Indoor target positioning method and system based on inertial navigation unit and map data matching

Abstract

The invention discloses an indoor target positioning method based on matching of an inertial navigation unit and map data. The indoor target positioning method mainly solves the problem that an existing inertial navigation unit cannot realize long-time accurate positioning in an indoor environment. According to the implementation scheme, an indoor path model is established, and related information of a path and a path intersection is recorded; establishing a measurement probability model and a transition probability model according to the recorded related information and the parameters of the inertial navigation unit; calculating the matching probability of the target position information to each path and the direction information of the first three probability matching paths of the target position information by using a measurement probability model, and calculating a transition probability matrix of the target transferred among the paths by using a transition probability model; calculating a real-time moving track of the target by using a Viterbi algorithm; and calculating position information of the target according to the target moving track. According to the invention, data failure caused by overlarge accumulative error in long-time positioning of the inertial navigation unit is reduced, the precision of the positioning system is improved, and the method can be used for vehicle positioning in the underground parking lot.

Description

Technical Field

[0001] The invention belongs to the technical field of inertial navigation, and in particular relates to an indoor positioning method and system, which can be used for indoor vehicle positioning in an underground parking lot. Background Art

[0002] Location information has always been one of the many essential information in human production and life. With the continuous improvement of science and technology, various types of robots such as drones and unmanned vehicles have appeared in large numbers in various positions, and the demand for precise positioning and navigation has continued to increase. Unlike the outdoor environment, the indoor environment is more complex, and indoor positioning solutions based on wireless technology have problems such as high cost and poor stability. The positioning solution based on inertial navigation technology can achieve positioning and navigation functions without the help of any external information, but its positioning accuracy is poor. In order to further improve the indoor positioning accuracy, based on the strapdown inertial navigation technology, inertial navigation has been deeply analyzed and studied. On the basis of the research on the traditional use of inertial navigation systems to achieve indoor positioning functions, it is necessary to compensate and optimize the error factors that affect positioning accuracy in the inertial navigation system.

[0003] Patent document No. 202310274750.7 discloses a mobile robot navigation system based on solid-state laser radar and point cloud map, which uses solid-state laser radar and odometer and point cloud map tightly coupled with IMU to update the target posture, eliminates motion distortion for the original radar point cloud through high-frequency IMU data, outputs dedistorted point cloud, matches the dedistorted point cloud with the accumulated point cloud, calculates the residual to update the posture state, and uses the dedistorted point cloud output by the laser radar inertial odometer module to splice into a sliding window structure to update the robot's posture in the map coordinate system. Since this method does not use point cloud map data to establish a connection between target position information and posture information, it ignores the solution that can further correct the position information based on posture information, so the accuracy of obtaining position information is low.

[0004] The patent document with application number 202311089118.1 discloses an indoor path planning method based on two-dimensional lidar and binocular camera. It adds IMU inertial unit and wheel encoder as corrections to the lidar framework, corrects the distortion of the lidar data, and then uses the Hector-based grid map construction method to complete the construction of the global map. The lidar data is then responsible for the implementation of the global path planning algorithm. In this method, the use of lidar for a long time or in a complex environment will cause cumulative errors, and because the map information feedback is not used to adjust the target's posture information, long-term use will inevitably cause the target's posture data to become invalid. Summary of the invention

[0005] The purpose of the present invention is to address the defects of the above-mentioned prior art and propose an indoor positioning method based on matching an inertial navigation unit with map data, so as to improve the accuracy of position information, eliminate the cumulative errors caused by long-term operation of inertial navigation, and achieve the goal of indoor positioning by matching the inertial navigation unit with map data.

[0006] The key technology to achieve the purpose of the present invention is to eliminate the displacement accumulation error caused by long-term operation of the inertial navigation unit by matching the map data with the target posture data. The method provides an indoor target positioning method in which a vehicle-mounted inertial navigation unit matches the map data, comprising the following steps:

[0007] Establish an indoor path model according to the indoor environment, and record the position O and turning angle of the path intersection on the indoor path model. and the direction L of each path;

[0008] According to the indoor environment and inertial navigation unit parameters, the measurement probability model p and the transfer probability model n are established;

[0009] Use the inertial navigation unit to measure the indoor target posture data, and substitute it into the measurement probability model and the transfer probability model to calculate the real-time movement trajectory of the target and determine the current location of the target;

[0010] The target heading angle change is measured using an inertial navigation unit and compared with the steering angle of a recorded path intersection near the target. When the two are close, the target position is updated to the path intersection position and the indoor positioning data of the target is corrected.

