Robot positioning method and device, computer equipment, storage medium and program product

Abstract

The embodiment of the invention provides a robot positioning method and device, computer equipment, a storage medium and a computer program product, and relates to the technical field of computer vision. The method comprises the following steps: acquiring a current temperature and point cloud fusion feature matrix; performing feature coding on the temperature and point cloud fusion feature matrix to obtain a current coding feature vector; according to the historical coding feature vector and the current coding feature vector of the previous time step adjacent to the current time step, a relative position change matrix of the target robot between the previous time step and the current time step is obtained, and a temperature feature difference matrix of the target robot between the previous time step and the current time step is obtained; and according to the relative position change matrix and the temperature characteristic difference matrix, obtaining target current pose information of the target robot at the current time step. According to the method, the positioning accuracy, the path planning efficiency and accuracy and the map updating accuracy of the corresponding space are improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a robot positioning method, device, computer equipment, storage medium and computer program product. Background Art

[0002] In some scenarios, such as data center environments, robots are needed to conduct inspections to fully perceive the environmental conditions of the data center, such as equipment operating temperature and room space layout. Among them, the positioning of the inspection robot is crucial.

[0003] The current positioning technology of inspection robots mainly relies on traditional laser vision technology, which uses a single data source, such as point cloud data, to position the inspection robots. In an environment with a similar layout to that of a data center, a "false loop" phenomenon is prone to occur, resulting in positioning errors and low positioning accuracy. Summary of the invention

[0004] Based on this, it is necessary to provide a robot positioning method, device, computer equipment, storage medium and computer program product to address the above technical problems.

[0005] In a first aspect, the present application provides a robot positioning method. The method comprises:

[0006] Obtain the current temperature and point cloud fusion feature matrix of the target robot in the target space at the current time step;

[0007] Performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix;

[0008] According to the historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector, a relative position change matrix of the target robot between the previous time step and the current time step is obtained, and a temperature feature difference matrix of the target robot between the previous time step and the current time step is obtained;

[0009] According to the relative position change matrix and the temperature characteristic difference matrix, the target current posture information of the target robot at the current time step is obtained.

[0010] In one of the embodiments, the feature encoding of the temperature and point cloud fusion feature matrix to obtain the current encoded feature vector of the temperature and point cloud fusion feature matrix includes: constructing an initial state space model for the target space; obtaining feature association parameters of the initial state space model based on the current temperature and point cloud fusion feature matrix; obtaining a target state space model for the target space based on the feature association parameters and the initial state space model; and using the target state space model to feature encode the temperature and point cloud fusion feature matrix to obtain the current encoded feature vector of the temperature and point cloud fusion feature matrix.

[0011] In one of the embodiments, the method of obtaining the relative position change matrix of the target robot between the previous time step and the current time step includes: determining a first point cloud subset in a historical point cloud data set corresponding to the previous time step; screening out a second point cloud subset matching the first point cloud subset in a current point cloud data set corresponding to the current time step based on the historical coding feature vector and the current coding feature vector; there is a corresponding relationship between a plurality of second point cloud data included in the second point cloud subset and a plurality of first point cloud data included in the first point cloud subset; and obtaining the relative position change matrix based on difference information between the second point cloud subset and the first point cloud subset.

[0012] In one embodiment, obtaining the temperature feature difference matrix of the target robot between the previous time step and the current time step includes: obtaining temperature change information of a current temperature feature matrix corresponding to the current time step and a historical temperature feature matrix corresponding to the previous time step according to the historical coding feature vector and the current coding feature vector; and obtaining the temperature feature difference matrix according to the temperature change information.

[0013] In one of the embodiments, obtaining the target current pose information of the target robot at the current time step according to the relative position change matrix and the temperature feature difference matrix includes: constructing an initial pose prediction function and an initial prediction error covariance function of the target robot in the target space; obtaining a target pose prediction function corresponding to the initial pose prediction function and a target prediction error covariance function corresponding to the initial prediction error covariance function according to the relative position change matrix and the temperature feature difference matrix; and obtaining the target current pose information of the target robot at the current time step using the target pose prediction function and the target prediction error covariance function.

[0014] In one of the embodiments, obtaining the target pose prediction function corresponding to the initial pose prediction function according to the relative position change matrix and the temperature feature difference matrix includes: obtaining the temperature observation residual of the temperature feature difference matrix for the initial pose prediction function according to the temperature feature difference matrix; obtaining the first optimized pose prediction function corresponding to the initial pose prediction function according to the temperature observation residual; obtaining the spatial observation residual of the relative position change matrix for the initial pose prediction function according to the relative position change matrix; obtaining the second optimized pose prediction function corresponding to the initial pose prediction function according to the spatial observation residual; obtaining the target pose prediction function corresponding to the initial pose prediction function according to the first optimized pose prediction function and the second optimized pose prediction function.

[0015] In one of the embodiments, obtaining a first optimized pose prediction function corresponding to the initial pose prediction function based on the temperature observation residual includes: obtaining the temperature Kalman gain of the temperature feature difference matrix for the initial pose prediction function; obtaining the first optimized pose prediction function corresponding to the initial pose prediction function based on the temperature observation residual and the temperature Kalman gain.

[0016] In one of the embodiments, obtaining a second optimized pose prediction function corresponding to the initial pose prediction function based on the spatial observation residual includes: obtaining the spatial Kalman gain of the relative position change matrix for the initial pose prediction function; obtaining the second optimized pose prediction function corresponding to the initial pose prediction function based on the spatial observation residual and the spatial Kalman gain.

[0017] In one embodiment, obtaining the target prediction error covariance function corresponding to the initial prediction error covariance function includes: obtaining the temperature Kalman gain of the temperature feature difference matrix for the initial posture prediction function; obtaining the first optimized prediction error covariance function corresponding to the initial prediction error covariance function based on the temperature Kalman gain; obtaining the spatial Kalman gain of the relative position change matrix for the initial posture prediction function; obtaining the second optimized prediction error covariance function corresponding to the initial prediction error covariance function based on the spatial Kalman gain; and obtaining the target prediction error covariance function corresponding to the initial prediction error covariance function based on the first optimized prediction error covariance function and the second optimized prediction error covariance function.

