A seamless indoor and outdoor positioning method based on multi-source fusion

Abstract

The present invention discloses a method for seamless indoor and outdoor positioning based on multi-source fusion, the method comprising: constructing a factor graph model, obtaining pre-processed data values ​​of each positioning signal, inputting them into the constructed factor graph model according to the time sequence, expanding the factor graph through defined system variable nodes and factor nodes, implementing the recursion and update of the system state corresponding to the factor nodes of different sensors, and realizing the correction of inertial navigation positioning; finally obtaining accurate positioning coordinates and completing fusion positioning. The present invention solves the problem of seamless indoor and outdoor positioning; it can effectively reduce the seamless positioning error of multi-source heterogeneous data. And the present invention proposes a fixed rate factor graph method in which the variable nodes are generated at a fixed rate and are not affected by the measured value. Asynchronous measurement can be associated with the corresponding variable nodes in the graph for optimization. The method avoids the time span reduction and frequent optimization of the traditional factor graph, improves the estimation accuracy, and reduces the amount of calculation.

Description

Technical Field

[0001] The present invention relates to the technical field of computers and location services, and in particular to an indoor and outdoor seamless positioning method based on multi-source fusion, an intelligent terminal and a storage medium. Background Art

[0002] In the era of the Internet of Things, indoor and outdoor positioning plays various important roles to determine the location of people, mobile devices, and various robots. Due to the popularity of social networks and the widespread use of mobile devices, the demand for location-based services (LBS) in indoor and outdoor environments has increased significantly. In addition, with the popularization of the Internet and the rapid increase in communication network infrastructure, human daily life is also closely related to positioning technology. Therefore, the value of seamless indoor and outdoor positioning technology is highlighted.

[0003] Indoor and outdoor positioning can be applied to universities, airports, military systems, security systems, subway service route maps, agriculture, forests and other fields. Applications such as emergency management, tracking mobile users, context perception, advertising, environmental monitoring, military monitoring, and health care are inseparable from positioning. Due to its important significance and application value, indoor and outdoor positioning has been valued by industry and academia.

[0004] In the prior art, single-source positioning systems have obvious disadvantages, whether indoors or outdoors. For example, the performance of inertial positioning systems degrades over time due to the accumulation of errors in inertial sensors. The Global Navigation Satellite System (GNSS) is one of the main means for autonomous systems with navigation needs to provide global reference positioning. However, in highly urbanized areas, the positioning accuracy is significantly reduced due to non-optical line of sight (NLOS) reception and multipath effects caused by signal reflections caused by static buildings and dynamic objects (such as double-decker buses). Video positioning has problems such as difficult data registration and high scene requirements. All types of sensors have their own problems.

[0005] Therefore, the existing navigation technology needs to be improved and developed. Summary of the invention

[0006] The technical problem to be solved by the present invention is that, in view of the above-mentioned defects of the prior art, a seamless indoor and outdoor positioning method, an intelligent terminal and a storage medium based on multi-source fusion are provided. In order to solve the problem of indoor and outdoor integrated positioning of pedestrians, robots (drainage pipe inspection robots, crawling robots) and drones, the present invention provides a positioning method based on a multi-source sensor fusion positioning framework based on a factor graph, which realizes high-precision positioning of pedestrians, robots and drones. The positioning method of the present invention has fast positioning efficiency and high positioning accuracy.

[0007] The technical solution adopted by the present invention to solve the problem is as follows:

[0008] A method for seamless indoor and outdoor positioning based on multi-source fusion, wherein the method comprises:

[0009] Based on the principle of factor graph, the factor graph model is constructed by presetting variable nodes, factor nodes and connections; and the corresponding relationship between each positioning signal and variable nodes and factor nodes is defined;

[0010] Acquire positioning signals of outdoor non-restricted spaces, wherein the positioning signals of outdoor non-restricted spaces include: inertial navigation signals, odometers, and control point signals; use the inertial navigation signals as the main signals of the positioning system, and other signals as auxiliary signals for correcting positioning results; solve the inertial navigation information through a pre-constructed factor graph model, use the positioning coordinates and control point coordinates calculated by the odometer as auxiliary correction information, and calculate the corrected inertial navigation positioning results;

[0011] When entering a confined space, obtain the positioning signal of the confined space, and use the positioning signal of the confined space in the joint positioning solution; wherein the confined space is a space where GNSS signals cannot be received; the positioning signal of the confined space includes: radio frequency signals, inertial navigation signals, odometers, control points, laser radars, and image information;

[0012] Preprocess the received positioning signals of confined space and unconfined space: obtain point cloud features by using point cloud preprocessing technology on the received LiDAR signals; calculate the orientation elements from the received image information to obtain image features; use these two features as the features of each frame, identify the change value between the two frames through feature matching technology, and calculate the coordinates of LiDAR and camera;

[0013] Obtain control point information and add the control point information as a smoothing correction item to the entire positioning process;

[0014] In a confined space, the preprocessed data values ​​of each positioning signal are obtained and input into the constructed factor graph model according to the chronological order. The factor graph is expanded through the defined system variable nodes and factor nodes. The recursion and update of the system state are realized by the factor nodes corresponding to different sensors, so as to realize the correction of inertial navigation positioning. Finally, the precise positioning coordinates are obtained to complete the fusion positioning.

[0015] The method for seamless indoor and outdoor positioning based on multi-source fusion, wherein the principle based on the factor graph is pre-set to construct a factor graph model through variable nodes, factor nodes and lines; and the steps of defining the corresponding relationship between each positioning signal and the variable nodes and factor nodes include:

[0016] Define the corresponding relationship between the odometer signal, wireless signal, radar signal, control point signal and variable node of the variable signal in the positioning signal, and define the corresponding factor node of the inertial navigation signal;

[0017] Based on the principle of factor graph, the factor graph model is constructed by connecting the inertial navigation signal and each variable node and the factor node.

[0018] The method for seamless indoor and outdoor positioning based on multi-source fusion, wherein the positioning signal of an outdoor non-restricted space is obtained, wherein the positioning signal of the outdoor non-restricted space includes: an inertial navigation signal, an odometer, and a control point signal; and the inertial navigation signal is used as the main signal of the positioning system, and other signals are used as auxiliary signals for correcting the positioning result; the inertial navigation information is solved by a pre-constructed factor graph model, and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information, and the steps of calculating the corrected inertial navigation positioning result include:

[0019] Obtain inertial navigation signals, odometers, and control point signals in outdoor non-restricted spaces as positioning signals in non-restricted spaces;

[0020] The inertial navigation signal is used as the main signal of the positioning system, and the odometer and control point signals are used as auxiliary signals for correcting the positioning results. The inertial navigation information is solved through a pre-constructed factor graph model, and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information to calculate the corrected inertial navigation positioning results.

