signal strength information.

The key to multi-source information fusion technology is to determine the relationship between multi-source information. In the navigation system, multi-source information mainly includes three-axis angular velocity, acceleration, Wi-Fi signal strength, magnetic field strength, etc. Constructing the pedestrian position model specifically includes:

in Represents the estimated position, x(t _k ), y(t _k ) represent the position coordinates corresponding to the first K position points, ω _j (t _k ) represents the weight coefficient of the corresponding data point, and the weight coefficient based on signal strength can be expressed/ > RSSI is signal strength.

S2. Extend the Kalman filter fusion model building function.

Using the established model, the extended Kalman filter algorithm is used to fuse position, velocity and other information to calculate the real-time position result.

The traditional Kalman filter is only suitable for linear systems, and the extended Kalman filter is a kind of Kalman filter algorithm, which applies the traditional Kalman filter to the nonlinear field. The system matrix F and the observation matrix in the traditional Kalman filter H will be replaced by f(x) and h(x), and the nonlinear system is then linearized according to a first-order Taylor expansion.

The system model of the extended Kalman filter is:

X _k+1 ＝f(X _k )+w _k

Among them, X _k+1 is the state variable of the k+1th step, X _k is the state variable of the kth step, w _k is the process noise; f() is the nonlinear state function in the extended Kalman filter.

Methods to achieve linearization of the system matrix include:

Among them, F _k-1 is the Jacobian matrix of the partial derivative of f to X, is the posterior state estimation of step k-1, and X is the state variable.

The observation model of the extended Kalman filter is:

Z _k ＝h(X _k )+v _k

Among them, Z _k is the observation variable, v _k is the observation noise, h() is the measurement function in the extended Kalman filter;

Methods to achieve linearization of the observation matrix include:

Among them, H _k is the Jacobian matrix of the partial derivative of h(X) to X, is the state estimate for the kth step.

As shown in Figure 2, Figure 2 shows the construction process of the extended Kalman filter fusion model of the present invention. In this embodiment, the error model is constructed using multi-source information such as angular velocity, acceleration, and position. The specific extended Kalman filter update mainly includes the following five steps:

Step 1, one-step state prediction update:

in, is the optimal value of the previous state, /> is the one-step forecast value of the current state;

Step 2, one-step forecast estimation error covariance matrix update:

Among them, P _k+1|k is Corresponding covariance one-step forecast value, φ _k+1|k is the state transition matrix, P _k|k is /> Corresponding covariance, /> is the transpose of φ _k+1|k , Q _k is the process noise covariance matrix;

Step 3, calculate the extended Kalman filter gain:

Among them, K _k+1 is the extended Kalman filter gain matrix, H _k+1 is the observation matrix, is the transposition of the observation matrix, and R _k+1 is the measurement noise covariance matrix;

Step 4, calculate the innovation from the observation vector, and update the state estimation:

in, is the optimal estimated value of the current state, Z _k+1 is the current observed value, /> is the predicted value for the current observation.

Step 5, update the estimated error covariance:

P _k+1 ＝[IK _k+1 H _k+1 ]P _k+1|k

Among them, P _k+1 is The corresponding covariance, I is the identity matrix.

Based on the above five steps, the update of the state vector and the optimal estimation of the error value in the extended Kalman filter are completed.

S3, based on particle filter and map matching fusion model construction function.

Based on the estimation results of the above models, the particle filter is carried out on the basis of the extended Kalman filter fusion model and map information is added to reduce the amount of calculation, speed up the calculation, reduce the error, and improve the accuracy of the model.

Compared with Kalman filtering, the idea of particle filtering is based on the Monte Carlo method, which uses particle sets to represent probability and can be used in any form of state space model. Its core idea is to express its distribution through random state particles drawn from the posterior probability, which is a sequential importance sampling method. In simple terms, the particle filter method refers to the process of approximating the probability density function by finding a group of random samples propagated in the state space, and replacing the integral operation with the sample mean value, so as to obtain the minimum variance distribution of the state. The samples here refer to particles, which can approach any form of probability density distribution when the number of samples is N→∞.

The state transition equation and observation equation of the particle filter are as follows:

x(t)=f(x(t-1),u(t),w(t))

y(t)=f(x(t),e(t))

Among them, x(t) is the state at time t, u(t) is the control quantity, w(t) and e(t) are the state noise and observation noise respectively, and the particle filter is obtained from the observation y(t) and the state x at the last time (t-1), u(t), w(t) filter out the state x(t) at time t.