[0018] In a second aspect, the present application provides a robot positioning device. The device comprises:

[0019] A fusion module is used to obtain the current temperature of the target robot in the target space at the current time step and the point cloud fusion feature matrix;

[0020] An encoding module, used for performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix;

[0021] A first calculation module is used to obtain a relative position change matrix of the target robot between the previous time step and the current time step, and to obtain a temperature feature difference matrix of the target robot between the previous time step and the current time step according to a historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector;

[0022] The second calculation module is used to obtain the target current posture information of the target robot at the current time step according to the relative position change matrix and the temperature feature difference matrix.

[0023] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0024] Obtain the current temperature and point cloud fusion feature matrix of the target robot in the target space at the current time step;

[0025] Performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix;

[0026] According to the historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector, a relative position change matrix of the target robot between the previous time step and the current time step is obtained, and a temperature feature difference matrix of the target robot between the previous time step and the current time step is obtained;

[0027] According to the relative position change matrix and the temperature characteristic difference matrix, the target current posture information of the target robot at the current time step is obtained.

[0028] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0029] Obtain the current temperature and point cloud fusion feature matrix of the target robot in the target space at the current time step;

[0030] Performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix;

[0031] According to the historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector, a relative position change matrix of the target robot between the previous time step and the current time step is obtained, and a temperature feature difference matrix of the target robot between the previous time step and the current time step is obtained;

[0032] According to the relative position change matrix and the temperature characteristic difference matrix, the target current posture information of the target robot at the current time step is obtained.

[0033] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0034] Obtain the current temperature and point cloud fusion feature matrix of the target robot in the target space at the current time step;

[0035] Performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix;

[0036] According to the historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector, a relative position change matrix of the target robot between the previous time step and the current time step is obtained, and a temperature feature difference matrix of the target robot between the previous time step and the current time step is obtained;

[0037] According to the relative position change matrix and the temperature characteristic difference matrix, the target current posture information of the target robot at the current time step is obtained.

[0038] In the above-mentioned robot positioning method, device, computer equipment, storage medium and computer program product, first, the current temperature and point cloud fusion feature matrix of the target robot in the target space at the current time step can be obtained; next, the temperature and point cloud fusion feature matrix can be feature encoded to obtain the current encoding feature vector of the temperature and point cloud fusion feature matrix; further, based on the historical encoding feature vector of the target robot in the previous time step adjacent to the current time step and the current encoding feature vector, the relative position change matrix of the target robot between the previous time step and the current time step can be obtained, as well as the temperature feature difference matrix of the target robot between the previous time step and the current time step; finally, based on the relative position change matrix and the temperature feature difference matrix, the target current posture information of the target robot at the current time step can be obtained. In the method provided in the embodiment of the present application, temperature data and point cloud data can be combined to generate a temperature and point cloud fusion feature matrix containing temperature features and spatial features, which can improve the distinguishability of the target robot to the environment and the positioning accuracy, and the temperature and point cloud fusion feature matrix can be feature encoded to improve the accuracy of feature extraction, thereby improving the positioning accuracy. In addition, the iterative closest point algorithm and Kalman filter can be used to predict the position and posture of the target robot, thereby improving the accuracy of the position and posture prediction, thereby further improving the positioning accuracy, the efficiency and accuracy of path planning, and the accuracy of map updating in the corresponding space. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments of the present application or the related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A schematic diagram of a flow chart of a robot positioning method provided in an embodiment of the present application;

[0041] Figure 2 A schematic diagram of a Mamba network coding process provided in an embodiment of the present application;

[0042] Figure 3 A schematic diagram of a process for obtaining a current encoding feature vector provided in an embodiment of the present application;

[0043] Figure 4 A schematic diagram of a process for obtaining a relative position change matrix provided in an embodiment of the present application;

[0044] Figure 5A schematic diagram of a process for obtaining a temperature characteristic difference matrix provided in an embodiment of the present application;

[0045] Figure 6 A schematic diagram of a process for obtaining current position information of a target provided in an embodiment of the present application;

[0046] Figure 7 A schematic diagram of a process for obtaining a target pose prediction function provided in an embodiment of the present application;

[0047] Figure 8 A structural block diagram of a robot positioning device provided in an embodiment of the present application;

[0048] Fig. 9 An internal structure diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] In an exemplary embodiment, Figure 1 As shown, a robot positioning method is provided. This embodiment takes the method applied to a server as an example. It can be understood that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0051] Step 102, obtaining the current temperature of the target robot in the target space at the current time step and the point cloud fusion feature matrix.

[0052] The target space can be any space that needs to be inspected, such as a computer room and a data center. The target robot can be equipped with multiple sensors, such as a high-precision infrared temperature measurement camera and a laser radar. In this step, the target robot synchronously collects temperature matrix and point cloud data during movement. The high-precision infrared temperature measurement camera can be used to collect temperature matrix data in real time. The camera measures the temperature by receiving infrared radiation emitted by the surface of objects in the environment and generates a two-dimensional temperature matrix. At the same time, the target robot can also continuously capture spatial point cloud data using laser radar. , where it is necessary to ensure that the temperature matrix and the acquisition timestamp of the point cloud data are strictly synchronized. At the same time, that is, at the same time step, the acquired temperature and point cloud data can correspond to each other, so as to facilitate the subsequent effective fusion processing. For example, a high-precision clock synchronization module can be used to calibrate the acquisition time of the infrared temperature measurement camera and the laser radar so that the two are consistent in time. Next, the collected temperature data and point cloud data can be environmentally compensated and spatially corrected to ensure that the two are aligned in a unified coordinate system. The environmental compensation mainly considers the influence of environmental factors (such as light, reflection, etc.) on temperature measurement and point cloud acquisition, and corrects them through corresponding algorithms; spatial correction converts the point cloud data and temperature matrix to the same spatial coordinate system, and usually requires coordinate transformation and calibration operations, such as determining the conversion relationship between the laser radar coordinate system and the infrared camera coordinate system, so that the two can accurately match, laying the foundation for subsequent fusion operations. In addition, the collected temperature data and point cloud data can also be standardized, the temperature matrix can be normalized, and the temperature value can be mapped to the standard range for subsequent fusion. The point cloud data is processed through filtering and downsampling to retain key points and reduce the amount of calculation.