[0021] The method for seamless indoor and outdoor positioning based on multi-source fusion, wherein when entering a confined space, the positioning signal of the confined space is obtained, and the positioning signal of the confined space is used in a joint positioning solution; wherein the confined space is a space where GNSS signals cannot be received; the positioning signal of the confined space includes: radio frequency signals, inertial navigation signals, odometers, control points, laser radars, and picture information. The steps include:

[0022] When entering a confined space, obtain the radio frequency signal, inertial navigation signal, odometer, control point, laser radar, and image information of the confined space as the positioning signal of the confined space;

[0023] And the positioning signal of the confined space is used in the joint positioning solution; wherein the confined space is a space where the GNSS signal cannot be received.

[0024] The method for seamless indoor and outdoor positioning based on multi-source fusion, wherein the received positioning signals of the confined space and the unconstrained space are preprocessed: the received laser radar signal is subjected to point cloud preprocessing technology to obtain point cloud features; the received image information is solved to obtain orientation elements to obtain image features; these two features are used as features of each frame, and the change value between the two frames is identified through feature matching technology to solve the laser radar and camera coordinates. The steps include:

[0025] Pre-process the received positioning signals of confined space and non-confined space:

[0026] The received LiDAR signal is processed through point cloud preprocessing technology to obtain point cloud features;

[0027] The received image information is solved to obtain the orientation elements to obtain the image features; these two features are used as the features of each frame, and the change value between the two frames is identified through feature matching technology;

[0028] A coordinate conversion method for multi-source heterogeneous data is established, and all heterogeneous positioning data are unified into the 2000 coordinate system to solve the laser radar and camera coordinates.

[0029] The method for seamless indoor and outdoor positioning based on multi-source fusion, wherein, in a confined space, the pre-processed data values ​​of each positioning signal are obtained, and are input into the constructed factor graph model according to the time sequence, and the factor graph is expanded through the defined system variable nodes and factor nodes, and the recursion and update of the system state corresponding to the factor nodes of different sensors are realized to realize the correction of inertial navigation positioning; finally, the accurate positioning coordinates are obtained, and the steps of completing the fusion positioning include:

[0030] In a confined space, obtaining preprocessed data values ​​of each positioning signal;

[0031] According to the chronological order, the data are input into the constructed factor graph model, and the factor graph is expanded through the defined system variable nodes and factor nodes. The factor nodes corresponding to different sensors realize the recursion and update of the system state, and realize the correction of inertial navigation positioning.

[0032] Finally, the corrected positioning data is fused to obtain accurate positioning coordinates and complete the fusion positioning.

[0033] A seamless indoor and outdoor positioning device based on multi-source fusion, comprising:

[0034] The factor graph model building module is used to build the factor graph model based on the principle of factor graph by presetting variable nodes, factor nodes and connections; and define the corresponding relationship between each positioning signal and the variable nodes and factor nodes;

[0035] The outdoor non-restricted space signal acquisition and processing module is used to acquire the positioning signal of the outdoor non-restricted space, wherein the positioning signal of the outdoor non-restricted space includes: inertial navigation signal, odometer, and control point signal; the inertial navigation signal is used as the main signal of the positioning system, and other signals are used as auxiliary signals for correcting the positioning result; the inertial navigation information is solved by the pre-constructed factor graph model, and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information to calculate the corrected inertial navigation positioning result;

[0036] The confined space signal acquisition and processing module is used to acquire the positioning signal of the confined space when entering the confined space, and use the positioning signal of the confined space in the joint positioning solution; wherein the confined space is a space where GNSS signals cannot be received; the positioning signal of the confined space includes: radio frequency signal, inertial navigation signal, odometer, control point, laser radar, and picture information;

[0037] The preprocessing module is used to preprocess the received positioning signals of the confined space and the unconstrained space: the received LiDAR signals are processed through the point cloud preprocessing technology to obtain the point cloud features; the received image information is solved to obtain the orientation elements to obtain the image features; these two features are used as the features of each frame, and the change value between the two frames is identified through the feature matching technology to solve the coordinates of the LiDAR and the camera;

[0038] A control point information acquisition module is used to acquire control point information and add the control point information as a smoothing correction item to the entire positioning process;

[0039] The fusion processing and positioning module is used to obtain the pre-processed data values ​​of each positioning signal in a confined space, and input them into the constructed factor graph model according to the chronological order. The factor graph is expanded through the defined system variable nodes and factor nodes, and the recursion and update of the system state corresponding to the factor nodes of different sensors are realized to realize the correction of inertial navigation positioning; finally, the precise positioning coordinates are obtained to complete the fusion positioning.

[0040] The indoor and outdoor seamless positioning device based on multi-source fusion, wherein it also includes:

[0041] A definition module is used to define the corresponding relationship between the odometer signal, wireless signal, radar signal, control point signal and variable node of the variable signal in the positioning signal, and define the corresponding factor node of the inertial navigation signal;

[0042] The coordinate conversion module is used to establish the coordinate conversion method of multi-source heterogeneous data and unify all heterogeneous positioning data into the 2000 coordinate system.

[0043] An intelligent terminal includes a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, including the method for executing any one of the methods described above.

[0044] A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute any one of the methods described above.

[0045] Beneficial effects of the present invention: The present invention provides a method, intelligent terminal and storage medium for seamless indoor and outdoor positioning based on multi-source fusion, and the method includes: 1) a fusion framework of multi-source positioning information (including radio frequency signals, inertial data, lidar data, camera photos, control points and global navigation system GNSS data) based on factor graphs. The method is based on inertial data, and integrates other positioning source data collected from time to time. It has the versatility of seamless indoor and outdoor positioning and can receive asynchronous signal processing at any time. 2) The framework includes a coordinate conversion method, which can unify all asynchronous data into the same coordinate system for processing. 3) The method uses a fixed-frequency factor graph model, and variable nodes are generated at a fixed rate, which avoids the problems of reduced time span and frequent optimization of traditional factor graphs, improves estimation accuracy, and reduces computational complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 It is a flow chart of a method for seamless indoor and outdoor positioning based on multi-source fusion provided in an embodiment of the present invention.

