A mobile device lifelong positioning method and device based on double-layer maps

Abstract

The scheme provides a mobile device lifelong positioning method and device based on a double-layer map, comprising: constructing a forward propagation model, based on the interframe difference of a current laser frame and the actual pose state of the last laser frame, obtaining the predicted pose state of the current laser frame; constructing point-surface residual constraint, IMU information residual constraint and wheel speed meter information residual constraint; based on the point-surface residual constraint, the IMU information residual constraint and the wheel speed meter information residual constraint, calculating the Kalman gain, and using the Kalman gain to update the predicted pose state of the current laser frame to obtain the actual pose state; based on the actual pose state, obtaining the spatial position and attitude of the mobile device in the target area. The scheme realizes accurate positioning of the mobile device by constructing a double-layer point cloud map of the prior point cloud map and the local point cloud sub-map, and performing point-surface constraint of the laser point cloud information and the double-layer point cloud map in the iESKF framework.

Description

Technical Field

[0001] This application relates to the field of mobile device positioning, and in particular to a method and apparatus for lifetime positioning of mobile devices based on a two-layer map. Background Technology

[0002] In recent years, with the widespread application of autonomous driving and intelligent navigation technologies in large industrial environments, high-precision, long-term stable positioning technologies have become a research focus. However, achieving high-precision positioning in large-scale environments faces multiple challenges. While Real-Time Kinematic (RTK) positioning technology is widely adopted, its reliability is limited by the availability of satellite signals, especially in areas near metal structures or with severe obstructions, where positioning accuracy can drop from centimeter-level to meter-level. Furthermore, map-based positioning methods often suffer from performance instability due to environmental changes; for example, deviations between the map and the actual environment can lead to positioning errors. These limitations have driven the development of sensor-based positioning methods.

[0003] Early sensor-based localization schemes primarily relied on single sensors, such as LiDAR or cameras. However, these single-sensor solutions are susceptible to perception degradation in feature-scarce areas or when the environment undergoes significant changes, making them unsuitable for large-scale deployments. To overcome the limitations of single sensors, multi-sensor fusion systems have emerged, combining multiple complementary sensing methods (such as LiDAR and inertial measurement units) to improve localization robustness. Nevertheless, existing multi-sensor fusion methods still face issues of accumulated errors and long-term consistency in kilometer-scale operations, especially in dynamic environments where interference from moving objects further exacerbates perception distortion, leading to localization drift. Furthermore, map building for large-scale scenes is costly and difficult to maintain, and the high repetitiveness of the environment places stringent demands on localization accuracy.

[0004] In summary, existing LiDAR-based Simultaneous Localization and Mapping (SLAM) methods are prone to failure in dynamic environments due to map misalignment or odometer drift, making it difficult to meet the actual needs of industrial applications. There is an urgent need for a positioning method that can achieve high accuracy and long-term stability in dynamic, complex, and large-scale environments, dynamically adapting to environmental changes and maintaining long-term consistency in mobile device positioning. Summary of the Invention

[0005] This application provides a method and apparatus for lifetime positioning of mobile devices based on a two-layer map. It achieves accurate positioning of mobile devices by constructing a two-layer point cloud map consisting of a prior point cloud map and a local point cloud sub-map, and by performing point and surface constraints on laser point cloud information and the two-layer point cloud map under the iESKF framework.

[0006] In a first aspect, embodiments of this application provide a method for lifetime positioning of a mobile device based on a two-layer map, the method comprising:

[0007] A forward propagation model is constructed by taking a complete cycle of laser point cloud information collected by the LiDAR of a mobile device as a laser frame. The predicted pose state of the current laser frame is obtained based on the inter-frame difference of the current laser frame and the actual pose state of the previous laser frame. The forward propagation model uses the inter-frame difference of the laser frames as the step size. The predicted pose state includes the predicted laser pose of the mobile device, the predicted IMU information, and the predicted wheel speed meter information. A prior point cloud map is constructed based on fixed geometric features within the target area where the mobile device is located. A local point cloud sub-map is obtained based on the laser point cloud information acquired in the current laser frame. Point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints are constructed. The point-area residual constraints include residual constraints that match the laser point cloud information of the current laser frame with the prior point cloud map and the local point cloud sub-map, respectively. The IMU information residual constraints are the residual constraints between the actual IMU information and the predicted IMU information of the mobile device in the current laser frame. The wheel speed meter information residual constraints are the residual constraints between the actual wheel speed meter information and the predicted wheel speed meter information of the mobile device in the current laser frame. The Kalman gain is calculated based on the point-to-surface residual constraints, the IMU information residual constraints, and the wheel speed meter information residual constraints. The Kalman gain is then used to update the predicted pose state of the current laser frame to obtain the actual pose state. The spatial position and orientation of the mobile device within the target area are obtained based on its actual pose.

