Mobile robot-based mapping method and system thereof

Abstract

The embodiment of the present application provides a mapping method based on a mobile robot and a system thereof, wherein the mapping method based on a mobile robot comprises: determining the score corresponding to the current particle posture of the mobile robot; judging whether the score is greater than a first threshold, if so, taking the laser radar posture of the mobile robot as the first posture, and reconstructing the map based on the first posture of the mobile robot; if not, judging whether it is greater than a second threshold, if the score is greater than the second threshold, correcting the posture of the mobile robot using the laser radar posture and the odometer posture to obtain a second posture, and reconstructing the map based on the second posture; if the score is not greater than the second threshold, recalculating the particle weight and updating it to a third posture, and the third posture is used to update the boundary information in the map. The embodiment of the present application can realize precise positioning and mapping for the robot.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and more particularly to a mobile robot-based mapping method and system thereof. Background Art

[0002] SLAM (Simultaneous Localization and Mapping) refers to the process of a robot equipped with certain sensors moving in an unfamiliar environment. Without prior information such as an environmental map, it uses only the sensor's observation of the environment to dynamically build a map model of the current environment in an incremental manner, and at the same time determines its own position in the map during the map construction process.

[0003] Currently, accurate map construction in unknown environments depends on accurate positioning information, and accurate positioning depends on accurate mapping. The robot's autonomous navigation and path planning rely on accurate positioning and environmental maps. Based on this, how to achieve accurate positioning and mapping for robots has become an urgent problem that needs to be solved. Summary of the Invention

[0004] In view of this, the embodiments of the present application provide a mapping method, apparatus, and related equipment based on a mobile robot, which can achieve precise positioning and mapping for the robot.

[0005] To achieve the above objectives, the present invention provides the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a mobile robot-based patterning method, comprising:

[0007] Determine the score corresponding to the current particle pose of the mobile robot;

[0008] Determine whether the score is greater than a first threshold. If so, take the lidar pose of the mobile robot as the first pose, and reconstruct the map based on the first pose of the mobile robot; if not, determine whether it is greater than a second threshold. If the score is greater than the second threshold, use the lidar pose and the odometer pose to correct the pose of the mobile robot to obtain a second pose, and reconstruct the map based on the second pose; if the score is not greater than the second threshold, recalculate the particle weight and update it to a third pose. The third pose is used to update the boundary information in the map, wherein the first threshold is greater than the second threshold, the lidar pose is the particle pose of the mobile robot under the lidar measurement mode, and the odometer pose is the particle pose of the mobile robot under the odometer measurement mode.

[0009] Optionally, the method of correcting the posture of the mobile robot using the lidar posture and the odometer posture to obtain the second posture includes:

[0010] Determine the first Sigma point corresponding to the mobile robot based on the lidar pose and odometer pose;

[0011] On the basis of the first Sigma point, the noise factor is superimposed to predict the second Sigma point, where the noise factor is the noise prediction under the actual state and the ideal state;

[0012] Calculate the Kalman gain of the unscented Kalman filter;

[0013] Observe the second Sigma point, combine it with the Kalman gain, and correct the second Sigma to update the second pose.

[0014] Optionally, determining the first Sigma point corresponding to the mobile robot based on the lidar pose and the odometer pose includes:

[0015] Use SIR filter to sample the lidar pose and odometer pose to obtain the first initial Sigma point;

[0016] Assign a weight to each particle in the first initial Sigma point;

[0017] Based on the weighted particles, resampling from the proposed distribution is used to obtain the first Sigma point.

[0018] Optionally, the step of obtaining the first Sigma point by resampling from the proposed distribution based on the weighted particles includes:

[0019] Determining whether the number of valid particles among the weighted particles is greater than a preset value;

[0020] If so, resampling from the proposed distribution is used to obtain the first Sigma point.

[0021] Optionally, determining the score corresponding to the current particle pose of the mobile robot includes:

[0022] Get the current position of the mobile robot;

[0023] Based on the comparison between the current posture and the preset posture, the parameters corresponding to the current posture are obtained, wherein the more similar the current posture is to the preset posture, the higher the score corresponding to the current posture.

[0024] Optionally, any particle pose includes a state vector and a weight, wherein the state vector is used to represent the heading of the mobile robot, and the weight is used to represent the importance of the current position of the mobile robot.