Referring to Figure 3, the update method in the particle filter mainly includes the following steps:

Step 1, initialization.

Taking the initial value x ⁱ (k) of the observed value as the initial value of the probability density function x(0), by right /> Sampling is carried out N times, and the initial w ⁱ (k) is set to />

Step 2, one-step prediction.

For each particle ^xi (k), a new particle ^xi (k+1) is obtained by converting the formula p(x(k+1)| ^xi (k)).

Step 3, importance sampling.

For any particle x ⁱ (k+1), solve their weight value w ⁱ (k+1)=p(z(k+1)| ^xi (k+1)).

Step 4, normalization.

Normalize the weights.

Step 5, resampling.

Particles are resampled according to the weight size, the probability of repetition of particles with large weight is higher, and the probability of repetition of particles with smaller weight is small.

Step 6, update the particles and add map information at the same time.

make and />

Due to the limitation of the indoor walking area, map matching is performed according to the map information. When the one-step predicted particle crosses the passable area (such as crossing the wall), its weight value is assigned to 0:

In step 7, the basic particle filter algorithm is used for calculation.

S4. A pedestrian indoor navigation method based on multi-source information fusion.

After the above three models are constructed, the pedestrian indoor navigation method based on multi-source information fusion is basically generated. When the satellite navigation fails, the navigation model is introduced to predict the error, correct the positioning error caused by various factors to the inertial navigation system in real time, and improve the positioning accuracy of indoor pedestrians.

After the Wi-Fi positioning model, the extended Kalman filter fusion model and the particle filter and map matching fusion model are constructed, the pedestrian indoor navigation method based on multi-source information fusion proposed by the present invention is basically completed. The specific flow chart is shown in Figure 1 shown. In the pedestrian indoor navigation system, the multi-source information is used to construct the extended Kalman filter fusion model and the particle filter and map matching fusion model. The navigation method of the present invention performs indoor navigation of pedestrians to obtain the current position information of pedestrians.

In summary, compared with the prior art, the method of this embodiment has at least the following advantages and beneficial effects:

(1) The method of this embodiment improves the traditional PDR positioning by introducing extended Kalman filter and particle filter, and adds map matching to reduce the amount of computation.

(2) The method of this embodiment is based on the pedestrian indoor navigation method based on multi-source information fusion, which can improve the efficiency and accuracy of pedestrian navigation and positioning.

(3) The model and method in the method of this embodiment can be applied to most indoor navigation systems for pedestrians.

This embodiment also provides a pedestrian indoor navigation device based on multi-source information fusion, including:

at least one processor;

at least one memory for storing at least one program;

When the at least one program is executed by the at least one processor, the at least one processor implements the above method.

A pedestrian indoor navigation device based on multi-source information fusion in this embodiment can execute a pedestrian indoor navigation method based on multi-source information fusion provided by the method embodiment of the present invention, and can execute any combination of implementation steps of the method embodiment , have the corresponding functions and beneficial effects of the method.

The embodiment of the present application also discloses a computer program product or computer program, where the computer program product or computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the method shown in FIG. 1 .

This embodiment also provides a storage medium, which stores an instruction or program that can execute a pedestrian indoor navigation method based on multi-source information fusion provided by the method embodiment of the present invention. When the instruction or program is executed, it can execute Any combination of implementation steps of the method embodiments has the corresponding functions and beneficial effects of the method.

In some alternative implementations, the functions/operations noted in the block diagrams may occur out of the order noted in the operational diagrams. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality/operations involved. Furthermore, the embodiments presented and described in the flowcharts of the present invention are provided by way of example in order to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logical flow presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and in which sub-operations described as part of larger operations are performed independently.

Furthermore, although the invention has been described in the context of functional modules, it should be understood that one or more of the described functions and/or features may be integrated into a single physical device and/or unless stated to the contrary. or software modules, or one or more functions and/or features may be implemented in separate physical devices or software modules. It will also be appreciated that a detailed discussion of the actual implementation of each module is not necessary to understand the present invention. Rather, given the attributes, functions and internal relationships of the various functional blocks in the devices disclosed herein, the actual implementation of the blocks will be within the ordinary skill of the engineer. Accordingly, those skilled in the art can implement the present invention set forth in the claims without undue experimentation using ordinary techniques. It is also to be understood that the particular concepts disclosed are illustrative only and are not intended to limit the scope of the invention which is to be determined by the appended claims and their full scope of equivalents.