[0008] Secondly, embodiments of this application provide a mobile device lifetime positioning device based on a two-layer map, comprising: The forward propagation module takes a complete cycle of the LiDAR of the mobile device acquiring laser point cloud information once as a laser frame, constructs a forward propagation model, and obtains the predicted pose state of the current laser frame based on the inter-frame difference of the current laser frame and the actual pose state of the previous laser frame. The forward propagation model uses the inter-frame difference of the laser frames as the step size. The predicted pose state includes the predicted laser pose of the mobile device, the predicted IMU information, and the predicted wheel speed meter information. The constraint construction module constructs a prior point cloud map based on fixed geometric features within the target area where the mobile device is located, and obtains a local point cloud sub-map based on the laser point cloud information acquired in the current laser frame. It then constructs point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints. Specifically, the point-area residual constraints include residual constraints that match the laser point cloud information of the current laser frame with the prior point cloud map and the local point cloud sub-map, respectively. The IMU information residual constraints are the residual constraints between the actual IMU information and the predicted IMU information of the mobile device in the current laser frame. The wheel speed meter information residual constraints are the residual constraints between the actual wheel speed meter information and the predicted wheel speed meter information of the mobile device in the current laser frame. The state update module calculates the Kalman gain based on the point-to-surface residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints, and uses the Kalman gain to update the predicted pose state of the current laser frame to obtain the actual pose state. The positioning module obtains the spatial position and attitude of the mobile device within the target area based on its actual pose.

[0009] Thirdly, embodiments of this application provide an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to execute a method for lifetime positioning of a mobile device based on a two-layer map.

[0010] The main contributions and innovations of this invention are as follows: This application's embodiments achieve accurate positioning of mobile devices by constructing a two-layer point cloud map consisting of a prior point cloud map and a local point cloud sub-map, and by applying point and surface constraints between the laser point cloud information and the two-layer point cloud map within the iESKF framework. This scheme uses the inter-frame difference of the laser frames as the step size and constructs a forward propagation model based on a uniform motion model. The advantage is that it synchronizes the time dimension of pose prediction with the acquisition rhythm of the laser radar, avoiding prediction deviations caused by time misalignment. The assumption of uniform motion within a short time between frames closely matches the actual motion state of the device. This scheme interpolates the real-time acquisition information from the IMU and wheel speedometer to obtain the actual information at the end of the laser frame, thereby unifying the time dimensions of the laser, IMU, and wheel speedometer, solving the problem of inconsistent acquisition frequencies among the three types of sensors, making sensor information matching more accurate, and ensuring the effectiveness of residual constraint construction. After acquiring the actual pose state, this scheme synchronously updates the covariance matrix and uses it as the prior covariance for predicting the pose in the next frame. The advantage is that it achieves a closed loop in the positioning logic, allowing each pose prediction to be based on the error reference of the previous frame, continuously optimizing prediction accuracy and reducing cumulative errors.

[0011] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0012] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of a mobile device lifetime positioning method based on a two-layer map according to an embodiment of this application; Figure 2 This is a schematic diagram of a priori point cloud map and a local point cloud sub-map according to an embodiment of this application; Figure 3 This is a structural block diagram of a mobile device lifetime positioning device based on a dual-layer map according to an embodiment of this application; Figure 4 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0013] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with one or more embodiments of this specification. Rather, they are merely examples of apparatuses and methods consistent with some aspects of one or more embodiments of this specification as detailed in the appended claims.