[0048] Figure 2 It is a schematic diagram of a multi-source sensor fusion framework based on a factor graph of the indoor and outdoor seamless positioning method based on multi-source fusion provided by an embodiment of the present invention.

[0049] Figure 3 It is a schematic diagram of a fixed rate factor graph model of the indoor and outdoor seamless positioning method based on multi-source fusion provided in an embodiment of the present invention.

[0050] Figure 4 It is a schematic diagram of a fixed frequency factor graph optimization framework of an indoor and outdoor seamless positioning method based on multi-source fusion provided in an embodiment of the present invention.

[0051] Figure 5 It is a principle block diagram of an indoor and outdoor seamless positioning device based on multi-source fusion provided by an embodiment of the present invention.

[0052] Figure 6 It is a block diagram of the internal structure principle of the intelligent terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0054] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0055] In the era of the Internet of Things, indoor and outdoor positioning plays various important roles to determine the location of people, mobile devices, and various robots. Due to the popularity of social networks and the widespread use of mobile devices, the demand for location-based services (LBS) in indoor and outdoor environments has increased significantly. In addition, with the popularization of the Internet and the rapid increase in communication network infrastructure, human daily life is also closely related to positioning technology. Therefore, the value of seamless indoor and outdoor positioning technology is highlighted.

[0056] Indoor and outdoor positioning can be applied to universities, airports, military systems, security systems, subway service route maps, agriculture, forests and other fields. Applications such as emergency management, tracking mobile users, context perception, advertising, environmental monitoring, military monitoring, and health care are inseparable from positioning. Due to its important significance and application value, indoor and outdoor positioning has been valued by industry and academia.

[0057] Whether indoors or outdoors, single-source positioning systems have obvious disadvantages. For example, the performance of inertial positioning systems degrades over time due to the accumulation of errors in inertial sensors. The Global Navigation Satellite System (GNSS) is one of the main means for autonomous systems with navigation needs to provide global reference positioning. However, in highly urbanized areas, the positioning accuracy is significantly reduced due to non-optical line of sight (NLOS) reception and multipath effects caused by signal reflections caused by static buildings and dynamic objects (such as double-decker buses). Video positioning has problems such as difficult data registration and high scene requirements. All types of sensors have their own problems.

[0058] Therefore, the present invention provides a best method to solve the problem of seamless indoor and outdoor positioning, using a multi-source sensor fusion method. The present invention proposes a positioning method of multiple sensor fusions and proposes a universal multi-source sensor fusion positioning framework that does not rely entirely on external GNSS signals, which makes indoor and outdoor hybrid integrated positioning possible.

[0059] In order to solve the problem of integrated indoor and outdoor positioning of pedestrians, robots (drain pipe inspection robots, crawling robots), and drones, the present invention provides a method for seamless indoor and outdoor positioning based on multi-source fusion, the method comprising: 1) constructing a fusion framework of multi-source positioning information (the input multi-source sensor data includes: radio frequency signals, inertial data, lidar data, camera photos, control points, and global navigation system GNSS data) based on factor graphs. The method is based on inertial data and integrates other positioning source data collected from time to time. It has the versatility of seamless indoor and outdoor positioning and can receive asynchronous signal processing at any time. 2) The framework includes a coordinate conversion method that can unify all asynchronous data into the same coordinate system for processing. 3) The method uses a fixed-frequency factor graph model, and variable nodes are generated at a fixed rate, which avoids the problems of reduced time span and frequent optimization of traditional factor graphs, improves estimation accuracy, and reduces computational complexity. The present invention achieves high-precision positioning of pedestrians, robots, and drones.

[0060] Exemplary Methods

[0061] like Figure 1 As shown in , an embodiment of the present invention provides an indoor and outdoor seamless positioning method based on multi-source fusion, and the indoor and outdoor seamless positioning method based on multi-source fusion can be applied to smart terminals. In an embodiment of the present invention, the method includes the following steps:

[0062] Step S100: Based on the principle of factor graph, variable nodes, factor nodes and connections are pre-set to construct a factor graph model; and the corresponding relationship between each positioning signal and the variable node and factor node is defined;

[0063] In the embodiment of the present invention, it is necessary to pre-construct a factor graph model. Specifically, the factor graph model is constructed based on the principle of factor graph, and the factor graph model is constructed through variable nodes, factor nodes and connections; Figure 2 As shown, the corresponding relationship between each positioning signal and the variable node and the factor node is defined: for example, the variable odometer signal, wireless signal, radar signal, control point signal and the variable node are defined as the corresponding relationship, and the inertial navigation signal is defined as the corresponding factor node. By connecting the inertial navigation signal and each variable node and factor node, the factor graph model is constructed.

[0064] Step S200, obtaining a positioning signal of an outdoor non-restricted space, wherein the positioning signal of the outdoor non-restricted space includes: an inertial navigation signal, an odometer, and a control point signal; using the inertial navigation signal as the main signal of the positioning system, and other signals as auxiliary signals for correcting the positioning result; solving the inertial navigation information through a pre-constructed factor graph model, using the positioning coordinates and control point coordinates calculated by the odometer as auxiliary correction information, and calculating the corrected inertial navigation positioning result.

[0065] like Figure 2 As shown, in the present invention, a positioning signal of an outdoor non-restricted space is obtained, wherein the positioning signal of the outdoor non-restricted space includes: an inertial navigation signal, an odometer, and a control point signal; the inertial navigation signal is used as the main signal of the positioning system, and other signals are used as auxiliary signals for correcting the positioning result; the inertial navigation information is solved by a preset factor graph model (probability graph), and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information to calculate the corrected inertial navigation positioning result.

[0066] Step S300: When entering a confined space, obtain the positioning signal of the confined space, and use the positioning signal of the confined space in a joint positioning solution; wherein the confined space is a space where GNSS signals cannot be received; the positioning signal of the confined space includes: radio frequency signals, inertial navigation signals, odometers, control points, lidars, and image information.

[0067] In this step, when entering a confined space, the positioning signal of the confined space is obtained, and the positioning signal of the confined space is used in the joint positioning solution. The confined space refers to a space where GNSS signals cannot be received, such as a tunnel or under a bridge; the positioning signal of the confined space includes: radio frequency signals (such as UWB signals), inertial navigation, odometer, control points, laser radar, pictures and other information.

[0068] In the embodiment of the present invention, the so-called involving the positioning signal of the confined space in the joint positioning solution is specifically positioning, and the advantage is that stable and accurate positioning can be achieved in the confined space.