[0053] Furthermore, the current temperature of the target robot in the target space at the current time step can be fused with the point cloud feature matrix based on the temperature matrix and point cloud data after data processing. Specifically, the following can be included: Step 1: projection calculation. Using the internal and external parameters of the infrared camera, the point cloud data is projected onto the two-dimensional pixel plane of the infrared camera. See equations (1) and (2). The internal parameters of the infrared camera include the focal length , optical center coordinates Etc., the external parameters describe the position and attitude information of the camera in space.

[0054]

[0055] Among them, is the pixel coordinate after projection, is the three-dimensional coordinate of the point cloud point, is the focal length, For the heart of light.

[0056] The second step is temperature binding, which is to associate pixel coordinates with temperature values. , find the corresponding temperature value in the temperature matrix , and attach it to the point cloud point to form the corresponding extension point, as shown in equation (3). Since the correspondence between the point cloud point and the pixel coordinates is established in the projection calculation, the temperature value corresponding to each point cloud point can be accurately found in the temperature matrix through this correspondence. Therefore, each extension point not only contains its own spatial coordinate information, but also carries the corresponding temperature information, realizing the fusion of temperature and spatial features at the point cloud level, and obtaining the initial temperature and point cloud fusion feature matrix. For example, for a point cloud point on a cabinet in a data center, through projection and temperature binding operations, the point cloud point not only has the coordinate information representing its position in space, but also is associated with the temperature value at that position, thereby forming point cloud data containing temperature-spatial features.

[0057]

[0058] In the third step, the initial temperature and point cloud fusion feature matrix can be filtered and downsampled. The purpose of filtering is to remove outliers and noise in the data. For example, through statistical filtering methods, according to the density distribution of points in the point cloud, outliers with density significantly lower or higher than the surrounding points are removed. At the same time, the temperature data is smoothed to remove abnormal temperature values ​​caused by sensor errors or environmental interference to ensure the stability and reliability of the data. Downsampling is to reduce the amount of data and improve computational efficiency while retaining the key features of the data. For example, the voxel filtering method is used to divide the point cloud space into voxel grids of a certain size, and a representative point is selected in each voxel grid, thereby reducing the number of point cloud points. Then, the current temperature and point cloud fusion feature matrix corresponding to the initial temperature and point cloud fusion feature matrix is ​​obtained, which can be seen in equation (4).

[0059]

[0060] Step 104 , feature encoding is performed on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix.

[0061] Among them, in this step, if Figure 2As shown, the Mamba network can be used to encode the temperature and point cloud fusion feature matrix, and the Mamba module can be used to obtain the encoded position and temperature features based on the fused temperature-point cloud feature matrix. The core principle of the Mamba network is based on the selective state space model (State Space Model, SSM). By expressing the parameters of the state space model as a function of the input features, the dynamic modeling of the input features is realized. Specifically, it can include: the first step is to build an initial state space model for the target space; the second step is to obtain the feature association parameters of the initial state space model according to the current temperature and point cloud fusion feature matrix. The Mamba network can express the parameters of the initial state space model as a function of the input features, that is, the function of the current temperature and point cloud fusion feature matrix, and obtain the feature association parameters. See formula (5):

[0062]

[0063] The third step is to obtain the target state space model for the target space according to the feature association parameters and the initial state space model. The initial state space model is updated using the feature association parameters to obtain the target state space model, which can be seen in formula (6):

[0064]

[0065] in, and is an adaptive parameter calculated based on the current input. At each time step, the Mamba network dynamically adjusts the current temperature of the current input and the point cloud fusion feature matrix and The value of , thereby updating the state . This process enables the Mamba network to track changes in environmental features. For example, when the robot moves to different areas of the data center and the temperature and spatial structure change, the network accurately reflects these changes by adjusting parameters and updating states, thereby effectively modeling the current temperature and point cloud fusion feature matrix. In the fourth step, the Mamba network uses the target state space model to encode the temperature and point cloud fusion feature matrix to obtain the current encoded feature vector of the temperature and point cloud fusion feature matrix. The current encoded feature vector is an abstract representation of the temperature and spatial features of the current frame. It integrates temperature and spatial information and can be compared with the encoded vectors of historical frames in subsequent processing.

[0066] Step 106, based on the historical coding feature vector of the target robot in the previous time step adjacent to the current time step and the current coding feature vector, obtain the relative position change matrix of the target robot between the previous time step and the current time step, and obtain the temperature feature difference matrix of the target robot between the previous time step and the current time step.

[0067] Wherein, according to the historical encoding feature vector of the target robot in the previous time step adjacent to the current time step and the current encoding feature vector, the relative position change matrix between the previous time step and the current time step is obtained, which may include: the point cloud data of the previous time step and the current time step are point cloud data of two adjacent frames, and the iterative closest point (ICP) algorithm can be used to match the historical point cloud data of the previous time step with the current point cloud data of the current time step. The basic idea of ​​the ICP algorithm is to find the optimal transformation (for example, translation and rotation) between two point clouds through iteration, so that one point cloud overlaps with the other point cloud as much as possible under the transformation. Specifically, firstly, a first point cloud subset is selected in the historical frame point cloud of the previous time step (for example, a certain proportion of points are randomly selected), and then, the corresponding points closest to the first point cloud subset containing multiple first point cloud data are found in the current frame point cloud of the current time step (based on a certain distance metric, for example, Euclidean distance), and the second point cloud subset is obtained. There is a one-to-one correspondence between the multiple second point cloud data contained in the second point cloud subset and the multiple first point cloud data contained in the first point cloud subset. There are multiple pairs of matching point cloud pairs in the first point cloud subset and the second point cloud subset. Then, an initial transformation matrix can be calculated based on the multiple pairs of matching point cloud pairs, which can be understood as the approximate position of transforming one frame of point cloud to another frame of point cloud. Next, the transformation matrix can be continuously adjusted based on the difference between the two point clouds in the transformed matching point cloud pairs, and this process can be repeated until convergence, that is, the difference is no longer significantly reduced, and the relative position change matrix of the target robot between the previous time step and the current time step is obtained. The relative position change matrix obtained by the ICP algorithm represents the relative posture transformation of the point cloud data between adjacent frames, thereby obtaining the relative position change between the point cloud data, which provides an important basis for the posture estimation of the target robot.