[0014] It should be noted that the steps of the corresponding methods are not necessarily performed in the order shown and described in this specification in other embodiments. In some other embodiments, the methods may include more or fewer steps than described in this specification. Furthermore, a single step described in this specification may be broken down into multiple steps in other embodiments; and multiple steps described in this specification may be combined into a single step in other embodiments.

[0015] Example 1 This application provides a method for lifetime positioning of mobile devices based on a two-layer map. It constructs a two-layer point cloud map consisting of a prior point cloud map and a local point cloud sub-map, and achieves accurate positioning of the mobile device by applying point and surface constraints between laser point cloud information and the two-layer point cloud map within the iESKF framework. Specifically, refer to... Figure 1 The method includes: A complete cycle of laser point cloud information acquisition by the LiDAR of a mobile device is taken as a laser frame. A forward propagation model is constructed to obtain the predicted pose state of the current laser frame based on the inter-frame difference of the current laser frame and the actual pose state of the previous laser frame. The forward propagation model uses the inter-frame difference of the laser frames as the step size. The predicted pose state includes the predicted laser pose of the mobile device, the predicted IMU information, and the predicted wheel speed meter information. A prior point cloud map is constructed based on fixed geometric features within the target area where the mobile device is located. A local point cloud sub-map of the target area is constructed based on the laser point cloud information acquired in the current laser frame. Point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints are constructed. The point-area residual constraints include residual constraints that match the laser point cloud information of the current laser frame with the prior point cloud map and the local point cloud sub-map, respectively. The IMU information residual constraints are the residual constraints between the actual IMU information and the predicted IMU information of the mobile device in the current laser frame. The wheel speed meter information residual constraints are the residual constraints between the actual wheel speed meter information and the predicted wheel speed meter information of the mobile device in the current laser frame. The Kalman gain is calculated based on the point-to-surface residual constraints, the IMU information residual constraints, and the wheel speed meter information residual constraints. The Kalman gain is then used to update the predicted pose state of the current laser frame to obtain the actual pose state. The spatial position and orientation of the mobile device within the target area are obtained based on its actual pose.

[0016] In the current embodiment, the mobile device is equipped with a lidar, an IMU sensor, and a wheel speed meter to acquire laser point cloud data, IMU information, and wheel speed meter information in real time during the movement of the mobile device.

[0017] In the current embodiment, this scheme calculates the Kalman gain and updates the predicted pose state within the iESKF framework.

[0018] In the current embodiment, the inter-frame difference of the laser frame is based on the timestamp difference between the end and beginning of the current laser frame.

[0019] Furthermore, a forward propagation model is constructed based on the uniform motion model.

[0020] Specifically, since the time of the inter-frame difference is short, the displacement of the mobile device within the inter-frame difference can be regarded as uniform motion, and the pose state of the mobile device after uniform motion within the inter-frame difference can be predicted by forward propagation as the predicted pose state.

[0021] Specifically, this scheme uses the inter-frame difference as the step size for forward propagation, which allows the time dimension of pose prediction to be synchronized with the acquisition rhythm of the lidar, thus avoiding prediction deviations caused by time misalignment.

[0022] In addition, the forward propagation model predicts the actual pose state based on the previous laser frame, which ensures that the predicted pose of the current frame always closely matches the actual motion state of the device, making the prior covariance matrix more accurate.

[0023] Specifically, the formula for obtaining the predicted pose state is expressed as:

[0024] in, This represents the predicted pose state of the current laser frame. For inter-frame difference, This represents the actual pose state of the previous laser frame. This is the forward propagation function.

[0025] Furthermore, after obtaining the predicted pose state of the current laser frame, the forward propagation covariance matrix is ​​updated based on the characteristics of the uniform motion model and the noise of various sensors. Specifically, the updated covariance matrix is ​​used as the prior covariance matrix to provide an error reference for the calculation of Kalman gain in subsequent observation updates.

[0026] Specifically, the predicted laser pose is the six-degree-of-freedom spatial position and attitude information predicted in the global coordinate system under the current laser frame, which is the core component of the predicted pose state.

[0027] In the current embodiment, a prior point cloud map is obtained by performing global point cloud modeling on the geometric features within the target area.

[0028] Specifically, the prior map is a static reference benchmark for global positioning, providing global position constraints for the positioning of mobile devices, thereby avoiding positioning drift caused by accumulated errors when continuously positioning mobile devices.