[0011] Furthermore, the indoor path model is established according to the indoor environment by first drawing an indoor path image M according to the indoor path, and then marking the path label E and the path intersection label P in M:

[0012] E=e 1 ,e 2 ,…,e i ,…,e m ,

[0013] P=p 1 ,p 2 ,…,p j ,…,p n ,

[0014] where e i represents the label of the i-th path, m represents the number of paths; p j represents the label of the jth path intersection, and n represents the number of path intersections.

[0015] O=(o 1x ,o1y ),(o 2x ,o 2y ),…,(o jx ,o jy ),…,(o nx ,o ny ),

[0016]

[0017] L = l 1 ,l 2 ,…,l i ,…,l m ,

[0018] Among them (o jx ,o jy ) represents the jth path intersection point p j The horizontal and vertical coordinates of the location; is the jth path intersection point p j Two steering angle values ​​at i represents the i-th path e i The path direction.

[0019] Furthermore, the measurement probability model p and the transition probability model n are established according to the indoor environment and the inertial navigation unit parameters, and the implementation thereof includes the following:

[0020] (4a) An error calibration experiment is conducted on the inertial navigation unit used. The straight-line driving is used as the experimental content, and the probability density distribution curve of the displacement error is drawn to establish the measurement probability model p:

[0021]

[0022] where x ij Indicates the straight-line distance between two consecutive target position measurement data, z ij Indicates the path e between two adjacent target positions that may match i With e j The path distance between the two paths, λ is the exponential parameter, Δθ represents the change value of the target heading angle within 5s, l i and l j Respectively represent the path e i With e j The path direction.

[0023] Furthermore, the real-time moving trajectory of the target is calculated to determine the current location of the target, and its implementation includes the following:

[0024] (5a) During the target movement process, the target position information and attitude information measured by the inertial navigation unit are substituted into the measurement probability model to obtain the matching probability of the target position information for each path;

[0025] (5b) Substitute the target position information, attitude information and direction information of the top three probability matching paths of the target position information measured by the stored inertial navigation unit into the transition probability model n, and calculate the state transfer matrix N:

[0026]

[0027] where n ia represents the transition probability model calculated from path e i To e a The path transition probability;

[0028] (5c) Taking the matching probability of the target location information to each path and the state transition matrix N as input, the Viterbi algorithm is used to calculate the real-time trajectory of the target and determine the current location of the target.

[0029] Further, the updating of the target position to the intersection position of the path is performed based on whether the target heading angle measured by the inertial navigation unit has a significant change:

[0030] When the target heading angle measured by the inertial navigation unit changes significantly, the steering angle information recorded at the intersection of the path closest to the target is queried (φ j1 ,φ j2 );

[0031] When the target's heading angle change value Δθ and φ within 5 seconds j1 or φ j2 When they are close, update the target position to the position coordinates of the intersection point of the path (o jx ,o jy ).

[0032] The present invention also provides an indoor target positioning system for matching a vehicle-mounted inertial navigation unit with map data, which comprises:

[0033] The inertial navigation unit module is used to obtain the position information and attitude information of the target. The acceleration value and angular velocity of the target are obtained through the accelerometer and gyroscope, and the position and attitude information of the target is obtained after integral calculation;

[0034] The map matching positioning module is used to calculate the target movement trajectory, determine the current location of the target, substitute the stored target posture information into the measurement probability model and the transition probability model, and use the output result as the input of the Viterbi algorithm to calculate the target location;

[0035] The position correction module is used to correct the position information of the target, that is, to compare the change value of the target heading angle with the recorded path intersection angle information, and to correct the target position to the position of the recorded path intersection when the two values ​​are close.

[0036] The present invention also provides an electronic device, including a memory, a processor, an inertial navigation unit, an LCD screen, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the indoor target positioning method of matching the inertial navigation unit with map data as described above is implemented.