If the functions described above are realized in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the essence of the technical solution of the present invention or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes. .

The logic and/or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium, For use with instruction execution systems, devices, or devices (such as computer-based systems, systems including processors, or other systems that can fetch instructions from instruction execution systems, devices, or devices and execute instructions), or in conjunction with these instruction execution systems, devices or equipment used. For the purposes of this specification, a "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use in or in conjunction with an instruction execution system, device or device.

More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connection with one or more wires (electronic device), portable computer disk case (magnetic device), random access memory (RAM), Read Only Memory (ROM), Erasable and Editable Read Only Memory (EPROM or Flash Memory), Fiber Optic Devices, and Portable Compact Disc Read Only Memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program can be printed, as it may be possible, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or other suitable processing if necessary. The program is processed electronically and stored in computer memory.

It should be understood that various parts of the present invention can be realized by hardware, software, firmware or their combination. In the embodiments described above, various steps or methods may be implemented by software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented by any one or combination of the following techniques known in the art: Discrete logic circuits, ASICs with suitable combinational logic gates, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

In the above description of this specification, the description with reference to the terms "one embodiment/example", "another embodiment/example" or "some embodiments/example" means that the description is described in conjunction with the embodiment or example. A particular feature, structure, material, or characteristic is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the described specific features, structures, materials or characteristics may be combined in any suitable manner in any one or more embodiments or examples.

Although the embodiments of the present invention have been shown and described, those skilled in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the invention is defined by the claims and their equivalents.

The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above-mentioned embodiments, and those skilled in the art can also make various equivalent deformations or replacements without violating the spirit of the present invention. Equivalent modifications or replacements are all within the scope defined by the claims of the present application.

Claims (10)

1. The pedestrian indoor navigation method based on multi-source information fusion is characterized by comprising the following steps of:

constructing map information of the current indoor environment, selecting a reference point on a map, and constructing a Wi-Fi fingerprint library;

acquiring multi-source information by using a sensor on an intelligent terminal, wherein the multi-source information comprises Wi-Fi signal strength, angular velocity, speed, acceleration and gesture information;

constructing a pedestrian position model according to the Wi-Fi signal intensity and the Wi-Fi fingerprint library to obtain pedestrian position information; judging the instantaneous gesture of the intelligent terminal through the gesture information, converting the acceleration from a carrier coordinate system to a navigation coordinate system, and obtaining the movement distance of the pedestrian through gait detection and step estimation;

constructing a motion model of the pedestrian according to the multi-source information and the motion distance to obtain motion information of the pedestrian; according to an extended Kalman filtering algorithm, a fusion model based on the extended Kalman filtering is constructed, multi-source information and pedestrian position information are fused, fused position information is obtained, and state vector updating and optimal estimation of error values in the extended Kalman filtering are completed;

and further correcting and updating the pedestrian position information according to the map information and the particle filter algorithm by utilizing the fused position information, so as to complete the pedestrian indoor navigation with multi-source information fusion.

2. The pedestrian indoor navigation method based on multi-source information fusion according to claim 1, wherein the expression of the pedestrian position model is:

wherein,,represents the estimated position, x (t _k )、y(t _k ) Representing the position coordinates, ω, corresponding to the first K position points _j (t _k ) Weight coefficients representing the corresponding data points, the weight coefficients based on signal strength may represent +.>RSSI is the signal strength.

3. The pedestrian indoor navigation method based on multi-source information fusion according to claim 1, wherein the acceleration is converted from the carrier coordinate system to the navigation coordinate system by:

wherein,,for navigating a directional cosine matrix from the navigation system to the carrier system, C _r C for rotating matrix around Y axis _p For a rotation matrix about the Z-axis, C _y For a rotation matrix about the X-axis, γ is the rotation angle about the Y-axis, p is the rotation angle about the X-axis, and Y is the rotation angle about the Z-axis.

4. The pedestrian indoor navigation method based on multi-source information fusion of claim 1, wherein the step of gait detection comprises:

calculating the overall acceleration:

wherein a is _x ，a _y ，a _z For the collected triaxial acceleration value, the influence of the sensor posture is reduced by calculating the overall acceleration a;

according to the overall acceleration, a potential peak value is obtained in the sliding window, and primary judgment is carried out by utilizing a preset acceleration threshold value;

calculating the time difference between the potential wave crest and the previous wave crest, and performing secondary judgment by using a preset walking one-step time threshold range;

and comparing the potential peak with the front and rear neighborhood acceleration, judging for three times to remove the pseudo peak, if the potential peak point is the maximum value, recording one step by the algorithm, otherwise, not performing step counting processing.