[0029] In the current embodiment, a priori point cloud map and a local point cloud sub-map are constructed in the global coordinate system. The laser point cloud information acquired in the current laser frame is transformed into the global coordinate system to construct point-surface residual constraints. The point-surface residual constraints are constructed based on the global residual and the local residual. The formula for the global residual is expressed as:

[0030] in, For global residuals, Let i be the predicted pose state of the current frame i. Let j be the coordinates of the j-th laser point in the predicted pose state information of the i-th frame. The noise of the j-th laser point in the predicted pose state information of the i-th frame is... for The normal vector of the plane formed by the nearest points in the prior point cloud map M. This is the transformation matrix between the global coordinate system and the corresponding IMU coordinate system. This is the transformation matrix between the coordinate system of the laser point cloud information and the corresponding IMU coordinate system. Let J be the associated point in the prior point cloud map M corresponding to the j-th laser point.

[0031] The formula for local residuals is expressed as:

[0032] in, For local residuals, for Local point cloud submap The normal vector of the plane formed by the nearest points in the vector. For the j-th laser point, the corresponding local point cloud sub-map Related points within.

[0033] The formula for point-to-surface residual constraints is expressed as: .

[0034] This solution ensures the accuracy of mobile device positioning within the target area by constraining the point-to-surface residual to be as close to zero as possible. The point-to-surface residual constraint can fully combine global and local information, making the positioning more accurate.

[0035] In the current embodiment, the IMU information collected in real time by the mobile device is interpolated to obtain the actual IMU information of the current laser frame end time. Based on the actual IMU information of the current laser frame end time, an IMU information residual constraint is constructed. This IMU information residual constraint is constructed based on the linear acceleration residual and the angular velocity residual. The formula for the linear acceleration residual is expressed as:

[0036] in, For linear acceleration residuals, Let i be the predicted pose state of the current frame i. This represents the actual linear acceleration at the end of the current laser frame. To predict linear acceleration in the pose state, For linear acceleration measurement noise, The deviation is the linear acceleration. Noise is processed for random walks of linear acceleration; The formula for the angular velocity residual is expressed as:

[0037] in, For angular velocity residuals, The actual angular velocity at the end of the current laser frame. To predict the angular velocity in the pose state, For angular velocity measurement noise, This is the measurement deviation of angular velocity. Noise is processed for random walks of angular velocities; The IMU information residual constraint is: .

[0038] Specifically, a single lidar may experience perception degradation in open environments, while an IMU is a high-frequency acquisition sensor. This solution can provide continuous motion state correction when laser observation fails by constructing IMU information residual constraints, thus avoiding positioning interruption or large inter-frame drift.

[0039] In other words, the failure of any of the sensors—laser, IMU, and wheel speed meter—will not affect the pose estimation of this solution, especially in degraded environments such as open spaces and long corridors, where laser observation may fail, but pose estimation can still be achieved based on the IMU and wheel speed meter.

[0040] In the current embodiment, the wheel speed meter information collected in real time by the mobile device is interpolated to obtain the actual wheel speed meter information at the end time of the current laser frame. Based on the actual wheel speed meter information at the end time of the current laser frame, a wheel speed meter information residual constraint is constructed. The formula for the wheel speed meter information residual constraint is expressed as:

[0041] in, For wheel speed gauge information residual, This is the actual wheel speed measurement information for the current laser frame tail time. The actual measurement noise of the actual wheel speed gauge information. This represents the actual measurement deviation of the actual wheel speed gauge information. To handle the random walk noise of the actual wheel speed meter information, This is for predicting wheel velocity information in the pose state.

[0042] Specifically, since the acquisition frequency of the IMU sensor and wheel speed meter is much higher than that of the lidar, in order to obtain the IMU information and wheel speed meter information at the end of the lidar frame in a unified manner, the wheel speed meter information and IMU information are obtained by interpolation.

[0043] Specifically, the wheel speed meter information in this scheme is linear velocity. By introducing wheel speed meter information residual constraints, the accumulated error of the IMU is suppressed, the translational state estimation is stabilized, and redundant speed observations are provided.