[0069] Step S400, preprocessing the received positioning signals of the confined space and the unconfined space: obtaining point cloud features from the received lidar signals through point cloud preprocessing technology; solving the orientation elements from the received image information to obtain image features; using these two features as features of each frame, identifying the change value between the two frames through feature matching technology, and solving the lidar and camera coordinates.

[0070] Preprocess the received positioning signals of confined space and unconfined space. Obtain point cloud features by using point cloud preprocessing technology for the received LiDAR signal; obtain image features by calculating the orientation elements from the received image information; use these two features as the features of each frame, and identify the change value between the two frames through feature matching technology, thereby calculating the coordinates of the LiDAR and camera. UWB signals can directly input ranging information values.

[0071] Step S500, obtaining control point information, and adding the control point information as a smoothing correction item to the entire positioning process;

[0072] Get the control point information and add it as a smoothing correction item to the entire positioning process. Since the control point coordinates are accurate, using control points to smooth the positioning results improves the robustness and reliability of the positioning results.

[0073] Step S600: In a confined space, obtain the pre-processed data values ​​of each positioning signal, input them into the constructed factor graph model according to the time sequence, expand the factor graph through the defined system variable nodes and factor nodes, implement the recursion and update of the system state corresponding to the factor nodes of different sensors, and realize the correction of inertial navigation positioning; finally, obtain the accurate positioning coordinates and complete the fusion positioning;

[0074] In a restricted space, after obtaining the preprocessed values ​​of each data, they are input into the constructed factor graph model according to the time sequence. The factor graph is expanded through the defined system variable nodes and factor nodes, and the recursion and update of the system state corresponding to the factor nodes of different sensors are realized to realize the correction of inertial navigation positioning. Finally, the precise positioning coordinates are obtained to complete the fusion positioning.

[0075] In the embodiment of the present invention, the specific steps are as follows:

[0076] Regarding the principle based on the factor graph in step S100, the steps of constructing a factor graph model through variable nodes, factor nodes and connections are pre-set; and defining the corresponding relationship between each positioning signal and the variable nodes and factor nodes are specifically as follows:

[0077] A probabilistic graphical model is a graphical model that expresses the joint probability distribution of random variables. It consists of two parts: a graph and its related parameters. Currently, commonly used probabilistic graphical models include Bayesian networks, Markov networks, factor graphs, etc. Factor graphs are a general model with strong multi-source sensor fusion capabilities.

[0078] Factor graphs can be used as probabilistic graph models to express the joint probability distribution of random variables. They are composed of variable nodes, factor nodes, and links. Suppose all variables in a factor graph are x1, x2, ..., x n, all local functions are f1(X1), f2(X2), ..., f k (X k ).in, 1≤k≤K, represents the kth local function f k (·) is the set of independent variable points, K represents the total number of local functions, then the joint probability density distribution specified by the factor graph is:

[0079]

[0080] Among them, f k (X k )(1≤k≤K) is a non-negative finite function, Z is a normalization factor, if the joint probability distribution function It can be expressed as the following factor decomposition form:

[0081]

[0082] The corresponding factor graph is Figure 3 As shown, the dotted hollow circular nodes on the middle line represent variables, and the solid circular nodes with words on the upper and lower sides of the middle line represent local functions.

[0083] In the present invention, the corresponding relationship between each positioning signal and the variable node and factor node is defined as follows: Figure 2 As shown, for example: define the odometer signal, wireless signal, radar signal, control point signal and variable node of the variable signal in the positioning signal as a corresponding relationship, define the factor node corresponding to the inertial navigation signal, and construct the factor graph model by connecting the inertial navigation signal and each variable node and factor node.

[0084] like Figure 4 As shown, Figure 4 is a schematic diagram of a fixed frequency factor graph optimization framework of an indoor and outdoor seamless positioning method based on multi-source fusion provided by an embodiment of the present invention. In an embodiment of the present invention, Figure 4 As shown, by acquiring multi-source sensor data. The multi-source sensor data is divided into two categories, one is absolute measurement value and the other is relative measurement value. The present invention puts the absolute measurement value into a set univariate factor and the relative measurement value into a set binary factor, and performs a pre-integration step at the same time to obtain a fixed frequency factor graph.

[0085] In an embodiment of the present invention, regarding step S200 of acquiring a positioning signal of an outdoor non-restricted space, wherein the positioning signal of the outdoor non-restricted space includes: an inertial navigation signal, an odometer, and a control point signal; and the inertial navigation signal is used as the main signal of the positioning system, and other signals are used as auxiliary signals for correcting the positioning result; the inertial navigation information is solved by a pre-constructed factor graph model, and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information to calculate the corrected inertial navigation positioning result; and step S300 of acquiring a positioning signal of the confined space when entering the confined space, and using the positioning signal of the confined space to participate in the joint positioning solution; wherein the confined space is a space where the GNSS signal cannot be received; the positioning signal of the confined space includes: a radio frequency signal, an inertial navigation signal, an odometer, a control point, a laser radar, and image information; and step S400 of acquiring the control point information, and adding the control point information as a smoothing correction item to the entire positioning process. The specific explanation is as follows:

[0086] With the increase in the number and types of sensors, different sensors have different update frequencies and are asynchronous in time. Traditional information fusion methods are difficult to meet the increasingly complex and changing needs of the system. The present invention combines the principle of factor graphs to construct a factor graph model, obtains measurement information from different sensors, expands the factor graph through defined system variable nodes and factor nodes, and implements the recursion and update of the system state corresponding to the factor nodes of different sensors, thereby completing the information fusion of multiple sensors.

[0087] A factor graph is a bipartite graph G = (F, X, E), which includes two types of nodes: factor nodes f i ∈F and variable node x i ∈X, where e ij ∈E represents the edge between the factor node and the variable node. The factor graph G defines the factorization of the function f(X) as:

[0088] f(X)=Π i f i (X i ),

[0089] Among them, X i are all variables X j The set of factors f i The concept of factor graph is introduced into navigation information fusion. The variable node represents the integrated navigation state variable, and the factor node represents the measurement update process of the integrated navigation. After receiving the measurement information from the sensor, the factor node is used to perform recursive estimation of the state.