[0037] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the indoor target positioning method of matching the inertial navigation unit with map data.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] Firstly, since the present invention adds the combination of target posture information into the model when establishing the measurement probability model and the transition probability model, it not only improves the matching probability of the target position information to the correct path calculated by the measurement probability model, but also improves the probability of the target transferring from a certain path to the correct path calculated by the transition probability model, thereby improving the accuracy of the subsequent calculation of the target real-time trajectory probability using the Viterbi algorithm and the accuracy of the position information.

[0040] Secondly, when establishing an indoor path model, the present invention records the directional attribute information of the path and the angle information of the path intersection. Therefore, such angle information can be compared with the posture information of the target to update the real-time position of the target and reduce the cumulative error caused by the long-term operation of the inertial navigation unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of a method for matching inertial navigation and map data provided by an embodiment of the present invention;

[0042] Figure 2 is an indoor path image of an underground parking lot provided by an embodiment of the present invention;

[0043] Figure 3 is a sub-flow chart of a map matching algorithm in an embodiment of the present invention;

[0044] Figure 4 is a structural block diagram of an inertial navigation and map data matching system provided by an embodiment of the present invention;

[0045] Figure 5is a structural block diagram of an electronic device provided by an embodiment of the present invention;

[0046] Figure 6 It is a simulation effect diagram of the present invention. DETAILED DESCRIPTION

[0047] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0048] Reference Figure 1 The inertial navigation and map data matching method of this example includes the following steps:

[0049] Step 1: Establish an indoor path model according to the indoor environment and mark it.

[0050] The indoor path includes a road in the indoor space that can be passed by vehicles and pedestrians.

[0051] The specific implementation of this step includes:

[0052] 1.1) Actually check the number and distribution of indoor paths, and draw an indoor path image M based on it;

[0053] 1.2) Label the paths in the path image M and mark the path labels E:

[0054] E=e 1 ,e 2 ,…,e i ,…,e m ,

[0055] Among them, e i represents the label of the i-th path, and m represents the number of paths;

[0056] 1.3) Label the path intersections in the path image M and mark the path label P:

[0057] P=p 1 ,p 2 ,…,p j ,…,p n ,

[0058] Among them, p j represents the label of the jth path intersection, and n represents the number of path intersections.

[0059] This example depicts the indoor path image of an underground parking lot. Figure 2 As shown. Figure 2 In the figure, the solid lines are paths, the dashed points are path intersections, and there is a path label E for the path on one side of each path, and a label P for the path intersection on the side of each path intersection.

[0060] Step 2, record relevant information about the paths and path intersections.

[0061] The path intersection includes the intersection position of paths in the indoor space.

[0062] 2.1) After constructing the indoor path image M, establish a two-dimensional rectangular coordinate system U with a certain point Q in the room as the origin, the north direction as the positive direction of the y-axis, and the east direction as the positive direction of the x-axis;

[0063] 2.2) Using the two-dimensional rectangular coordinate system U as a reference, record the coordinate information O of the path intersection:

[0064] O=(o 1x ,o 1y ),(o 2x ,o 2y ),…,(o jx ,o jy ),…,(o nx ,o ny )

[0065] Among them (o jx ,o jy ) represents the jth path intersection point p j The horizontal and vertical coordinates of the location;

[0066] 2.3) The intersection angle of the two paths at the actual measurement path intersection

[0067]

[0068] in, represents the jth path intersection point p j The steering angle value between the two paths at ;

[0069] 2.4) With point Q as the origin and due north as 0 degrees, establish a two-dimensional polar coordinate system R;

[0070] 2.5) Using the two-dimensional polar coordinate system R as a reference, mark the path direction L in the indoor path image M:

[0071] L = l 1 ,l 2 ,…,l i ,…,l m ,

[0072] Among them, l i represents the i-th path e i The path direction.

[0073] Step 3: Establish a measurement probability model and a transfer probability model based on the recorded information about the paths and path intersections and the parameters of the inertial navigation unit.

[0074] The inertial navigation unit parameters include the size of the displacement error that can be generated during straight-line movement. The inertial navigation unit is subjected to a displacement error calibration experiment, with a straight-line driving of 70 meters as the experimental content, and its experimental data is collected.