[0044] In the current embodiment, the Kalman gain is calculated based on point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints within the iESKF framework. Relying on the core characteristics of iESKF's incremental processing of nonlinearity and focused state error estimation, the three complementary types of residual constraints participate in the construction of the observation model and the solution of the gain. The core benefit is that the weight allocation of the Kalman gain is more accurate, the state correction is more in line with the real motion, and the filter is more stable and resistant to divergence, thus perfectly adapting to the lifelong positioning requirements of large scenes and high dynamic environments.

[0045] Specifically, after using Kalman gain to update the predicted pose state of the current laser frame to obtain the actual pose state, the covariance matrix is ​​updated, and the updated covariance matrix is ​​used as the prior covariance for obtaining the predicted pose state of the next laser frame, so as to achieve the overall logical closed loop.

[0046] In the current embodiment, after obtaining the spatial position and attitude of the mobile device in the target area, the laser point cloud information of the current laser frame obtained by the mobile device is inserted into the local dynamic sub-map, and the expired point cloud data in the local dynamic sub-map is deleted to obtain the updated local dynamic sub-map. The updated local dynamic sub-map is used as the local point cloud sub-map of the next laser frame. The expired point cloud data is point cloud data whose spatial distance from the mobile device is greater than a first threshold, or point cloud data whose existence time is greater than a second threshold.

[0047] Furthermore, the local point cloud sub-maps in this solution can be cached on mobile devices for use. This solution rasterizes the prior point cloud map and the local dynamic sub-map, and uses an octree for data management within each grid cell. A schematic diagram of the prior point cloud map and the local dynamic sub-map is shown below. Figure 2 As shown, in Figure 2 The purple area represents the prior point cloud map, while the other colors represent different local dynamic sub-maps.

[0048] In other words, the prior point cloud map in this solution records all fixed geometric features within the target area, such as walls, fixed shelves, facilities, etc., while the local dynamic sub-map records real-time obstacles near the mobile device, such as pedestrians or other debris. The mobile device is accurately located based on the constraints of the prior point cloud map and the local dynamic sub-map.

[0049] Specifically, the laser point cloud information of the current laser frame is converted to the global coordinate system and inserted into the local dynamic sub-map. This scheme updates the local dynamic sub-map in real time to ensure the lightweight and accurate features of the local dynamic sub-map, thereby providing a highly matched local environment reference for the lidar to construct point-area constraints, and ultimately improving the positioning accuracy.

[0050] Specifically, before using the laser point cloud information acquired by the mobile device, this solution preprocesses the real-time acquired laser point cloud information, removing those with a distance less than a threshold from the laser origin. and greater than The data is used to interpolate the laser pose of each laser point at the corresponding time based on the predicted pose information estimated by the forward propagation model, thereby compensating for laser distortion caused by motion.

[0051] In the current embodiment, after obtaining the spatial position and attitude of the mobile device in the target area, a preset size area is cropped out of the prior point cloud map with the spatial position of the mobile device on the prior point cloud map as the center. The cropped prior point cloud map is used to calculate the residual constraint between the laser point cloud information of the next laser frame and the prior point cloud map.

[0052] Specifically, directly loading the complete prior point cloud map into the mobile device would directly lead to memory overload; if the residual constraints of the laser point cloud are directly calculated based on the complete prior map, the computational load of nearest point search and point-area constraint construction will increase exponentially, completely exceeding the processing capacity of the onboard computing power, resulting in a significant decrease in the real-time performance of positioning. Therefore, this solution uses the spatial position of the mobile device on the prior point cloud map as the center to crop the complete prior point cloud map, in order to ensure the real-time performance and stability of the positioning system in large-scale scenarios.