[0025] In a second aspect, an embodiment of the present invention provides a mapping system based on a mobile robot, comprising:

[0026] The score calculation module is used to determine the score corresponding to the current particle posture of the mobile robot;

[0027] A first judging module, configured to judge whether the score is greater than a first threshold;

[0028] If yes, executing a first reconstruction module, wherein the first reconstruction module is configured to use the lidar pose of the mobile robot as the first pose and reconstruct the map based on the first pose of the mobile robot;

[0029] If not, the second judgment module is executed to determine whether it is greater than the second threshold. If the score is greater than the second threshold, the second reconstruction module is executed to use the lidar pose and the odometer pose to correct the pose of the mobile robot to obtain a second pose, and reconstruct the map based on the second pose; if the score is not greater than the second threshold, the third reconstruction module is executed to recalculate the particle weight and update it to the third pose. The third pose is used to update the boundary information in the map, wherein the first threshold is greater than the second threshold, the lidar pose is the particle pose of the mobile robot under the lidar measurement mode, and the odometer pose is the particle pose of the mobile robot under the odometer measurement mode.

[0030] Optionally, the second reconstruction module is used to correct the posture of the mobile robot using the lidar posture and the odometer posture to obtain a second posture, including:

[0031] A sampling module is used to determine the first Sigma point corresponding to the mobile robot based on the lidar pose and the odometry pose;

[0032] A prediction module is used to superimpose a noise factor on the first Sigma point to predict a second Sigma point, where the noise factor is a noise prediction under an actual state and an ideal state;

[0033] A calculation module, used for calculating the Kalman gain of the unscented Kalman filter;

[0034] The observation module is used to observe the second Sigma point and correct the second Sigma in combination with the Kalman gain to update the second pose.

[0035] Optionally, the sampling module is used to determine the first Sigma point corresponding to the mobile robot based on the lidar pose and the odometer pose, including:

[0036] SIR filter, used to sample and process the lidar pose and odometer pose to obtain the first initial Sigma point;

[0037] A weight assignment module, for assigning a weight to each particle in the first initial Sigma point;

[0038] The resampling module is used to obtain the first Sigma point by resampling from the proposed distribution based on the particles after assigning weights.

[0039] Optionally, the SIR filter is used to obtain the first Sigma point by resampling from the proposed distribution based on the weighted particles, including:

[0040] A preset value judgment module is used to judge whether the number of effective particles in the weighted particles is greater than a preset value;

[0041] If so, resampling from the proposed distribution is used to obtain the first Sigma point.

[0042] An embodiment of the present invention provides a mapping method based on a mobile robot, comprising: determining a score corresponding to the current particle pose of the mobile robot; determining whether the score is greater than a first threshold; if so, using the laser radar pose of the mobile robot as a first pose and reconstructing a map based on the first pose of the mobile robot; if not, determining whether the score is greater than a second threshold; if the score is greater than the second threshold, correcting the pose of the mobile robot using the laser radar pose and the odometry pose to obtain a second pose, and reconstructing the map based on the second pose; if the score is not greater than the second threshold, recalculating the particle weight and updating it to a third pose, the third pose being used to update boundary information in the map, wherein the first threshold is greater than the second threshold, the laser radar pose is the particle pose of the mobile robot in the laser radar measurement mode, and the odometry pose is the particle pose of the mobile robot in the odometry measurement mode. Because the embodiment of the present invention uses the laser radar pose and the odometry pose to correct the pose of the mobile robot for poses greater than the second threshold and not greater than the first threshold, the robot's pose is corrected, which can better fit the actual situation, improve the accurate positioning of the mobile robot's pose, and further improve the efficiency of map reconstruction. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0044] Figure 1 A schematic diagram of the steps of a mobile robot-based mapping method provided in an embodiment of the present invention.

[0045] Figure 2 This is a comparison diagram of the composition method in this application and the existing composition method in an embodiment of the present invention.

[0046] Figure 3 A schematic diagram of the process of the composition process provided in an embodiment of the present application.

[0047] Figure 4 A structural diagram of a mobile robot-based mapping method system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0049] The inventors discovered that during the mapping process, LiDAR motion distortion is primarily caused by the relative motion between the mobile robot and the target. When the mobile robot is stationary or there is no relative motion between the targets, LiDAR motion distortion does not occur. However, the mobile robot is constantly moving while emitting laser light, and the sensor assumes this motion does not correspond to the robot's motion. LiDAR motion distortion is related to both the robot's speed and the laser frequency. The lower the laser frame rate and the faster the speed, the more significant the LiDAR motion distortion.