[0075] Reference Figure 3 , the implementation of this step includes the following:

[0076] 3.1) Based on the collected experimental data, a probability density distribution curve H of the distance difference between the target measured position and the actual position is drawn;

[0077] 3.2) In view of the similarity between the shape of the probability density distribution curve H and the shape of the Gaussian distribution function curve, the Gaussian distribution function is chosen to fit the probability density distribution curve H. The Gaussian distribution function G is expressed as follows:

[0078]

[0079] Among them, σ is the standard deviation of the Gaussian function, μ is the mean value of the Gaussian function, and x is the independent variable of the Gaussian distribution function;

[0080] 3.3) Since the center point of the probability density distribution curve H is at the 0 coordinate position, the average value μ of the Gaussian distribution function G is set to 0, so that the center position of the Gaussian distribution function G coincides with the center position of the probability density distribution curve H;

[0081] 3.4) adjusting the standard deviation value σ of the Gaussian distribution function G so that the shape of the curve of the Gaussian distribution function G is closest to the shape of the probability density distribution curve H;

[0082] 3.5) Modify the expression of Gaussian distribution function G by adding the target heading angle information measured by the inertial navigation unit and the i-th path e i Direction information, get the observation probability model P:

[0083]

[0084] Where, θ represents the target heading angle information measured by the inertial navigation unit, x i Indicates the path from the target measurement position to the i-th path e i The distance, l i represents the i-th path e i Direction information;

[0085] 3.6) Substitute the target position information and heading angle information measured by the inertial navigation unit into the observation probability model P to obtain the matching probability of the target position information for each path, record the top three with the highest matching probabilities, and obtain the observation probability matrix Q:

[0086]

[0087] Among them, q i Indicates the target location information for the i-th path p i The matching probability of q j Indicates the target location information for the jth path p j The matching probability of q k Indicates the target location information for the kth path p k The matching probability of

[0088] 3.7) Plotting the probability density distribution curve r of the straight-line distance between two adjacent target position data and the path distance difference of the path with the maximum probability matching path;

[0089] 3.8) In view of the similarity between the shape of the probability density distribution curve r and the shape of the exponential function curve, the exponential function is selected to fit the probability density distribution curve r. The exponential function s is expressed as follows:

[0090] s=λe -λ|x|

[0091] Among them, λ represents the index and x represents the independent variable;

[0092] 3.9) Adjust the exponent λ of the exponential function s so that the curve shape of the exponential function s is closest to the shape of the probability density distribution curve r;

[0093] 3.10) Modify the expression of the exponential function s by adding the target heading angle change information measured by the inertial navigation unit and the i-th path e i With the jth path e j Direction i and l j , and rewrite the independent variable x as x ij -z ij , and obtain the observation probability model n:

[0094]

[0095] Among them, λ represents the exponent, x ij Represents the distance between two adjacent target position data, z ij Indicates the maximum probability matching path e of these two target location data i With e j The path distance between irepresents the i-th path e i Direction information, l j represents the jth path e j , Δθ represents the change of the target heading angle within 5 seconds.

[0096] Step 4: Calculate the real-time trajectory of the target and determine the current location of the target based on the target position information and attitude information measured by the inertial navigation unit and the established observation probability model and transition probability model.

[0097] 4.1) The target position information and attitude information measured by the inertial navigation unit are compared with the direction information of the top three probability matching paths of the target position information. i , l j , l k Substituting into the transition probability model, we get the state transfer matrix N:

[0098]

[0099] Among them, n ia represents the transition probability model calculated from the i-th path e i To the ath path e a The path transition probability, n ja represents the transition probability model calculated from the jth path e j To the ath path e a The path transition probability, n ka represents the transition probability model calculated from the kth path e k To the ath path e a The path transition probability, n ib represents the transition probability model calculated from the i-th path e i To the bth path e b The path transition probability, n jb represents the transition probability model calculated from the jth path e j To the bth path e b The path transition probability, n kb represents the transition probability model calculated from the kth path e k To the bth path e b The path transition probability, n ic represents the transition probability model calculated from the i-th path e i To the cth path e c The path transition probability, n jc represents the transition probability model calculated from the jth path e j To the cth path e c The path transition probability, n kcrepresents the transition probability model calculated from the kth path e k To the cth path e c The path transition probability;