[0068] According to the historical coding feature vector of the target robot in the previous time step adjacent to the current time step and the current coding feature vector, the temperature feature difference matrix between the previous time step and the current time step of the target robot is obtained, which can include: according to the historical coding feature vector and the current coding feature vector, obtaining the temperature change information of the current temperature feature matrix corresponding to the current time step and the historical temperature feature matrix corresponding to the previous time step; according to the temperature change information, obtaining the temperature feature difference matrix. Between adjacent frames, the difference value of the temperature feature is calculated by comparing the change amount of the temperature feature matrix. This can be achieved by calculating the difference of the temperature values ​​at the corresponding pixel positions to obtain a difference matrix representing the temperature change. For example, for the temperature matrix of two adjacent time steps and ,calculate The difference matrix is ​​obtained, in which each element represents the change in temperature at the corresponding position. The temperature feature difference value contains information about the change in ambient temperature during the movement of the target robot, and reflects the change in the position of the target robot from another perspective, because the target robot may perceive different temperature distributions at different positions, such as the temperature increases when it is close to the heating device and decreases when it is far away.

[0069] Step 108, obtaining the target current posture information of the target robot at the current time step according to the relative position change matrix and the temperature feature difference matrix.

[0070] In this step, the first step is to construct the initial position prediction function and initial prediction error covariance function of the target robot in the target space. Specifically, it is assumed that the target robot has a certain motion model from the previous time step to the current time step. This motion model can be established based on the kinematic principle or empirical data of the target robot. If the target robot is a wheeled inspection robot, its motion model can consider the relationship between factors such as the wheel speed and steering angle and the robot's position and posture changes. Based on this motion model, the initial position prediction function and initial prediction error covariance function of the target robot in the target space are constructed, which can be seen from equations (7) and (8):

[0071]

[0072] in, is the state transfer matrix, which describes the transformation relationship between the target robot's posture at different time steps; It is process noise, which is used to consider the uncertainty factors in the actual motion process, such as motion deviation caused by wheel slippage and uneven ground. The process noise follows Gaussian distribution; is the covariance matrix of process noise, which is used to describe the distribution characteristics of process noise; is the initial predicted pose; is the initial forecast error covariance.

[0073] The second step is to obtain the temperature observation residual of the temperature feature difference matrix for the initial pose prediction function according to the temperature feature difference matrix; and obtain the first optimized pose prediction function corresponding to the initial pose prediction function according to the temperature observation residual. Specifically, the main function of the temperature matrix data is to provide environmental feature positioning. According to the temperature feature difference matrix, the corresponding temperature observation value can be obtained. Then, according to the temperature observation value and the initial predicted pose, the temperature observation residual of the temperature feature difference matrix for the initial pose prediction function can be obtained, which can be seen in formula (9):

[0074]

[0075] in, It is the posture information provided by the temperature matrix (reflecting the posture change through the difference of environmental characteristics); is the observation matrix, which maps the state space (the robot's posture state) to the observation space (temperature feature space). The determination of the observation matrix depends on the specific relationship model between the temperature feature and the robot posture. For example, if the temperature feature is directly related to the position of the robot in a certain direction, then The corresponding elements in will reflect this relationship; is the observed residual for this temperature.

[0076] Next, based on the uncertainty of the temperature in predicting the position and posture of the target robot, the Kalman filter can be used to obtain the temperature Kalman gain, which can be seen in formula (10):

[0077]

[0078] in, is the covariance matrix of temperature observation noise, which is used to describe the error distribution characteristics of temperature observation data; is the temperature Kalman gain, It reflects the degree of trust the system has in the temperature matrix observation at this step. Furthermore, the temperature observation residual and the temperature Kalman gain can be used to update the initial pose prediction function and the initial prediction error covariance function, respectively, to obtain the first optimized pose prediction function and the first optimized prediction error covariance function, which can be seen in equations (11) and (12):

[0079]

[0080] The third step is to obtain the spatial observation residual of the relative position change matrix for the initial pose prediction function according to the relative position change matrix; and obtain the second optimized pose prediction function corresponding to the initial pose prediction function according to the spatial observation residual. Specifically, the spatial observation value can be obtained according to the relative position change matrix, and then the spatial observation residual can be obtained according to the spatial observation value, as shown in formula (13):

[0081]

[0082] in, is the residual of the spatial observation; It is the relative pose information calculated by point cloud ICP, which is used as an external pose observation for mapping update. This step is similar to the calculation of temperature matrix observation residual. By comparing the pose calculated by point cloud ICP with the pose estimate after the current temperature matrix update, the difference between the two is obtained, which provides a basis for the subsequent fusion of point cloud observation information.

[0083] Next, based on the uncertainty of the point cloud data for the position prediction of the target robot, the Kalman filter can be used to obtain the spatial Kalman gain, which can be seen in equation (14):

[0084]

[0085] in, is the measurement noise of point cloud ICP, which is used to describe the error characteristics of point cloud ICP observation data; is the spatial Kalman gain, which reflects the degree of trust the system has in the point cloud ICP observation data under the current pose estimation. It is similar to the Kalman gain calculation during temperature matrix observation, but takes into account the influence of point cloud ICP measurement noise.

[0086] Furthermore, in a possible implementation, the spatial observation residual and the spatial Kalman gain can be used to update the initial pose prediction function and the initial prediction error covariance function, respectively, to obtain the second optimized pose prediction function and the second optimized prediction error covariance function, as shown in equations (15) and (16); further, the target pose prediction function can be obtained based on the first optimized pose prediction function and the second optimized pose prediction function, and the target prediction error covariance function can be obtained based on the first optimized prediction error covariance function and the second optimized prediction error covariance function, as shown in equations (17) and (18). In another possible implementation, the spatial observation residual and the spatial Kalman gain can be used to update the first optimized pose prediction function and the first optimized prediction error covariance function, respectively, to obtain the target pose prediction function and the target prediction error covariance function, as shown in equations (19) and (20).