[0053] Example 2 Based on the same concept, referencing Figure 3 This application also proposes a mobile device lifetime positioning device based on a two-layer map, comprising: The forward propagation module takes a complete cycle of the LiDAR of the mobile device acquiring laser point cloud information once as a laser frame, constructs a forward propagation model, and obtains the predicted pose state of the current laser frame based on the inter-frame difference of the current laser frame and the actual pose state of the previous laser frame. The forward propagation model uses the inter-frame difference of the laser frames as the step size. The predicted pose state includes the predicted laser pose of the mobile device, the predicted IMU information, and the predicted wheel speed meter information. The constraint construction module constructs a prior point cloud map based on fixed geometric features within the target area where the mobile device is located, and obtains a local point cloud sub-map based on the laser point cloud information acquired in the current laser frame. It then constructs point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints. Specifically, the point-area residual constraints include residual constraints that match the laser point cloud information of the current laser frame with the prior point cloud map and the local point cloud sub-map, respectively. The IMU information residual constraints are the residual constraints between the actual IMU information and the predicted IMU information of the mobile device in the current laser frame. The wheel speed meter information residual constraints are the residual constraints between the actual wheel speed meter information and the predicted wheel speed meter information of the mobile device in the current laser frame. The state update module calculates the Kalman gain based on the point-to-surface residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints, and uses the Kalman gain to update the predicted pose state of the current laser frame to obtain the actual pose state. The positioning module obtains the spatial position and attitude of the mobile device within the target area based on its actual pose.

[0054] Example 3 This embodiment also provides an electronic device, see reference. Figure 4 It includes a memory 404 and a processor 402, the memory 404 storing a computer program and the processor 402 being configured to run the computer program to perform the steps in any of the above method embodiments.

[0055] Specifically, the processor 402 may include a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0056] Memory 404 may include a mass storage device for data or instructions. For example, and not limitingly, memory 404 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 404 may include removable or non-removable (or fixed) media. Where appropriate, memory 404 may be internal or external to a data processing device. In a particular embodiment, memory 404 is non-volatile memory. In a particular embodiment, memory 404 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0057] The memory 404 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 402.

[0058] The processor 402 reads and executes computer program instructions stored in the memory 404 to implement any of the two-layer map-based mobile device lifetime positioning methods in the above embodiments.

[0059] Optionally, the electronic device may further include a transmission device 406 and an input / output device 408, wherein the transmission device 406 is connected to the processor 402, and the input / output device 408 is connected to the processor 402.

[0060] The transmission device 406 can be used to receive or send data via a network. Specific examples of the network described above may include wired or wireless networks provided by the communication provider of the electronic device. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 406 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0061] The input / output device 408 is used to input or output information. In this embodiment, the input information may be the predicted pose state, etc., and the output information may be the spatial position and attitude of the mobile device within the target area, etc.

[0062] Optionally, in this embodiment, the processor 402 can be configured to perform the following steps via a computer program: A forward propagation model is constructed by taking a complete cycle of laser point cloud information collected by the LiDAR of a mobile device as a laser frame. The predicted pose state of the current laser frame is obtained based on the inter-frame difference of the current laser frame and the actual pose state of the previous laser frame. The forward propagation model uses the inter-frame difference of the laser frames as the step size. The predicted pose state includes the predicted laser pose of the mobile device, the predicted IMU information, and the predicted wheel speed meter information. A prior point cloud map is constructed based on fixed geometric features within the target area where the mobile device is located. A local point cloud sub-map is obtained based on the laser point cloud information acquired in the current laser frame. Point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints are constructed. The point-area residual constraints include residual constraints that match the laser point cloud information of the current laser frame with the prior point cloud map and the local point cloud sub-map, respectively. The IMU information residual constraints are the residual constraints between the actual IMU information and the predicted IMU information of the mobile device in the current laser frame. The wheel speed meter information residual constraints are the residual constraints between the actual wheel speed meter information and the predicted wheel speed meter information of the mobile device in the current laser frame. The Kalman gain is calculated based on the point-to-surface residual constraints, the IMU information residual constraints, and the wheel speed meter information residual constraints. The Kalman gain is then used to update the predicted pose state of the current laser frame to obtain the actual pose state. The spatial position and orientation of the mobile device within the target area are obtained based on its actual pose.

[0063] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0064] Generally, various embodiments can be implemented in hardware or dedicated circuitry, software, logic, or any combination thereof. Some aspects of the invention can be implemented in hardware, while others can be implemented by firmware or software executed by a controller, microprocessor, or other computing device, but the invention is not limited thereto. Although various aspects of the invention may be shown and described as block diagrams, flowcharts, or using some other graphical representation, it should be understood that, by way of non-limiting example, these blocks, apparatuses, systems, techniques, or methods described herein can be implemented in hardware, software, firmware, dedicated circuitry or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.