[0100] 4.2) Using the Viterbi algorithm, the observation probability matrix Q and the state transition matrix N are used as input to calculate the target's moving trajectory T:

[0101] T=ψ 1 ,ψ 2 ,…,ψ γ ,…,ψ δ ,

[0102] Among them, ψ γ represents the γth path passed by the target, γ∈[1,δ], δ represents the total number of paths passed by the target;

[0103] 4.3) The δth path ψ from the target's latest position measured by the inertial navigation unit to the target δ Make a perpendicular line and get the perpendicular line and path ψ δ The intersection point is the target position information F:

[0104] F=(x,y),

[0105] Among them, x represents the horizontal coordinate of the intersection position, and y represents the vertical coordinate of the intersection position;

[0106] Step 5: According to the target attitude information measured by the inertial navigation unit and the recorded path intersection position information O and steering angle Correct the target's position information.

[0107] When using the Viterbi algorithm to calculate the target's position information, it can only rely on the output results of the measurement probability model and the transition probability model for calculation, and the input variables of the measurement probability model and the transition probability model are only the original posture data measured by the inertial navigation unit. When the inertial navigation unit fails to provide relatively accurate posture data due to the accumulated error caused by long-term operation, the Viterbi algorithm will fail. Therefore, a method is needed to correct the original position information of the target. The specific implementation includes the following:

[0108] 5.1) When the target heading angle measured by the inertial navigation unit exceeds When the arc changes, the steering angle information recorded at the path intersection closest to the target position is queried (φ j1 ,φ j2 );

[0109] 5.2) Record the steering angle information (φ) at the intersection of the path j1 ,φ j2) is compared with the target heading angle information measured by the inertial navigation unit to correct the target position information:

[0110] If the absolute value of the angle change Δθ produced by the target heading angle within 5 seconds is equal to φ j1 or φ j2 Difference not more than radians, the target location information is updated to the location information recorded at the intersection of the path (o jx ,o jy );

[0111] Otherwise, do nothing.

[0112] Reference Figure 4 The inertial navigation and map data matching system provided in this example includes: an inertial navigation unit module 1, a map matching and positioning module 2 and a position correction module 3, wherein the map matching and positioning module 2 includes a measurement probability submodule 21, a transfer probability submodule 22 and a trajectory calculation submodule 23.

[0113] These modules work together and their working principle is as follows:

[0114] The inertial navigation unit module 1 obtains the position information and attitude information of the target, and obtains the acceleration value and angular velocity of the target through the accelerometer and gyroscope, and obtains the position and attitude information of the target after integral calculation; the position and attitude information is transmitted to the map matching positioning module 2, which is used to calculate the target movement trajectory, determine the current position of the target, substitute the stored target posture information into the measurement probability model and the transfer probability model, and use the output result as the input of the Viterbi algorithm to calculate the position of the target. Specifically, the measurement probability submodule 21 calculates the matching probability of the target position information for each path, substitutes the position data and attitude information measured by the stored inertial navigation unit as input into the measurement probability model, and calculates the matching probability of the target position information for each path. The matching probability is matched with the direction information of the top three probability matching paths of the target position information, and the matching probability of the top three probability matching paths of the target position information is transmitted to the trajectory calculation submodule 23. The direction information of the top three probability matching paths of the target position information is calculated through the transfer probability submodule 22 to calculate the transfer probability matrix N of the target transfer between each path, and the target position information and posture information measured by the stored inertial navigation unit and the direction information of the top three probability matching paths of the target position information are substituted into the transfer probability model n, and the transfer probability matrix N is transmitted to the trajectory calculation submodule 23 to calculate the moving trajectory of the target; the Viterbi algorithm is used to calculate the real-time moving trajectory of the target with the results calculated by the measurement probability module and the transfer probability module as input. The real-time moving trajectory of the target and the position and posture information of the target obtained by the inertial navigation unit module 1 are transmitted to the position correction module 3, and the position information of the target is corrected, and the change value of the target heading angle is compared with the recorded path intersection angle information. When the two values ​​are close, the target position is corrected to the position of the recorded path intersection.