[0090] like Figure 2 As shown, Figure 2It is a factor graph framework for information fusion, in which the sensors involved include radio frequency signals (wireless signals), inertial data (inertial navigation), lidar data (radar), camera photos (vision), control points and global navigation system GNSS (Global Navigation Satellite System) data. Figure 2 As shown, the odometer signal, wireless signal, radar signal, control point signal and variable nodes of the variable signals in the positioning signal are defined as corresponding relationships, and the factor nodes corresponding to the inertial navigation signal are defined. By connecting the inertial navigation signal and each variable node and factor node, the factor graph model is constructed.

[0091] Among them, the state variables of the system are set as variable nodes:

[0092]

[0093] The variable nodes include the platform error angle φ E ,φ η ,φ υ , speed error δυ E ,δυ η ,δυ υ , position error δL, δλ, δh, gyro constant drift error ε bx , ε by , ε bz , gyro first-order Markov drift error ε rx , ε ry , ε rz , and the accelerometer bias

[0094] In the navigation framework, the measurement model h(·) is defined to predict the sensor's measurement value based on the given state estimate. The factor node obtains the difference between the predicted measurement information and the actual measurement information, and constructs the corresponding indicator function to obtain the estimate of the state variable. Based on the Gaussian white noise model, a measurement factor node can be expressed as follows:

[0095] f i (X i )=d[h i (X i )-z i ]

[0096] In the formula, h i (X i ) represents the measurement model, which is a function related to the state variables; z i represents the actual observation information; d(·) represents the cost function.

[0097] After receiving the observation information from the IMU (inertial measurement unit), define the factor node f IMU .f IMUConnect different times t k and t k+1 Two variable nodes, namely the navigation state x k and x k+1 . Update the state and expand the variable nodes according to the state equation of the system:

[0098] x k+1 =F(a,w)x k

[0099] Where a and w represent the outputs of the accelerometer and gyroscope, respectively, and F(·) represents the state transfer function matrix of the system.

[0100] Then the factor node IMU can be expressed as:

[0101] f IMU (x k+1 , z k ) = d[x k+1 -h(x k , z k )]

[0102] The unknown vectors involved in the above formula all represent variable nodes in the factor graph, among which the IMU factor is the factor node connecting the variables.

[0103] In t l After receiving the GPS observation information at the moment, define the factor node f GPS .f GPS Connect to variable node x l , the GPS measurement equation is as follows:

[0104] z GPS =h GPS (x l )+υ GPS

[0105] In the formula, z GPS is the GPS observation information, h GPS is the observation function, u GPS is the observation noise.

[0106] Therefore, the factor node f GPS It can be expressed as:

[0107] f GPS (x l )=d[z GPS -h GPS (x l )]

[0108] The cost function of the selected factor node is as follows:

[0109]

[0110] Similarly, after receiving the lidar data, camera photos, control points, and GNSS data at other times, define the factor node f baro 、f flow 、f mag 、f star Expand the factor graph and update the state of the variable node according to the sensor's observation equation and the corresponding cost function. The multi-sensor information fusion method based on the factor graph can easily process the data from asynchronous sensors. After receiving the output data of the sensor, expand the factor graph node, quickly and effectively update the system state according to the system's state equation and measurement equation, and realize the comprehensive data processing of multiple sensors.

[0111] In the embodiment of the present invention, the steps of preprocessing the received positioning signals of the confined space and the non-confined space in step S400 are as follows: the received laser radar signal is subjected to point cloud preprocessing technology to obtain point cloud features; the received image information is solved to obtain orientation elements to obtain image features; these two features are used as features of each frame, and the change value between the two frames is identified through feature matching technology to solve the laser radar and camera coordinates.

[0112] In this step, the received positioning signals of the confined space and the unconstrained space are preprocessed, and a unified multi-source sensor data coordinate system is adopted to establish a coordinate conversion method for multi-source heterogeneous data, and all heterogeneous positioning data are unified into the 2000 coordinate system.

[0113] The movement of an object is described relative to a certain reference system. The task of the navigation system is to determine the motion parameters of the carrier, that is, to determine the position of the carrier in a certain coordinate system, as well as the change of the position. For different carriers, different coordinate systems should be selected due to their different ranges of movement. Due to the different shapes, sizes and geoid refinement methods of the earth ellipsoid selected by different countries and regions, independent geodetic coordinate systems are formed. The same ground point corresponds to different coordinate values ​​in different coordinate systems. Therefore, determining a unified coordinate system is particularly important in seamless positioning.

[0114] The use of different coordinate systems by various positioning sources will cause large errors in seamless positioning. A method for seamless indoor and outdoor positioning based on multi-source fusion in an embodiment of the present invention unifies all positioning data into the 2000 coordinate system. The measurement of various physical quantities in navigation is also carried out in a certain coordinate system and between coordinate systems, such as the geocentric inertial system, the geocentric fixed system, the local geographic system, etc. The unification of the navigation coordinate system in the embodiment of the present invention requires the use of a coordinate transformation method. According to the theory of spatial geometry, the coordinate system transformation in navigation can basically be divided into three categories: the conversion between geodetic coordinates of different coordinate systems, the conversion between rectangular coordinates of different coordinate systems, and the conversion between geodetic coordinates and rectangular coordinates in the same coordinate system.

[0115] a. In the embodiment of the present invention, the mathematical expression for conversion between geodetic coordinates of different coordinate systems is:

[0116]

[0117] λ 2000 =λ original +Δλ

[0118] h 2000 =h original +Δh

[0119] in, represents the longitude value, λ represents the latitude value, and h represents the altitude value.

[0120] b. In the embodiment of the present invention, the conversion between different rectangular coordinate systems is expressed as follows:

[0121]

[0122] Among them, ΔX, ΔY, ΔZ are the translation parameters between the origins of the two coordinate systems, K is the scale factor between the two coordinate systems, and ε X , ε Y , ε Z is the Euler angle.

[0123] The seven-parameter conversion method has a relatively accurate mathematical model, and the conversion formula is as follows:

[0124]

[0125] In the embodiment of the present invention, in step S600, the pre-processed data values ​​of each positioning signal are obtained in a confined space, and are input into the constructed factor graph model according to the time sequence. The factor graph is expanded through the defined system variable nodes and factor nodes, and the recursion and update of the system state corresponding to the factor nodes of different sensors are realized to realize the correction of the inertial navigation positioning; finally, the accurate positioning coordinates are obtained, and the fusion positioning is completed as follows:

[0126] The indoor and outdoor seamless positioning method based on factor graph, the graph model in the proposed factor graph is as follows Figure 3 As shown. The box timu represents a variable node constructed at a fixed rate. In the factor graph, the construction process of the variable node is equivalent to estimating the state vector using nonlinear optimization, that is, establishing the state vector, deriving the measurement model corresponding to the state vector, and forming the residual of the fixed update rate. Based on these residuals, the state vector is estimated. If the fixed rate is rate f Hz, then at 1 / rate f The state estimation is performed by nonlinear optimization at fixed time intervals of 10 seconds. The circles tabs represent unary factors from absolute measurements. The circles trel are binary factors from relative measurements. The dotted circles represent the asynchrony of different measurements.