[0050] In order to eliminate the motion distortion produced by the laser radar and realize the accurate positioning of the posture change of the mobile robot, the present invention proposes a composition method based on the mobile robot.

[0051] like Figure 1 , which is a schematic diagram of the steps of the mobile robot-based mapping method provided in an embodiment of the present invention.

[0052] Step S11, determining the score corresponding to the current particle posture of the mobile robot;

[0053] In an optional implementation, the current particle posture of the mobile robot refers to the posture of the mobile robot during the movement. For example, the mobile robot performs a dive, roll, heading, etc., which will affect the spatial posture of the robot's movement.

[0054] Any current particle posture of the mobile robot includes a state vector and a weight, wherein the state vector is used to represent the heading of the mobile robot, and the weight is used to represent the importance of the current position of the mobile robot.

[0055] Step S12: Determine whether the score is greater than a first threshold.

[0056] The first threshold is a set target value, that is, when the score exceeds the target value, it is determined that the first threshold is exceeded.

[0057] If yes, then execute step S13, take the LiDAR pose of the mobile robot as the first pose, and reconstruct the map based on the first pose of the mobile robot;

[0058] If not, step S14: determine whether it is greater than a second threshold;

[0059] If so, it means that the score is greater than the second threshold, and step S15 is executed to correct the posture of the mobile robot using the lidar posture and the odometer posture to obtain a second posture, and reconstruct the map based on the second posture.

[0060] In an optional implementation, step S15, correcting the posture of the mobile robot using the lidar posture and the odometer posture to obtain the second posture, may specifically include:

[0061] Determine the first Sigma point corresponding to the mobile robot based on the lidar pose and odometer pose;

[0062] In an optional embodiment, a SIR filter may be used to sample and process the lidar pose and the odometer pose to obtain a first initial Sigma point.

[0063] Assign a weight to each particle in the first initial Sigma point;

[0064] Based on the weighted particles, resampling from the proposed distribution is used to obtain the first Sigma point.

[0065] In an optional embodiment, it is further possible to determine, based on the weighted particles, whether the number of valid particles in the weighted particles is greater than a preset value;

[0066] If so, resampling from the proposed distribution is used to obtain the first Sigma point.

[0067] On the basis of the first Sigma point, the noise factor is superimposed to predict the second Sigma point, where the noise factor is the noise prediction under the actual state and the ideal state;

[0068] Calculate the Kalman gain of the unscented Kalman filter;

[0069] Observe the second Sigma point, combine it with the Kalman gain, and correct the second Sigma to update the second pose.

[0070] It should be noted that the process of correcting the mobile robot's posture can be achieved by using an unscented Kalman filter. The unscented Kalman filter (UKF) is a combination of the unscented transform (UT transform) and the standard Kalman filter system. Through the lossless transformation, the nonlinear system equations are adapted to the standard Kalman system under the linear assumption.

[0071] Before estimating and predicting the system state, the state estimate needs to be subjected to a UT transform. The UT transform is commonly used in nonlinear systems. Its essence is that the principle of the UT transform is to select a set of points (Sigma points) around the mean and covariance matrix of the Gaussian distribution, calculate the expectation and covariance of the nonlinear function through these points, generate an estimate of the current state based on the covariance, and then use these statistics to approximate the output of the nonlinear function. The advantage of this method is that compared to the traditional Taylor series expansion method, the UT transform can better handle high-dimensional and non-Gaussian distributions and can reduce estimation errors.

[0072] In the UT transformation, x needs to be set as an n-dimensional random vector. The UT transformation first takes 2n+1 Sigma points and brings the Sigma points into the nonlinear equation to calculate χ i =(i=1, 2, ..., n), the Sigma point and its weight coefficient can be obtained according to the following formula:

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079] λ=α 2 (L+K)-n Formula Group 1

[0080] In formula group 1, α controls the sigma distribution, indicating that the degree of dispersion of the sigma points ranges from 0 to 1. In an optional embodiment, α is set to 0.001. K is a scaling factor. In an optional embodiment, K is set to 0 during calculation. β is the accuracy of the variance matrix. In this paper, β is set to 2. L is the number of sigma points. In this paper, L is set to 2*n+1, where n is 4. ----- is the weight coefficient of the first-order statistical characteristics, used to estimate the system state; ----- is the weight coefficient of the second-order statistical characteristics, used to estimate the covariance matrix;

[0081] The system nonlinear equation y = f(x) performs nonlinear transformation on each Sigma point, and corrects the second Sigma considering noise and Kalman gain to update the second pose.