5. The pedestrian indoor navigation method based on multi-source information fusion of claim 1, wherein the step size estimation is:

where SL is the step size, a _max And a _min The maximum and minimum of the longitudinal acceleration of the pedestrian, respectively.

6. The pedestrian indoor navigation method based on multi-source information fusion of claim 1, wherein the extended kalman filtering algorithm comprises:

X _k+1 ＝f(X _k )+w _k

wherein X is _k+1 X is the state variable of the (k+1) th step _k Is the state variable of the kth step, w _k Is process noise; f () is a nonlinear state function in extended kalman filtering;

the method for realizing the linearization of the system matrix comprises the following steps:

wherein Z is _k To observe the variable, v _k H () is a measurement function in extended kalman filtering for observing noise;

the method for realizing the linearization of the observation matrix comprises the following steps:

wherein H is _k Is a jacobian matrix of h (X) to X bias,and (5) estimating the state of the kth step.

7. The pedestrian indoor navigation method based on multi-source information fusion of claim 1, wherein the step of expanding the state vector update in the kalman filter comprises:

one-step state prediction update:

in the method, in the process of the invention,is the optimal value of the last state, +.>A one-step predicted value for the current state;

one-step predictive estimation error covariance matrix update:

wherein P is _k+1|k Is thatCorresponding covariance one-step predictor, phi _k+1|k For state transition matrix, P _k|k Is->Corresponding covariance,/>Is phi _k+1|k Transpose of Q _k A process noise covariance matrix;

calculating an extended Kalman filtering gain:

wherein K is _k+1 To expand the Kalman filter gain matrix, H _k+1 In order to observe the matrix,for transpose of the observation matrix, R _k+1 Measuring a noise covariance matrix;

calculating information from the observation vectors and updating the state estimate:

in the method, in the process of the invention,z is the optimal estimated value of the current state _k+1 For the current observation +.>The current observation predicted value;

updating the estimated error covariance:

P _k+1 ＝[I-K _k+1 H _k+1 ]P _k+1|k

wherein P is _k+1 Is thatCorresponding covariance, I is the identity matrix.

8. The pedestrian indoor navigation method based on multi-source information fusion according to claim 1, wherein the further correcting and updating of the pedestrian position information according to the map information and the particle filter algorithm comprises:

the state transfer equation and the observation equation for particle filtering are as follows:

x(t)＝f(x(t-1),u(t),w(t))

y(t)＝f(x(t),e(t))

wherein x (t) is a state at time t, u (t) is a control quantity, w (t) and e (t) are state noise and observation noise respectively, and particle filtering filters the state at time t x (t) from observation y (t) and the state at the last time x (t-1), u (t) and w (t);

the updating method in particle filtering mainly comprises the following steps:

a1, initializing: initial value x of observed value ⁱ (k) As an initial value of the probability density function x (0), byFor->Sampling N times, initial w ⁱ (k) Set to->

A2, one-step prediction: for each particle x ⁱ (k) By converting the formula p (x (k+1) |x ⁱ (k) Obtaining a new particle;

a3, importance sampling: for any particle x ⁱ (k+1) solving for their weights w ⁱ (k+1)＝p(z(k+1)|x ⁱ (k+1))；

A4, normalization: and (3) carrying out normalization processing on the weight values:

a5, resampling: resampling the particles according to the weight, wherein the probability of repeating the particles with the weight is high, and the probability of repeating the particles with the weight is low;

a6, updating particles and simultaneously adding map information:

order theIs->

Because of the limit value of the indoor walking area, map matching is carried out according to map information, and when the particles predicted in one step cross the passable area, the weight value of the particles is given as 0:

and A7, calculating by adopting a basic particle filtering algorithm, and completing the indoor positioning and navigation of the pedestrians according to the calculation result.

9. A pedestrian indoor navigation device based on multi-source information fusion, comprising:

at least one of a processor;

at least one memory for storing at least one program;

the at least one program, when executed by the at least one processor, causes the at least one processor to implement the method of any one of claims 1-8.

10. A computer readable storage medium, in which a processor executable program is stored, characterized in that the processor executable program is for performing the method according to any of claims 1-8 when being executed by a processor.