[0087]

[0088] Finally, the target current pose information of the target robot at the current time step can be obtained according to the target pose prediction function and the target prediction error covariance function. The target current pose information can include the position information and pose information of the target robot, which can be expressed as formula (21):

[0089]

[0090] in, Represents the position coordinates, Indicates the direction (attitude) angle.

[0091] Furthermore, the current pose information of the target robot at the current time step can be used to perform map update and loop detection. Map update: Update the Simultaneous Localization and Mapping (SLAM) map based on the fused pose. If the temperature matrix and point cloud features indicate repeated positions, mark the loop position and adjust the global coordinates to reduce the cumulative error. Loop detection and map optimization: Use temperature matrix features to match loops. After detecting loops, adjust the map through back-end optimization to reduce drift.

[0092] In the method of this embodiment, first, the current temperature and point cloud fusion feature matrix of the target robot in the target space at the current time step can be obtained; next, the temperature and point cloud fusion feature matrix can be feature encoded to obtain the current encoding feature vector of the temperature and point cloud fusion feature matrix; further, based on the historical encoding feature vector of the target robot in the previous time step adjacent to the current time step and the current encoding feature vector, the relative position change matrix of the target robot between the previous time step and the current time step can be obtained, as well as the temperature feature difference matrix of the target robot between the previous time step and the current time step; finally, based on the relative position change matrix and the temperature feature difference matrix, the target current posture information of the target robot at the current time step can be obtained. In the method provided in the embodiment of the present application, temperature data and point cloud data can be combined to generate a temperature and point cloud fusion feature matrix containing temperature features and spatial features, which can improve the distinguishability of the target robot to the environment and the positioning accuracy, and the temperature and point cloud fusion feature matrix can be feature encoded to improve the accuracy of feature extraction, thereby improving the positioning accuracy. In addition, the iterative closest point algorithm and Kalman filter can be used to predict the position and posture of the target robot, thereby improving the accuracy of the position and posture prediction, thereby further improving the positioning accuracy, the efficiency and accuracy of path planning, and the accuracy of map updating in the corresponding space.

[0093] In an exemplary embodiment, Figure 3 As shown, step 104 may include steps 302 to 308. Among them:

[0094] Step 302: construct an initial state space model for the target space.

[0095] Step 304: Acquire feature association parameters of the initial state space model based on the current temperature and the point cloud fusion feature matrix.

[0096] Among them, the Mamba network can express the parameters of the initial state space model as a function of the input features, that is, a function of the current temperature and the point cloud fusion feature matrix, and obtain the feature association parameters, which can be seen in formula (5).

[0097] Step 306: Obtain a target state space model for the target space according to the feature association parameters and the initial state space model.

[0098] According to the feature association parameters and the initial state space model, the target state space model for the target space is obtained, and the initial state space model is updated using the feature association parameters to obtain the target state space model, as shown in equation (6). and is an adaptive parameter calculated based on the current input. At each time step, the Mamba network dynamically adjusts the current temperature of the current input and the point cloud fusion feature matrix and The value of , thereby updating the state This process enables the Mamba network to track changes in environmental features. For example, when the robot moves to different areas of the data center and the temperature and spatial structure change, the network accurately reflects these changes by adjusting parameters and updating states, thereby effectively modeling the current temperature and point cloud fusion feature matrix.

[0099] Step 308 , using the target state space model, feature encoding is performed on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix.

[0100] Among them, the Mamba network uses the target state space model to perform feature encoding on the temperature and point cloud fusion feature matrix to obtain the current encoding feature vector of the temperature and point cloud fusion feature matrix. The current encoding feature vector is an abstract representation of the temperature and spatial features of the current frame. It integrates temperature and spatial information and can be compared with the encoding vector of the historical frame in subsequent processing.

[0101] In the method of this embodiment, a Mamba network can be introduced to perform feature encoding on the temperature and point cloud fusion feature matrix to improve the accuracy of feature extraction.

[0102] In an exemplary embodiment, Figure 4 As shown, obtaining the relative position change matrix of the target robot between the previous time step and the current time step in step 106 may include steps 402 to 406. Among them:

[0103] Step 402 : determining a first point cloud subset in the historical point cloud data set corresponding to the previous time step.

[0104] Step 404 , based on the historical encoding feature vector and the current encoding feature vector, a second point cloud subset matching the first point cloud subset is screened out from the current point cloud data set corresponding to the current time step.

[0105] There is a corresponding relationship between the multiple second point cloud data included in the second point cloud subset and the multiple first point cloud data included in the first point cloud subset.

[0106] Step 406: Obtain a relative position change matrix based on the difference information between the second point cloud subset and the first point cloud subset.

[0107] Among them, the point cloud data of the previous time step and the current time step are point cloud data of two adjacent frames. The iterative closest point (ICP) algorithm can be used to match the historical point cloud data of the previous time step with the current point cloud data of the current time step. The basic idea of ​​the ICP algorithm is to find the optimal transformation (for example, translation and rotation) between two point clouds through iteration, so that one point cloud overlaps with the other point cloud as much as possible under the transformation. Specifically, firstly, a first point cloud subset is selected from the historical frame point cloud of the previous time step (for example, a certain proportion of points are randomly selected), and then, the corresponding points closest to the first point cloud subset containing multiple first point cloud data are found in the current frame point cloud of the current time step (based on a certain distance metric, for example, Euclidean distance), and the second point cloud subset is obtained. There is a one-to-one correspondence between the multiple second point cloud data contained in the second point cloud subset and the multiple first point cloud data contained in the first point cloud subset. There are multiple pairs of matching point cloud pairs in the first point cloud subset and the second point cloud subset. Then, an initial transformation matrix can be calculated based on the multiple pairs of matching point cloud pairs, which can be understood as the approximate position of transforming one frame of point cloud to another frame of point cloud. Next, the transformation matrix can be continuously adjusted based on the difference between the two point clouds in the transformed matching point cloud pairs, and this process can be repeated until convergence, that is, the difference is no longer significantly reduced, and the relative position change matrix of the target robot between the previous time step and the current time step is obtained. The relative position change matrix obtained by the ICP algorithm represents the relative posture transformation of the point cloud data between adjacent frames, thereby obtaining the relative position change between the point cloud data, which provides an important basis for the posture estimation of the target robot.