[0082] If not, it means that the score is not greater than the second threshold, then step S16, recalculate the particle weight and update it to the third pose, the third pose is used to update the boundary information in the map, wherein the first threshold is greater than the second threshold, the lidar pose is the particle pose of the mobile robot under the lidar measurement mode, and the odometer pose is the particle pose of the mobile robot under the odometer measurement mode.

[0083] An embodiment of the present invention provides a mapping method based on a mobile robot, comprising: determining a score corresponding to the current particle pose of the mobile robot; determining whether the score is greater than a first threshold; if so, using the laser radar pose of the mobile robot as a first pose and reconstructing a map based on the first pose of the mobile robot; if not, determining whether the score is greater than a second threshold; if the score is greater than the second threshold, correcting the pose of the mobile robot using the laser radar pose and the odometry pose to obtain a second pose, and reconstructing the map based on the second pose; if the score is not greater than the second threshold, recalculating the particle weight and updating it to a third pose, the third pose being used to update boundary information in the map, wherein the first threshold is greater than the second threshold, the laser radar pose is the particle pose of the mobile robot in the laser radar measurement mode, and the odometry pose is the particle pose of the mobile robot in the odometry measurement mode. Because the embodiment of the present invention uses the laser radar pose and the odometry pose to correct the pose of the mobile robot for poses greater than the second threshold and not greater than the first threshold, the robot's pose is corrected, which can better fit the actual situation, improve the accurate positioning of the mobile robot's pose, and further improve the efficiency of map reconstruction.

[0084] Further integration Figure 2 As shown, the left figure is a map formed after the implementation of the existing composition method, and the right figure is a map formed by the implementation of the composition method of the present application. In the embodiment of the present invention, the fusion of the lidar and odometer data under the premise that the data is greater than the second threshold and not greater than the first threshold is more in line with the actual situation. For this reason, the map formed by the composition of the embodiment of the present invention is clearer, which is a significant improvement compared with the existing composition method.

[0085] In a further implementation, the present invention also uses an unscented Kalman filter to fuse lidar and odometry data. For highly nonlinear models like mobile robots, using an unscented Kalman filter to fuse lidar and odometry data is more realistic and effective. The fused data is then proportionally calculated with the laser scan matching data to produce more reliable pose data. Experiments have shown that this improved algorithm achieves better positioning and clearer mapping, significantly reducing positioning errors compared to existing mapping methods.

[0086] The mobile robot-based mapping method proposed in the embodiment of the present application provides a use case for indoor implementation. In order to better operate and detect the positioning accuracy of the algorithm during the mapping process, some obstacles need to be provided so that the effect can be reflected.

[0087] Before implementing the case, you need to configure relevant tools. This application embodiment is based on a mobile robot. First, configure a mobile robot and control it with Linux-ROS.

[0088] The ROS computation graph is a graphical structure composed of nodes and topics, primarily including messages, services, and the ROS master. Services allow a node to request certain operations from other nodes, while parameters allow nodes to dynamically configure parameters at runtime, thus affecting their behavior. Tools such as Rviz can be used to visualize the ROS computation graph structure.

[0089] Before implementation, you need to set the GMapping algorithm parameters. The threshold parameter, miniumScore, for determining whether scan matching is successful is set to 80, the parameter m_Min, for determining whether to perform unscented Kalman filter fusion, is set to 30, the number of particles is set to 30, and the parameter map_update_interval indicates the interval between map updates.

[0090] In an indoor venue, the mapping method based on mobile robots disclosed in the embodiment of the present application is used to build a map. Start the mapping node and enter the mapping command. In the embodiment of the present application, the mapping and navigation algorithms are both in the turn_on_wheeltec_robot function package. The mapping node is named mapping. Place the robot in a corner of the map, enter the command roslaunchturn_on_wheeltec_robotmapping.launch, and open RVIZ to view the mapping effect. Enter the command roslaunchwheeltec_robot_rc keyboard_teleop.launch, start the keyboard control node, and use the keyboard to control the robot movement. The linear speed and angular speed of the movement can be set according to the actual situation. When the robot has walked three-quarters around the outermost perimeter, the robot's position and the effect of the constructed map are as shown below. Figure 3 As shown in the figure, the mobile robot-based mapping method provided by the embodiment of the present invention, compared with traditional methods, can handle outliers and has a high fault tolerance. Furthermore, the improved GMapping algorithm creates maps more accurately and clearly defines map outlines.