[0127] from Figure 3 It can be seen that asynchronous absolute and relative measurements are received at different rates. Figure 3 The variable nodes in are generated according to the arrival of new measurement sources. When the window size is 5, the time span decreases. As the time span decreases, the information obtained during the optimization process is limited, thereby reducing the estimation accuracy. At the same time, frequent optimization will lead to an increase in the amount of calculation. In the embodiment of the present invention, in the proposed method, the variable nodes are constructed at a fixed rate, such as Figure 3 The five boxes in timu are shown in the figure. The addition of variable nodes is not affected by asynchronous measurements, which ensures the time span in the window. In addition, since asynchronous measurements can be associated with corresponding variable nodes through the derived mathematical model to form error terms, the method can easily adapt to the addition and removal of information sources. Information with a long time span can be fully utilized for iterative optimization to improve the estimation accuracy. In addition, the optimization can be performed at a relatively low rate, thereby reducing the computational load. Therefore, better plug-and-play capabilities can be achieved for asynchronous measurements.

[0128] As can be seen from the above, the present invention proposes to use factor graphs to fuse and locate the acquired multi-source sensor data (including: radio frequency signals, inertial data, lidar data, camera photos, control points and global navigation system GNSS data) to solve the problem of seamless indoor and outdoor positioning.

[0129] Furthermore, the present invention establishes a coordinate conversion method for multi-source heterogeneous data and unifies all heterogeneous positioning data into a 2000 coordinate system, which can effectively reduce the seamless positioning error of multi-source heterogeneous data.

[0130] The present invention also proposes a fixed rate factor graph method in which variable nodes are generated at a fixed rate and are not affected by the measured values. Asynchronous measurements can be associated with corresponding variable nodes in the graph for optimization. The method avoids the time span reduction and frequent optimization of traditional factor graphs, improves estimation accuracy, and reduces the amount of calculation. Therefore, better plug-and-play capabilities can be achieved for asynchronous measurements.

[0131] Exemplary Devices

[0132] like Figure 5 As shown in, an embodiment of the present invention provides an indoor and outdoor seamless positioning device based on multi-source fusion, the device comprising:

[0133] The factor graph model construction module 510 is used to construct the factor graph model by presetting the variable nodes, factor nodes and connections based on the principle of the factor graph; and define the corresponding relationship between each positioning signal and the variable nodes and factor nodes;

[0134] The outdoor non-restricted space signal acquisition and processing module 520 is used to acquire the positioning signal of the outdoor non-restricted space, wherein the positioning signal of the outdoor non-restricted space includes: inertial navigation signal, odometer, and control point signal; and the inertial navigation signal is used as the main signal of the positioning system, and other signals are used as auxiliary signals for correcting the positioning result; the inertial navigation information is solved by the pre-constructed factor graph model, and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information to calculate the corrected inertial navigation positioning result;

[0135] The confined space signal acquisition and processing module 530 is used to acquire the positioning signal of the confined space when entering the confined space, and use the positioning signal of the confined space in the joint positioning solution; wherein the confined space is a space where the GNSS signal cannot be received; the positioning signal of the confined space includes: radio frequency signal, inertial navigation signal, odometer, control point, laser radar, and picture information;

[0136] The preprocessing module 540 is used to preprocess the received positioning signals of the confined space and the unconfined space: the received laser radar signal is processed by the point cloud preprocessing technology to obtain the point cloud features; the received image information is solved to obtain the orientation elements to obtain the image features; these two features are used as the features of each frame, and the change value between the two frames is identified by the feature matching technology to solve the laser radar and camera coordinates;

[0137] A control point information acquisition module 550 is used to acquire control point information and add the control point information as a smoothing correction item to the entire positioning process;

[0138] The fusion processing and positioning module 560 is used to obtain the pre-processed data values ​​of each positioning signal in a confined space, input them into the constructed factor graph model according to the chronological order, expand the factor graph through the defined system variable nodes and factor nodes, and realize the recursion and update of the system state corresponding to the factor nodes of different sensors, so as to realize the correction of inertial navigation positioning; finally, the precise positioning coordinates are obtained to complete the fusion positioning, as described above.

[0139] Furthermore, the indoor and outdoor seamless positioning device based on multi-source fusion is characterized in that it also includes:

[0140] A definition module is used to define the corresponding relationship between the odometer signal, wireless signal, radar signal, control point signal and variable node of the variable signal in the positioning signal, and define the corresponding factor node of the inertial navigation signal;

[0141] The coordinate conversion module is used to establish the coordinate conversion method of multi-source heterogeneous data and unify all heterogeneous positioning data into the 2000 coordinate system.

[0142] Based on the above embodiments, the present invention further provides an intelligent terminal, whose principle block diagram can be shown as follows: Figure 6 As shown. The intelligent terminal includes a processor, a memory, a network interface, a display screen, and a sensor connected through a system bus. Among them, the processor of the intelligent terminal is used to provide computing and control capabilities. The memory of the intelligent terminal includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the intelligent terminal is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a seamless indoor and outdoor positioning method based on multi-source fusion is implemented. The display screen of the intelligent terminal can be a liquid crystal display screen or an electronic ink display screen.