[0108] In an exemplary embodiment, Figure 5 As shown, obtaining the temperature feature difference matrix of the target robot between the previous time step and the current time step in step 106 may include steps 502 to 504. Among them:

[0109] Step 502: Obtain temperature change information of a current temperature feature matrix corresponding to a current time step and a historical temperature feature matrix corresponding to a previous time step according to the historical coding feature vector and the current coding feature vector.

[0110] Step 504: Obtain a temperature feature difference matrix based on the temperature change information.

[0111] Among them, according to the historical coding feature vector and the current coding feature vector, the temperature change information of the current temperature feature matrix corresponding to the current time step and the historical temperature feature matrix corresponding to the previous time step is obtained; according to the temperature change information, the temperature feature difference matrix is ​​obtained. Between adjacent frames, the difference value of the temperature feature is calculated by comparing the change amount of the temperature feature matrix. This can be achieved by calculating the difference of the temperature values ​​at the corresponding pixel positions to obtain a difference matrix representing the temperature change. For example, for the temperature matrices of two adjacent time steps and ,calculate The difference matrix is ​​obtained, in which each element represents the change in temperature at the corresponding position. This temperature feature difference value contains information about the change in ambient temperature during the movement of the target robot, and reflects the change in the position of the target robot from another perspective, because the target robot may perceive different temperature distributions at different positions, such as the temperature increases when it is close to the heat-generating device and decreases when it is far away.

[0112] In the method of this embodiment, the relative position change matrix obtained by the ICP algorithm represents the relative pose transformation of the point cloud data between adjacent frames, thereby obtaining the relative position change between the point cloud data, which provides an important basis for the pose estimation of the target robot. The temperature feature difference value can also be calculated to represent the information of the ambient temperature change of the target robot during the movement, thereby improving the accuracy of the pose prediction.

[0113] In an exemplary embodiment, Figure 6 As shown, step 108 may include steps 602 to 606:

[0114] Step 602, constructing an initial pose prediction function and an initial prediction error covariance function of the target robot in the target space.

[0115] It is assumed that the target robot has a certain motion model from the previous time step to the current time step. This motion model can be established based on the kinematic principle or empirical data of the target robot. If the target robot is a wheeled inspection robot, its motion model can consider the relationship between factors such as wheel speed and steering angle and the robot's position and posture changes. Based on this motion model, the initial position prediction function and initial prediction error covariance function of the target robot in the target space are constructed, which can be seen in equations (7) and (8), where, is the state transfer matrix, which describes the transformation relationship between the target robot's posture at different time steps; It is process noise, which is used to consider the uncertainty factors in the actual motion process, such as motion deviation caused by wheel slippage and uneven ground. The process noise follows Gaussian distribution; is the covariance matrix of process noise, which is used to describe the distribution characteristics of process noise; is the initial predicted pose; is the initial forecast error covariance.

[0116] Step 604, obtaining a target pose prediction function corresponding to the initial pose prediction function according to the relative position change matrix and the temperature feature difference matrix, and obtaining a target prediction error covariance function corresponding to the initial prediction error covariance function.

[0117] Step 606, using the target posture prediction function and the target prediction error covariance function, obtain the target current posture information of the target robot at the current time step.

[0118] In an exemplary embodiment, Figure 7 As shown, in step 604, obtaining the target pose prediction function corresponding to the initial pose prediction function according to the relative position change matrix and the temperature feature difference matrix may include steps 702 to 710. Among them:

[0119] Step 702: Obtain the temperature observation residual of the temperature feature difference matrix for the initial pose prediction function according to the temperature feature difference matrix.

[0120] In this step, the main function of the temperature matrix data is to provide environmental feature positioning. According to the temperature feature difference matrix, the corresponding temperature observation value can be obtained. Then, according to the temperature observation value and the initial predicted pose, the temperature observation residual of the temperature feature difference matrix for the initial pose prediction function can be obtained, which can be seen in formula (9), where: It is the posture information provided by the temperature matrix (reflecting the posture change through the difference of environmental characteristics); is the observation matrix, which maps the state space (the robot's posture state) to the observation space (temperature feature space). The determination of the observation matrix depends on the specific relationship model between the temperature feature and the robot posture. For example, if the temperature feature is directly related to the position of the robot in a certain direction, then The corresponding elements in will reflect this relationship; is the temperature observation residual. Furthermore, the first optimized pose prediction function corresponding to the initial pose prediction function can be obtained according to the temperature observation residual.

[0121] Step 704: Obtain a first optimized pose prediction function corresponding to the initial pose prediction function based on the temperature observation residual.

[0122] Step 706: Obtain the spatial observation residual of the relative position change matrix for the initial pose prediction function according to the relative position change matrix.

[0123] In this step, the spatial observation value can be obtained according to the relative position change matrix, and then the spatial observation residual can be obtained according to the spatial observation value, as shown in formula (13):

[0124]

[0125] in, is the residual of the spatial observation; It is the relative pose information calculated by point cloud ICP, which is used as an external pose observation for mapping update. This step is similar to the calculation of temperature matrix observation residual. By comparing the pose calculated by point cloud ICP with the pose estimate after the current temperature matrix update, the difference between the two is obtained, which provides a basis for the subsequent fusion of point cloud observation information.

[0126] Step 708: Obtain a second optimized pose prediction function corresponding to the initial pose prediction function based on the spatial observation residual.

[0127] Step 710: Obtain a target pose prediction function corresponding to the initial pose prediction function according to the first optimized pose prediction function and the second optimized pose prediction function.

[0128] In an exemplary embodiment, step 704 may include:

[0129] Obtain the temperature Kalman gain of the temperature feature difference matrix for the initial pose prediction function; obtain the first optimized pose prediction function corresponding to the initial pose prediction function based on the temperature observation residual and the temperature Kalman gain.

[0130] Specifically, based on the uncertainty of the temperature in predicting the position and posture of the target robot, the Kalman filter can be used to obtain the temperature Kalman gain, which can be seen in formula (10), where: is the covariance matrix of temperature observation noise, which is used to describe the error distribution characteristics of temperature observation data; is the temperature Kalman gain, It reflects the degree of trust the system has in the temperature matrix observation at this step. Furthermore, the temperature observation residual and the temperature Kalman gain can be used to update the initial pose prediction function to obtain the first optimized pose prediction function, which can be seen in equation (11).