[0091] As an optional implementation of the disclosed content of the embodiment of the present invention, the embodiment of the present invention provides a mapping system based on a mobile robot, which is applied to reconstruct the map corresponding to the mobile robot, such as Figure 4 As shown, Figure 4 A schematic structural diagram of a mobile robot-based mapping system according to one embodiment of the present invention includes:

[0092] The score calculation module 41 is used to determine the score corresponding to the current particle posture of the mobile robot;

[0093] A first judgment module 42 is used to judge whether the score is greater than a first threshold;

[0094] If yes, a first reconstruction module 43 is executed, wherein the first reconstruction module is configured to use the LiDAR pose of the mobile robot as the first pose and reconstruct the map based on the first pose of the mobile robot;

[0095] If not, the second judgment module 44 is executed to determine whether the score is greater than a second threshold. If the score is greater than the second threshold, the second reconstruction module 45 is executed to correct the posture of the mobile robot using the lidar posture and the odometer posture to obtain a second posture, and reconstruct the map based on the second posture;

[0096] If the score is not greater than the second threshold, the third reconstruction module 46 is executed to recalculate the particle weight and update it to a third pose, where the third pose is used to update the boundary information in the map, wherein the first threshold is greater than the second threshold, the lidar pose is the particle pose of the mobile robot under the lidar measurement mode, and the odometer pose is the particle pose of the mobile robot under the odometer measurement mode.

[0097] In the mobile robot-based mapping system provided in an embodiment of the present invention, the second reconstruction module is configured to correct the posture of the mobile robot using the lidar posture and the odometer posture to obtain a second posture, including:

[0098] A sampling module is used to determine the first Sigma point corresponding to the mobile robot based on the lidar pose and the odometry pose;

[0099] A prediction module is used to superimpose a noise factor on the first Sigma point to predict a second Sigma point, where the noise factor is a noise prediction under an actual state and an ideal state;

[0100] A calculation module, used for calculating the Kalman gain of the unscented Kalman filter;

[0101] The observation module is used to observe the second Sigma point and correct the second Sigma in combination with the Kalman gain to update the second pose.

[0102] In the mobile robot-based mapping system provided in an embodiment of the present invention, the sampling module is configured to determine the first Sigma point corresponding to the mobile robot based on the laser radar pose and the odometer pose, including:

[0103] SIR filter, used to sample and process the lidar pose and odometer pose to obtain the first initial Sigma point;

[0104] A weight assignment module, for assigning a weight to each particle in the first initial Sigma point;

[0105] The resampling module is used to obtain the first Sigma point by resampling from the proposed distribution based on the particles after assigning weights.

[0106] In the mobile robot-based mapping system provided by an embodiment of the present invention, the SIR filter is used to obtain the first Sigma point by resampling from the proposed distribution based on the weighted particles, including:

[0107] A preset value judgment module is used to judge whether the number of effective particles in the weighted particles is greater than a preset value;

[0108] If so, resampling from the proposed distribution is used to obtain the first Sigma point.

[0109] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0110] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined in the embodiments of this application may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not limited to the embodiments shown in the examples of this application, but is intended to conform to the widest scope consistent with the principles and novel features disclosed in the examples of this application.