[0143] Those skilled in the art will understand that Figure 6 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the scheme of the present invention, and does not constitute a limitation on the smart terminal to which the scheme of the present invention is applied. The specific smart terminal may include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0144] In one embodiment, a smart terminal is provided, comprising a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, and the one or more programs include instructions for performing the following operations:

[0145] Based on the principle of factor graph, the factor graph model is constructed by presetting variable nodes, factor nodes and connections; and the corresponding relationship between each positioning signal and variable nodes and factor nodes is defined;

[0146] Acquire positioning signals of outdoor non-restricted spaces, wherein the positioning signals of outdoor non-restricted spaces include: inertial navigation signals, odometers, and control point signals; use the inertial navigation signals as the main signals of the positioning system, and other signals as auxiliary signals for correcting positioning results; solve the inertial navigation information through a pre-constructed factor graph model, use the positioning coordinates and control point coordinates calculated by the odometer as auxiliary correction information, and calculate the corrected inertial navigation positioning results;

[0147] When entering a confined space, obtain the positioning signal of the confined space, and use the positioning signal of the confined space in the joint positioning solution; wherein the confined space is a space where GNSS signals cannot be received; the positioning signal of the confined space includes: radio frequency signals, inertial navigation signals, odometers, control points, laser radars, and image information;

[0148] Preprocess the received positioning signals of confined space and unconfined space: obtain point cloud features by using point cloud preprocessing technology on the received LiDAR signals; calculate the orientation elements from the received image information to obtain image features; use these two features as the features of each frame, identify the change value between the two frames through feature matching technology, and calculate the coordinates of LiDAR and camera;

[0149] Obtain control point information and add the control point information as a smoothing correction item to the entire positioning process;

[0150] In a confined space, the preprocessed data values ​​of each positioning signal are obtained and input into the constructed factor graph model according to the chronological order. The factor graph is expanded through the defined system variable nodes and factor nodes. The recursion and update of the system state are realized by the factor nodes corresponding to different sensors, so as to realize the correction of inertial navigation positioning. Finally, the precise positioning coordinates are obtained and the fusion positioning is completed, as described above.

[0151] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0152] In summary, the present invention discloses a method for seamless indoor and outdoor positioning based on multi-source fusion, an intelligent terminal and a storage medium, the method comprising: based on the principle of factor graph, pre-setting variable nodes, factor nodes and lines to construct a factor graph model; and defining the corresponding relationship between each positioning signal and the variable nodes and factor nodes; obtaining the positioning signal of an outdoor non-restricted space, wherein the positioning signal of the outdoor non-restricted space comprises: an inertial navigation signal, an odometer, and a control point signal; and using the inertial navigation signal as the main signal of the positioning system, and other signals as auxiliary signals for correcting the positioning result; solving the inertial navigation information through a pre-constructed factor graph model, and using the positioning coordinates and control point coordinates calculated by the odometer as auxiliary correction information to calculate the corrected inertial navigation positioning result; when entering a confined space, obtaining the positioning signal of the confined space, and using the positioning signal of the confined space to participate in the joint positioning solution; wherein the confined space is a space where a GNSS signal cannot be received; the confined space The spatial positioning signals include: radio frequency signals, inertial navigation signals, odometers, control points, lidar, and image information; the received positioning signals of confined space and unconstrained space are preprocessed: the received lidar signals are processed through point cloud preprocessing technology to obtain point cloud features; the received image information is solved to obtain orientation elements to obtain image features; these two features are used as the features of each frame, and the change values ​​between the two frames are identified through feature matching technology to solve the lidar and camera coordinates; the control point information is obtained and added to the entire positioning process as a smoothing correction item; in a confined space, the preprocessed data values ​​of each positioning signal are obtained, and input into the constructed factor graph model according to the chronological order, and the factor graph is expanded through the defined system variable nodes and factor nodes, and the recursion and update of the system state corresponding to the factor nodes of different sensors are realized to realize the correction of inertial navigation positioning; finally, accurate positioning coordinates are obtained to complete the fusion positioning. The embodiment of the present invention proposes to use factor graphs to perform fusion positioning on the acquired multi-source sensor data (including: radio frequency signals, inertial data, lidar data, camera photos, control points and global navigation system GNSS data) to solve the problem of seamless positioning indoors and outdoors; it can effectively reduce the seamless positioning error of multi-source heterogeneous data. In addition, the present invention proposes a fixed rate factor graph method in which variable nodes are generated at a fixed rate and are not affected by the measurement value. Asynchronous measurements can be associated with corresponding variable nodes in the graph for optimization. This method avoids the time span reduction and frequent optimization of traditional factor graphs, improves estimation accuracy, and reduces the amount of calculation.

[0153] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A seamless indoor and outdoor positioning method based on multi-source fusion, characterized in that: The method comprises: Based on the principle of factor graph, the factor graph model is constructed by presetting variable nodes, factor nodes and connections; and the corresponding relationship between each positioning signal and variable nodes and factor nodes is defined; Acquire positioning signals of outdoor non-restricted spaces, wherein the positioning signals of outdoor non-restricted spaces include: inertial navigation signals, odometers, and control point signals; use the inertial navigation signals as the main signals of the positioning system, and other signals as auxiliary signals for correcting positioning results; solve the inertial navigation information through a pre-constructed factor graph model, use the positioning coordinates and control point coordinates calculated by the odometer as auxiliary correction information, and calculate the corrected inertial navigation positioning results; When entering a confined space, obtain the positioning signal of the confined space, and use the positioning signal of the confined space in the joint positioning solution; wherein the confined space is a space where GNSS signals cannot be received; the positioning signal of the confined space includes: radio frequency signals, inertial navigation signals, odometers, control points, laser radars, and image information; Preprocess the received positioning signals of confined space and unconfined space: obtain point cloud features by using point cloud preprocessing technology on the received LiDAR signals; calculate the orientation elements from the received image information to obtain image features; use these two features as the features of each frame, identify the change value between the two frames through feature matching technology, and calculate the coordinates of LiDAR and camera; Obtain control point information and add the control point information as a smoothing correction item to the entire positioning process; In a confined space, the pre-processed data values ​​of each positioning signal are obtained and input into the constructed factor graph model according to the time sequence. The factor graph is expanded through the defined system variable nodes and factor nodes, and the recursion and update of the system state corresponding to the factor nodes of different sensors are realized to realize the correction of inertial navigation positioning; finally, the precise positioning coordinates are obtained to complete the fusion positioning; The step of constructing a factor graph model by presetting variable nodes, factor nodes and connections based on the principle of factor graph and defining the corresponding relationship between each positioning signal and the variable nodes and factor nodes comprises: Define the corresponding relationship between the odometer signal, wireless signal, radar signal, control point signal and variable node of the variable signal in the positioning signal, and define the corresponding factor node of the inertial navigation signal; Based on the principle of factor graph, the factor graph model is constructed by connecting the inertial navigation signal and each variable node and the factor node.