[0131] In an exemplary embodiment, step 708 may include:

[0132] Obtain the spatial Kalman gain of the relative position change matrix for the initial pose prediction function; and obtain the second optimized pose prediction function corresponding to the initial pose prediction function based on the spatial observation residual and the spatial Kalman gain.

[0133] Specifically, based on the uncertainty of the point cloud data for the position prediction of the target robot, the Kalman filter can be used to obtain the spatial Kalman gain, which can be seen in formula (14), where: is the measurement noise of point cloud ICP, which is used to describe the error characteristics of point cloud ICP observation data; is the spatial Kalman gain, which reflects the degree of trust the system has in the point cloud ICP observation data under the current pose estimation. It is similar to the Kalman gain calculation during temperature matrix observation, but takes into account the influence of point cloud ICP measurement noise.

[0134] Furthermore, in one possible implementation, the spatial observation residual and the spatial Kalman gain can be used to update the initial pose prediction function to obtain a second optimized pose prediction function, as shown in equation (15); further, the target pose prediction function can be obtained based on the first optimized pose prediction function and the second optimized pose prediction function, as shown in equation (17). In another possible implementation, the spatial observation residual and the spatial Kalman gain can be used to update the first optimized pose prediction function to obtain a target pose prediction function, as shown in equation (19).

[0135] In an exemplary embodiment, obtaining a target prediction error covariance function corresponding to the initial prediction error covariance function in step 604 may include:

[0136] Obtain the temperature Kalman gain of the temperature feature difference matrix for the initial posture prediction function; obtain the first optimized prediction error covariance function corresponding to the initial prediction error covariance function based on the temperature Kalman gain; obtain the spatial Kalman gain of the relative position change matrix for the initial posture prediction function; obtain the second optimized prediction error covariance function corresponding to the initial prediction error covariance function based on the spatial Kalman gain; obtain the target prediction error covariance function corresponding to the initial prediction error covariance function based on the first optimized prediction error covariance function and the second optimized prediction error covariance function.

[0137] In the method of this embodiment, the iterative closest point algorithm and the Kalman filter can be used to predict the position and posture of the target robot, thereby improving the accuracy of the position and posture prediction, and further improving the positioning accuracy, the efficiency and accuracy of path planning, and the accuracy of map updates in the corresponding space.

[0138] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0139] Based on the same inventive concept, the embodiment of the present application also provides a robot positioning device for implementing the robot positioning method involved above. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above method, so the specific limitations in one or more robot positioning device embodiments provided below can refer to the limitations of the robot positioning method above, and will not be repeated here.

[0140] In one embodiment, Figure 8 As shown, a robot positioning device is provided, including: a fusion module 802, an encoding module 804, a first calculation module 806 and a second calculation module 808, wherein:

[0141] A fusion module 802 is used to obtain the current temperature of the target robot in the target space at the current time step and the point cloud fusion feature matrix;

[0142] The encoding module 804 is used to perform feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix;

[0143] A first calculation module 806 is used to obtain a relative position change matrix of the target robot between the previous time step and the current time step, and to obtain a temperature feature difference matrix of the target robot between the previous time step and the current time step, based on a historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector;

[0144] The second calculation module 808 is used to obtain the target current posture information of the target robot at the current time step according to the relative position change matrix and the temperature feature difference matrix.

[0145] In one embodiment, the encoding module 804 is also used to: construct an initial state space model for the target space; obtain feature association parameters of the initial state space model based on the current temperature and point cloud fusion feature matrix; obtain a target state space model for the target space based on the feature association parameters and the initial state space model; and use the target state space model to feature encode the temperature and point cloud fusion feature matrix to obtain the current encoded feature vector of the temperature and point cloud fusion feature matrix.

[0146] In one embodiment, the first computing module 806 is further used to: determine a first point cloud subset in the historical point cloud data set corresponding to the previous time step; screen out a second point cloud subset matching the first point cloud subset in the current point cloud data set corresponding to the current time step according to the historical coding feature vector and the current coding feature vector; there is a corresponding relationship between the multiple second point cloud data contained in the second point cloud subset and the multiple first point cloud data contained in the first point cloud subset; and obtain the relative position change matrix according to the difference information between the second point cloud subset and the first point cloud subset.

[0147] In one embodiment, the first calculation module 806 is further used to: obtain temperature change information of the current temperature feature matrix corresponding to the current time step and the historical temperature feature matrix corresponding to the previous time step based on the historical coding feature vector and the current coding feature vector; and obtain the temperature feature difference matrix based on the temperature change information.

[0148] In one embodiment, the second calculation module 808 is also used to: construct an initial pose prediction function and an initial prediction error covariance function of the target robot in the target space; obtain a target pose prediction function corresponding to the initial pose prediction function according to the relative position change matrix and the temperature feature difference matrix, and obtain a target prediction error covariance function corresponding to the initial prediction error covariance function; and use the target pose prediction function and the target prediction error covariance function to obtain the target current pose information of the target robot at the current time step.

[0149] In one embodiment, the second calculation module 808 is also used to: obtain the temperature observation residual of the temperature feature difference matrix for the initial pose prediction function according to the temperature feature difference matrix; obtain the first optimized pose prediction function corresponding to the initial pose prediction function according to the temperature observation residual; obtain the spatial observation residual of the relative position change matrix for the initial pose prediction function according to the relative position change matrix; obtain the second optimized pose prediction function corresponding to the initial pose prediction function according to the spatial observation residual; obtain the target pose prediction function corresponding to the initial pose prediction function according to the first optimized pose prediction function and the second optimized pose prediction function.

[0150] In one embodiment, the second calculation module 808 is also used to: obtain the temperature Kalman gain of the temperature feature difference matrix for the initial pose prediction function; and obtain the first optimized pose prediction function corresponding to the initial pose prediction function based on the temperature observation residual and the temperature Kalman gain.

[0151] In one embodiment, the second calculation module 808 is also used to: obtain the spatial Kalman gain of the relative position change matrix for the initial pose prediction function; and obtain the second optimized pose prediction function corresponding to the initial pose prediction function based on the spatial observation residual and the spatial Kalman gain.