Claims

1. A composition method based on a mobile robot, characterized in that: include: Determining a score corresponding to a current particle posture of the mobile robot includes: obtaining a current posture of the mobile robot; obtaining a parameter corresponding to the current posture based on a comparison between the current posture and a preset posture, wherein the more similar the current posture is to the preset posture, the higher the score corresponding to the current posture; Determine whether the score is greater than a first threshold value. If so, use the laser radar posture of the mobile robot as the first posture, and reconstruct the map based on the first posture of the mobile robot; if not, determine whether it is greater than a second threshold value. If the score is greater than the second threshold value, use the laser radar posture and the odometer posture to correct the posture of the mobile robot to obtain the second posture, and reconstruct the map based on the second posture; if the score is not greater than the second threshold value, recalculate the particle weight and update it to a third posture, and the third posture is used to update the boundary information in the map, wherein the first threshold value is greater than the second threshold value, the laser radar posture is the particle posture of the mobile robot under the laser radar measurement mode, and the odometer posture is the particle posture of the mobile robot under the odometer measurement mode; The method of correcting the posture of the mobile robot by using the laser radar posture and the odometer posture to obtain the second posture includes: Determine the first Sigma point corresponding to the mobile robot based on the LiDAR pose and the odometer pose; On the basis of the first Sigma point, a noise factor is superimposed to predict the second Sigma point, where the noise factor is a noise prediction under the actual state and the ideal state; Calculate the Kalman gain of the unscented Kalman filter; Observe the second Sigma point and correct the second Sigma in combination with the Kalman gain to update the second pose.

2. The mobile robot-based patterning method according to claim 1, characterized in that: The method of determining the first Sigma point corresponding to the mobile robot based on the laser radar posture and the odometer posture includes: The SIR filter is used to sample the lidar pose and odometer pose to obtain the first initial Sigma point; Assign a weight to each particle in the first initial Sigma point; based on the weighted particles, resample from the proposed distribution to obtain the first Sigma point.

3. The mobile robot-based patterning method according to claim 2, characterized in that: The particles after weight assignment are resampled from the proposed distribution to obtain the first Sigma point, including: It is determined whether the number of effective particles in the weighted particles is greater than a preset value; if so, resampling from the suggested distribution is adopted to obtain the first Sigma point.

4. The mobile robot-based patterning method according to claim 1, characterized in that: Any particle posture includes a state vector and a weight, wherein the state vector is used to represent the heading of the mobile robot, and the weight is used to represent the importance of the current position of the mobile robot.

5. A mapping system based on a mobile robot, characterized in that: include: The score calculation module is used to determine the score corresponding to the current particle posture of the mobile robot, including: obtaining the current posture of the mobile robot; based on the comparison between the current posture and the preset posture, obtaining the parameters corresponding to the current posture, wherein the more similar the current posture is to the preset posture, the higher the score corresponding to the current posture; a first judgment module is used to judge whether the score is greater than a first threshold; If yes, a first reconstruction module is executed, wherein the first reconstruction module is used to use the laser radar pose of the mobile robot as the first pose and reconstruct the map based on the first pose of the mobile robot; If not, the second judgment module is executed to determine whether it is greater than the second threshold. If the score is greater than the second threshold, the second reconstruction module is executed to use the laser radar posture and the odometer posture to correct the posture of the mobile robot to obtain the second posture, and reconstruct the map based on the second posture; if the score is not greater than the second threshold, the third reconstruction module is executed to recalculate the particle weight and update it to the third posture, and the third posture is used to update the boundary information in the map, wherein the first threshold is greater than the second threshold, the laser radar posture is the particle posture of the mobile robot under the laser radar measurement mode, and the odometer posture is the particle posture of the mobile robot under the odometer measurement mode; The second reconstruction module is used to correct the posture of the mobile robot using the laser radar posture and the odometer posture to obtain a second posture, including: A sampling module, used to determine the first Sigma point corresponding to the mobile robot based on the laser radar posture and the odometer posture; A prediction module is used to superimpose a noise factor on the first Sigma point to predict a second Sigma point, where the noise factor is a noise prediction under an actual state and an ideal state; A calculation module, used for calculating the Kalman gain of the unscented Kalman filter; The observation module is used to observe the second Sigma point and correct the second Sigma in combination with the Kalman gain to update the second pose.

6. The mobile robot-based mapping system according to claim 5, characterized in that: The sampling module is used to determine the first Sigma point corresponding to the mobile robot based on the laser radar posture and the odometer posture, including: SIR filter, used to sample and process the LiDAR pose and odometer pose to obtain the first initial Sigma point; A weight assignment module, used for assigning a weight to each particle in the first initial Sigma point; The resampling module is used to obtain the first Sigma point by resampling from the proposed distribution based on the particles after assigning weights.

7. The mobile robot-based mapping system according to claim 6, characterized in that: The SIR filter is used to obtain the first Sigma point based on the weighted particles by resampling from the proposed distribution, including: A preset value judgment module is used to judge whether the number of effective particles in the weighted particles is greater than a preset value; If so, resampling from the proposed distribution is adopted to obtain the first Sigma point.

Images