2. The method for seamless indoor and outdoor positioning based on multi-source fusion according to claim 1, characterized in that: The step of obtaining a positioning signal of an outdoor non-restricted space, wherein the positioning signal of the outdoor non-restricted space includes: an inertial navigation signal, an odometer, and a control point signal; using the inertial navigation signal as the main signal of the positioning system, and other signals as auxiliary signals for correcting the positioning result; solving the inertial navigation information through a pre-constructed factor graph model, using the positioning coordinates and the control point coordinates calculated by the odometer as auxiliary correction information, and calculating the corrected inertial navigation positioning result includes: Obtain inertial navigation signals, odometers, and control point signals in outdoor non-restricted spaces as positioning signals in non-restricted spaces; The inertial navigation signal is used as the main signal of the positioning system, and the odometer and control point signals are used as auxiliary signals for correcting the positioning results. The inertial navigation information is solved through a pre-constructed factor graph model, and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information to calculate the corrected inertial navigation positioning results.

3. The method for seamless indoor and outdoor positioning based on multi-source fusion according to claim 1, characterized in that: When entering a confined space, obtaining a positioning signal of the confined space, and using the positioning signal of the confined space in a joint positioning solution; wherein the confined space is a space where a GNSS signal cannot be received; the positioning signal of the confined space includes: a radio frequency signal, an inertial navigation signal, an odometer, a control point, a laser radar, and image information. The steps include: When entering a confined space, obtain the radio frequency signal, inertial navigation signal, odometer, control point, laser radar, and image information of the confined space as the positioning signal of the confined space; And the positioning signal of the confined space is used in the joint positioning solution; wherein the confined space is a space where the GNSS signal cannot be received.

4. The method for seamless indoor and outdoor positioning based on multi-source fusion according to claim 1, characterized in that: The received positioning signals of the confined space and the unconfined space are preprocessed: the received laser radar signals are processed by point cloud preprocessing technology to obtain point cloud features; the received image information is solved to obtain orientation elements to obtain image features; The steps of using these two features as the features of each frame, identifying the change value between the two frames through feature matching technology, and solving the laser radar and camera coordinates include: Pre-process the received positioning signals of confined space and non-confined space: The received LiDAR signal is processed through point cloud preprocessing technology to obtain point cloud features; The received image information is solved to obtain the orientation elements to obtain the image features; these two features are used as the features of each frame, and the change value between the two frames is identified through feature matching technology; A coordinate conversion method for multi-source heterogeneous data is established, and all heterogeneous positioning data are unified into the 2000 coordinate system to solve the laser radar and camera coordinates.

5. The method for seamless indoor and outdoor positioning based on multi-source fusion according to claim 1, characterized in that: The method of obtaining the preprocessed data values ​​of each positioning signal in a confined space and inputting them into the constructed factor graph model according to the time sequence, expanding the factor graph through the defined system variable nodes and factor nodes, implementing the recursion and update of the system state corresponding to the factor nodes of different sensors, and realizing the correction of the inertial navigation positioning; finally, obtaining the accurate positioning coordinates, and completing the steps of fusion positioning include: In a confined space, obtaining preprocessed data values ​​of each positioning signal; According to the chronological order, the data are input into the constructed factor graph model, and the factor graph is expanded through the defined system variable nodes and factor nodes. The recursion and update of the system state are realized by the factor nodes corresponding to different sensors, so as to realize the correction of inertial navigation positioning. Finally, the corrected positioning data is fused to obtain accurate positioning coordinates and complete the fusion positioning.

6. A seamless indoor and outdoor positioning device based on multi-source fusion, characterized in that: include: The factor graph model building module is used to build the factor graph model based on the principle of factor graph by presetting variable nodes, factor nodes and connections; And define the corresponding relationship between each positioning signal and variable node and factor node; Based on the principle of factor graph, variable nodes, factor nodes and connections are pre-set to construct a factor graph model; The steps of defining the corresponding relationship between each positioning signal and the variable node and the factor node include: Define the corresponding relationship between the odometer signal, wireless signal, radar signal, control point signal and variable node of the variable signal in the positioning signal, and define the corresponding factor node of the inertial navigation signal; Based on the principle of factor graph, the factor graph model is constructed by connecting the inertial navigation signal and each variable node and the factor node; The outdoor non-restricted space signal acquisition and processing module is used to acquire the positioning signal of the outdoor non-restricted space, wherein the positioning signal of the outdoor non-restricted space includes: inertial navigation signal, odometer, and control point signal; the inertial navigation signal is used as the main signal of the positioning system, and other signals are used as auxiliary signals for correcting the positioning result; the inertial navigation information is solved by the pre-constructed factor graph model, and the positioning coordinates and control point coordinates calculated by the odometer are used as auxiliary correction information to calculate the corrected inertial navigation positioning result; The confined space signal acquisition and processing module is used to acquire the positioning signal of the confined space when entering the confined space, and use the positioning signal of the confined space in the joint positioning solution; wherein the confined space is a space where GNSS signals cannot be received; the positioning signal of the confined space includes: radio frequency signal, inertial navigation signal, odometer, control point, laser radar, and picture information; The preprocessing module is used to preprocess the received positioning signals of the confined space and the unconstrained space: the received LiDAR signals are processed through the point cloud preprocessing technology to obtain the point cloud features; the received image information is solved to obtain the orientation elements to obtain the image features; these two features are used as the features of each frame, and the change value between the two frames is identified through the feature matching technology to solve the coordinates of the LiDAR and the camera; A control point information acquisition module is used to acquire control point information and add the control point information as a smoothing correction item to the entire positioning process; The fusion processing and positioning module is used to obtain the pre-processed data values ​​of each positioning signal in a confined space, and input them into the constructed factor graph model according to the chronological order. The factor graph is expanded through the defined system variable nodes and factor nodes, and the recursion and update of the system state corresponding to the factor nodes of different sensors are realized to realize the correction of inertial navigation positioning; finally, the precise positioning coordinates are obtained to complete the fusion positioning.

7. The indoor and outdoor seamless positioning device based on multi-source fusion according to claim 6, characterized in that: It also includes: A definition module is used to define the corresponding relationship between the odometer signal, wireless signal, radar signal, control point signal and variable node of the variable signal in the positioning signal, and define the corresponding factor node of the inertial navigation signal; The coordinate conversion module is used to establish the coordinate conversion method of multi-source heterogeneous data and unify all heterogeneous positioning data into the 2000 coordinate system.

8. An intelligent terminal, characterized in that: The device comprises a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, and the one or more programs include being used to execute the method according to any one of claims 1 to 5.

9. A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the method as claimed in any one of claims 1 to 5.

Images