[0115] Reference Figure 5 The electronic device provided by an embodiment of the present invention includes a memory, a processor, an inertial navigation unit, an LCD screen, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the indoor target positioning method of matching the inertial navigation unit with map data is implemented.

[0116] The processor may be a general-purpose central processing unit (CPU) or a microprocessor, wherein the microprocessor includes an MMU memory management unit for executing related programs to implement the technical solutions provided in the embodiments of this specification.

[0117] The memory may be implemented in the form of a read-only memory ROM, a static storage device, a dynamic storage device, etc., and is used to store an operating system and application programs. The technical solutions provided in the embodiments of this specification are implemented through software, and the relevant program codes are stored in the memory and called and executed by the processor.

[0118] The inertial navigation unit, which may be a 6-axis inertial navigation unit or a 9-axis inertial navigation unit, is used to measure the position information and attitude information of the target, provide the target's attitude information to the processor, and implement the technical solution provided in the embodiments of this specification through software.

[0119] The LCD screen, which may be an IPS LCD screen, a TFT LCD screen or a TN LCD screen, is used to receive the result output by the processor after the calculation is completed, so as to demonstrate the specific effect of the technical solution provided in the embodiments of this specification.

[0120] It should be noted that although the above device only shows a processor, a memory, an inertial navigation unit and an LCD screen, in the specific implementation process, the device may also include other components necessary for normal operation. This example provides a computer-readable storage medium, which stores multiple instructions that can be loaded by a processor to execute the steps in any indoor target positioning method for matching an inertial navigation unit with map data provided in an embodiment of the present invention. The computer storage medium includes, but is not limited to, a static random access memory (SRAM), a dynamic random access memory (DRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM) or other memory technology, which can be used to store information that can be accessed by a computing device.

[0121] It should be noted that computer-readable storage media include permanent media and non-permanent media, as well as removable media and non-removable media. These media can store information by any method or technology, and the information can be computer-readable instructions, data structures, program modules or other data.

[0122] The above descriptions are only a few specific examples of the present invention and do not constitute any limitation to the present invention. Obviously, for professionals in this field, after understanding the content and principles of the present invention, they may make various modifications and changes in form and details without departing from the principles and structures of the present invention. However, these modifications and changes based on the ideas of the present invention are still within the scope of protection of the claims of the present invention.

[0123] It should be noted that the step numbers in the specification and claims of the present invention are only for a clear description of the real-time solution of the present invention to facilitate understanding, and the order of the step numbers is not limited.

Claims

1. A method for indoor target positioning by matching a vehicle-mounted inertial navigation unit with map data, characterized in that: include: Establish an indoor path model according to the indoor environment, and record the position O and turning angle of the path intersection on the indoor path model. and the direction L of each path; According to the indoor environment and inertial navigation unit parameters, the measurement probability model p and the transfer probability model n are established; Use the inertial navigation unit to measure the indoor target posture data, and substitute it into the measurement probability model and the transfer probability model to calculate the real-time movement trajectory of the target and determine the current location of the target; The target heading angle change is measured using an inertial navigation unit and compared with the steering angle of a recorded path intersection near the target. When the two are close, the target position is updated to the path intersection position and the indoor positioning data of the target is corrected.

2. The method according to claim 1, characterized in that The indoor path model is established according to the indoor environment, which is to first draw an indoor path image M according to the indoor path, and then mark the path label E and the path intersection label P in M: E1,e2,…,e i ,…,the m , P=p1,p2,…,p j ,…,p n , where e i represents the label of the i-th path, m represents the number of paths; p j represents the label of the jth path intersection, and n represents the number of path intersections.

3. The method according to claim 1, characterized in that The recorded path intersection location information O, steering angle and the directions L of each path, which are expressed as follows: O=(o 1x ,o 1y ),(o 2x ,o 2y ),…,(o jx ,o jy ),…,(o nx ,o ny ), L＝l1,l2,…,l i ,…,l m ， Among them (o jx ,o jy ) represents the jth path intersection point p j The horizontal and vertical coordinates of the location; is the jth path intersection point p j Two steering angle values ​​at i represents the i-th path e i The path direction.