[0065] Embodiments of the present invention can be implemented by computer software, which may be executable by a data processor of a mobile device, such as a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products) including software routines, applets, and / or macros can be stored in any device-readable data storage medium, and they include program instructions for performing specific tasks. The computer program product may include one or more computer-executable components configured to perform the embodiments when the program is run. The one or more computer-executable components may be at least one piece of software code or a portion thereof. Additionally, it should be noted in this respect that, as Figure 4Any box in the logical flow can represent a program step, or interconnected logic circuits, boxes and functions, or a combination of program steps and logic circuits, boxes and functions. Software can be stored on physical media such as memory chips or blocks of storage implemented within a processor, magnetic media such as hard disks or floppy disks, and optical media such as DVDs and their data variants, CDs, etc. The physical medium is a non-transient medium.

[0066] Those skilled in the art should understand that the technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0067] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for lifetime positioning of mobile devices based on a two-layer map, characterized in that, Includes the following steps: A forward propagation model is constructed by taking a complete cycle of laser point cloud information collected by the LiDAR of a mobile device as a laser frame. The predicted pose state of the current laser frame is obtained based on the inter-frame difference of the current laser frame and the actual pose state of the previous laser frame. The forward propagation model uses the inter-frame difference of the laser frames as the step size. The predicted pose state includes the predicted laser pose of the mobile device, the predicted IMU information, and the predicted wheel speed meter information. A prior point cloud map is constructed based on fixed geometric features within the target area where the mobile device is located. A local point cloud sub-map is obtained based on the laser point cloud information acquired in the current laser frame. Point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints are constructed. The point-area residual constraints include residual constraints that match the laser point cloud information of the current laser frame with the prior point cloud map and the local point cloud sub-map, respectively. The IMU information residual constraints are the residual constraints between the actual IMU information and the predicted IMU information of the mobile device in the current laser frame. The wheel speed meter information residual constraints are the residual constraints between the actual wheel speed meter information and the predicted wheel speed meter information of the mobile device in the current laser frame. The Kalman gain is calculated based on the point-to-surface residual constraints, the IMU information residual constraints, and the wheel speed meter information residual constraints. The Kalman gain is then used to update the predicted pose state of the current laser frame to obtain the actual pose state. The spatial position and orientation of the mobile device within the target area are obtained based on its actual pose.

2. The method for lifetime positioning of mobile devices based on a dual-layer map according to claim 1, characterized in that, The difference between the timestamps of the current laser frame's end and beginning is used as the inter-frame difference of the laser frame, and a forward propagation model is constructed based on the uniform motion model.

3. The method for lifetime positioning of mobile devices based on a dual-layer map according to claim 1, characterized in that, A priori point cloud map and a local point cloud sub-map are constructed in the global coordinate system. The laser point cloud information acquired in the current laser frame is transformed into the global coordinate system to construct point-surface residual constraints. The point-surface residual constraints are constructed based on the global residual and the local residual. The formula for the global residual is expressed as: in, For global residuals, Let i be the predicted pose state of the current frame i. Let j be the coordinates of the j-th laser point in the predicted pose state information of the i-th frame. The noise of the j-th laser point in the predicted pose state information of the i-th frame is... for The normal vector of the plane formed by the nearest points in the prior point cloud map M. This is the transformation matrix between the global coordinate system and the corresponding IMU coordinate system. This is the transformation matrix between the coordinate system of the laser point cloud information and the corresponding IMU coordinate system. Let the j-th laser point be the associated point in the prior point cloud map M; The formula for local residuals is expressed as: in, For local residuals, for Local point cloud submap The normal vector of the plane formed by the nearest points in the vector. For the j-th laser point, the corresponding local point cloud sub-map Related points within; The formula for point-to-surface residual constraints is expressed as: 。 4. The method for lifetime positioning of mobile devices based on a dual-layer map according to claim 1, characterized in that, Interpolate the IMU information collected in real time by the mobile device to obtain the actual IMU information at the end time of the current laser frame, and construct IMU information residual constraints based on the actual IMU information at the end time of the current laser frame; interpolate the wheel speed meter information collected in real time by the mobile device to obtain the actual wheel speed meter information at the end time of the current laser frame, and construct wheel speed meter information residual constraints based on the actual wheel speed meter information at the end time of the current laser frame.