[0152] In one embodiment, the second calculation module 808 is also used to: obtain the temperature Kalman gain of the temperature feature difference matrix for the initial posture prediction function; obtain the first optimized prediction error covariance function corresponding to the initial prediction error covariance function based on the temperature Kalman gain; obtain the spatial Kalman gain of the relative position change matrix for the initial posture prediction function; obtain the second optimized prediction error covariance function corresponding to the initial prediction error covariance function based on the spatial Kalman gain; obtain the target prediction error covariance function corresponding to the initial prediction error covariance function based on the first optimized prediction error covariance function and the second optimized prediction error covariance function.

[0153] Each module in the above robot positioning device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0154] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Fig. 9As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store robot positioning related data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a robot positioning method is implemented.

[0155] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0156] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0157] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0158] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0159] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0160] 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 the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0161] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are 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.

[0162] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A robot positioning method, characterized in that: The method comprises: Obtain the current temperature and point cloud fusion feature matrix of the target robot in the target space at the current time step; Performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix; According to the historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector, a relative position change matrix of the target robot between the previous time step and the current time step is obtained, and a temperature feature difference matrix of the target robot between the previous time step and the current time step is obtained; According to the relative position change matrix and the temperature characteristic difference matrix, the target current posture information of the target robot at the current time step is obtained.

2. The method according to claim 1, characterized in that The step of performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix includes: constructing an initial state space model for the target space; Acquire characteristic association parameters of the initial state space model according to the current temperature and the point cloud fusion feature matrix; Acquire a target state space model for the target space according to the feature association parameters and the initial state space model; The target state space model is used to perform feature encoding on the temperature and point cloud fusion feature matrix to obtain the current encoded feature vector of the temperature and point cloud fusion feature matrix.

3. The method according to claim 1, characterized in that The step of obtaining a relative position change matrix of the target robot between the previous time step and the current time step includes: Determining a first point cloud subset in the historical point cloud data set corresponding to the previous time step; According to the historical encoding feature vector and the current encoding feature vector, a second point cloud subset matching the first point cloud subset is screened out from the current point cloud data set corresponding to the current time step; there is a corresponding relationship between the plurality of second point cloud data included in the second point cloud subset and the plurality of first point cloud data included in the first point cloud subset; The relative position change matrix is ​​acquired according to difference information between the second point cloud subset and the first point cloud subset.

4. The method according to claim 1, characterized in that: The step of obtaining a temperature feature difference matrix of the target robot between the previous time step and the current time step includes: According to the historical coding feature vector and the current coding feature vector, obtaining temperature change information of a current temperature feature matrix corresponding to the current time step and a historical temperature feature matrix corresponding to the previous time step; The temperature characteristic difference matrix is ​​obtained according to the temperature change information.

5. The method according to any one of claims 1 to 4, characterized in that: The step of obtaining the target current posture information of the target robot at the current time step according to the relative position change matrix and the temperature characteristic difference matrix includes: Constructing an initial posture prediction function and an initial prediction error covariance function of the target robot in the target space; According to the relative position change matrix and the temperature feature difference matrix, obtaining a target pose prediction function corresponding to the initial pose prediction function, and obtaining a target prediction error covariance function corresponding to the initial prediction error covariance function; The target current pose information of the target robot at the current time step is obtained by using the target pose prediction function and the target prediction error covariance function.

6. The method according to claim 5, characterized in that The step of obtaining a target pose prediction function corresponding to the initial pose prediction function according to the relative position change matrix and the temperature feature difference matrix includes: According to the temperature feature difference matrix, obtaining the temperature observation residual of the temperature feature difference matrix for the initial posture prediction function; According to the temperature observation residual, obtaining a first optimized pose prediction function corresponding to the initial pose prediction function; According to the relative position change matrix, obtaining the spatial observation residual of the relative position change matrix for the initial posture prediction function; According to the spatial observation residual, obtaining a second optimized pose prediction function corresponding to the initial pose prediction function; According to the first optimized pose prediction function and the second optimized pose prediction function, the target pose prediction function corresponding to the initial pose prediction function is obtained.

7. The method according to claim 6, characterized in that The step of obtaining, according to the temperature observation residual, a first optimized pose prediction function corresponding to the initial pose prediction function comprises: Obtaining the temperature Kalman gain of the temperature feature difference matrix for the initial posture prediction function; The first optimized pose prediction function corresponding to the initial pose prediction function is obtained according to the temperature observation residual and the temperature Kalman gain.

8. The method according to claim 6, characterized in that The step of obtaining, according to the spatial observation residual, a second optimized pose prediction function corresponding to the initial pose prediction function comprises: Obtaining the spatial Kalman gain of the relative position change matrix for the initial posture prediction function; According to the spatial observation residual and the spatial Kalman gain, the second optimized pose prediction function corresponding to the initial pose prediction function is obtained.

9. The method according to claim 5, characterized in that The obtaining of a target prediction error covariance function corresponding to the initial prediction error covariance function includes: Obtaining the temperature Kalman gain of the temperature feature difference matrix for the initial posture prediction function; According to the temperature Kalman gain, obtaining a first optimized prediction error covariance function corresponding to the initial prediction error covariance function; Obtaining the spatial Kalman gain of the relative position change matrix for the initial posture prediction function; According to the spatial Kalman gain, obtaining a second optimized prediction error covariance function corresponding to the initial prediction error covariance function; According to the first optimized prediction error covariance function and the second optimized prediction error covariance function, the target prediction error covariance function corresponding to the initial prediction error covariance function is obtained.

10. A robot positioning device, characterized in that: The device comprises: A fusion module is used to obtain the current temperature of the target robot in the target space at the current time step and the point cloud fusion feature matrix; An encoding module, used for performing feature encoding on the temperature and point cloud fusion feature matrix to obtain a current encoding feature vector of the temperature and point cloud fusion feature matrix; A first calculation module is used to obtain a relative position change matrix of the target robot between the previous time step and the current time step, and to obtain a temperature feature difference matrix of the target robot between the previous time step and the current time step according to a historical coding feature vector of the target robot at a previous time step adjacent to the current time step and the current coding feature vector; The second calculation module is used to obtain the target current posture information of the target robot at the current time step according to the relative position change matrix and the temperature feature difference matrix.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

13. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

Images