4. The method according to claim 1, characterized in that: The measurement probability model p and the transition probability model n are established according to the indoor environment and the inertial navigation unit parameters, and the implementation thereof includes the following: (4a) An error calibration experiment is conducted on the inertial navigation unit used. The straight-line driving is used as the experimental content, and the probability density distribution curve of the displacement error is drawn to establish the measurement probability model p: where σ z is the standard deviation, x i Indicates the measurement position of the target to the i-th path e i distance, θ represents the heading angle of the target, l i represents the i-th path e i The path direction. (4b) Substitute the target position information and attitude information measured by the inertial navigation unit into the measurement probability model to obtain the matching probability of the target position information for each path, and describe the distance x between two adjacent target position data. ij The path distance z to its most probable matching path ij Difference probability density distribution curve, establish the transition probability model n: where x ij Indicates the straight-line distance between two consecutive target position measurement data, z ij Indicates the path e between two adjacent target positions that may match i With e j The path distance between the two paths, λ is the exponential parameter, Δθ represents the change value of the target heading angle within 5s, l i and l j Respectively represent the path e i With e j The path direction.

5. The method according to claim 1, characterized in that The real-time moving trajectory of the target is calculated to determine the current location of the target, which is implemented as follows: (5a) During the target movement process, the target position information and attitude information measured by the inertial navigation unit are substituted into the measurement probability model to obtain the matching probability of the target position information for each path; (5b) Substitute the target position information, attitude information and direction information of the top three probability matching paths of the target position information measured by the stored inertial navigation unit into the transition probability model n, and calculate the state transfer matrix N: Where n ia represents the transition probability model calculated from path e i To e a The path transition probability; (5c) Taking the matching probability of the target location information to each path and the state transition matrix N as input, the Viterbi algorithm is used to calculate the real-time trajectory of the target and determine the current location of the target.

6. The method according to claim 1, characterized in that The updating target position is the intersection position of the path, which is performed based on whether the target heading angle measured by the inertial navigation unit has a significant change: When the target heading angle measured by the inertial navigation unit changes significantly, the steering angle information recorded at the path intersection closest to the target is queried (φ j1 ,φ j2 ); When the target's heading angle change value Δθ and φ within 5 seconds j1 or φ j2 When they are close, update the target position to the position coordinates of the intersection point of the path (o jx ,o jy ).

7. An indoor target positioning system that matches a vehicle-mounted inertial navigation unit with map data, characterized in that: include: The inertial navigation unit module is used to obtain the position information and attitude information of the target. The acceleration value and angular velocity of the target are obtained through the accelerometer and gyroscope, and the position and attitude information of the target is obtained after integral calculation; The map matching positioning module is used to calculate the target movement trajectory, determine the current location of the target, substitute the stored target posture information into the measurement probability model and the transition probability model, and use the output result as the input of the Viterbi algorithm to calculate the target location; The position correction module is used to correct the position information of the target, that is, to compare the change value of the target heading angle with the recorded path intersection angle information, and to correct the target position to the position of the recorded path intersection when the two values ​​are close.

8. The system according to claim 7, characterized in that The map matching and positioning module includes The measurement probability submodule is used to calculate the matching probability of the target position information for each path, substitute the position data and attitude information measured by the stored inertial navigation unit as input into the measurement probability model, calculate the matching probability of the target position information for each path and the direction information of the top three probability matching paths of the target position information; The transfer probability submodule is used to calculate the probability of the target transferring from one path to another. The target position information and attitude information measured by the inertial navigation unit and the direction information of the top three probability matching paths of the target position information are substituted into the transfer probability model to calculate the transfer probability matrix of the target transferring between various paths. The trajectory calculation submodule is used to calculate the moving trajectory of the target. It uses the Viterbi algorithm and takes the results calculated by the measurement probability module and the transfer probability module as input to calculate the real-time moving trajectory of the target.

9. An electronic device comprising a memory, a processor, an inertial navigation unit, an LCD screen, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the indoor target positioning method of matching the inertial navigation unit with map data as described in any one of claims 1 to 6 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the indoor target positioning method of matching an inertial navigation unit with map data as described in any one of claims 1 to 6.

Images