5. The method for lifetime positioning of mobile devices based on a dual-layer map according to claim 1, characterized in that, The IMU information residual constraint is constructed based on the linear acceleration residual and the angular velocity residual, and the formula for the linear acceleration residual is expressed as: in, For linear acceleration residuals, Let i be the predicted pose state of the current frame i. This represents the actual linear acceleration at the end of the current laser frame. To predict linear acceleration in the pose state, For linear acceleration measurement noise, The deviation is the linear acceleration. Noise is processed for random walks of linear acceleration; The formula for the angular velocity residual is expressed as: in, For angular velocity residuals, The actual angular velocity at the end of the current laser frame. To predict the angular velocity in the pose state, For angular velocity measurement noise, The deviation is the angular velocity. Noise is processed for random walks of angular velocities; The IMU information residual constraint is: 。 6. The method for lifetime positioning of mobile devices based on a dual-layer map according to claim 1, characterized in that, The formula for the residual constraint of the wheel speed gauge information is expressed as follows: in, For wheel speed gauge information residual, This is the actual wheel speed measurement information for the current laser frame tail time. The actual measurement noise of the actual wheel speed gauge information. This represents the actual measurement deviation of the actual wheel speed gauge information. To handle the random walk noise of the actual wheel speed meter information, This is for predicting wheel velocity information in the pose state.

7. The method for lifetime positioning of mobile devices based on a dual-layer map according to claim 1, characterized in that, After obtaining the spatial position and attitude of the mobile device within the target area, the laser point cloud information of the current laser frame obtained by the mobile device is inserted into the local dynamic sub-map, and expired point cloud data in the local dynamic sub-map is deleted to obtain an updated local dynamic sub-map. The updated local dynamic sub-map serves as the local point cloud sub-map for the next laser frame. The expired point cloud data refers to point cloud data whose spatial distance from the mobile device is greater than a first threshold, or point cloud data whose existence time is greater than a second threshold.

8. The method for lifetime positioning of mobile devices based on a dual-layer map according to claim 1, characterized in that, After obtaining the spatial position and attitude steps of the mobile device in the target area, a preset size area is cropped from the spatial position of the mobile device on the prior point cloud map as the center. This cropped prior point cloud map is used to calculate the residual constraint between the laser point cloud information of the next laser frame and the prior point cloud map.

9. A lifetime positioning device for mobile devices based on a dual-layer map, characterized in that, include: The forward propagation module takes a complete cycle of the LiDAR of the mobile device acquiring laser point cloud information once as a laser frame. It constructs a forward propagation model to obtain the predicted pose state of the current laser frame based on the inter-frame difference of the current laser frame and the actual pose state of the previous laser frame. The forward propagation model uses the inter-frame difference of the laser frames as the step size. The predicted pose state includes the predicted laser pose of the mobile device, the predicted IMU information, and the predicted wheel speed meter information. The constraint construction module constructs a prior point cloud map based on fixed geometric features within the target area where the mobile device is located, and obtains a local point cloud sub-map based on the laser point cloud information acquired in the current laser frame. It then constructs point-area residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints. Specifically, the point-area residual constraints include residual constraints that match the laser point cloud information of the current laser frame with the prior point cloud map and the local point cloud sub-map, respectively. The IMU information residual constraints are the residual constraints between the actual IMU information and the predicted IMU information of the mobile device in the current laser frame. The wheel speed meter information residual constraints are the residual constraints between the actual wheel speed meter information and the predicted wheel speed meter information of the mobile device in the current laser frame. The state update module calculates the Kalman gain based on the point-to-surface residual constraints, IMU information residual constraints, and wheel speed meter information residual constraints, and uses the Kalman gain to update the predicted pose state of the current laser frame to obtain the actual pose state. The positioning module obtains the spatial position and attitude of the mobile device within the target area based on its actual pose.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform a method for lifetime positioning of a mobile device based on a dual-layer map as described in any one of claims 1-8.

Images