Mobile robot navigation and positioning system and method based on electronic compass

Abstract

The present application relates to the field of intelligent navigation technology, and specifically discloses a mobile robot navigation and positioning system and method based on an electronic compass, which uses artificial intelligence technology based on deep learning to analyze and process the robot heading angle data, acceleration value and angular velocity value collected by the electronic compass, accelerometer and gyroscope, and mines the robot's direction and posture information, and uses a laser scanner for real-time environmental perception and map construction, thereby combining the surrounding environment information with the robot's direction and posture information to plan the robot's navigation path. In this way, the navigation strategy can be adjusted in real time to improve the robot's navigation and positioning accuracy and adaptability in complex environments.

Description

Technical Field

[0001] The present application relates to the field of intelligent navigation technology, and more specifically, to a mobile robot navigation and positioning system and method based on an electronic compass. Background Art

[0002] With the continuous development of robot technology, mobile robots are increasingly used in various fields. Especially in complex environments, such as warehouses and factories, the navigation and positioning capabilities of mobile robots are particularly important.

[0003] As a sensor that uses the geomagnetic field for orientation, the electronic compass has the advantage of not being affected by factors such as light and occlusion, so it has broad application prospects in the field of navigation and positioning. However, using an electronic compass alone for navigation and positioning may be affected by problems such as geomagnetic field interference and error accumulation. Therefore, it is necessary to combine other sensor information to achieve more accurate and stable navigation and positioning. In addition, traditional mobile robot navigation and positioning methods mostly rely on pre-set maps or environmental models. When the environment changes, the robot may not be able to work effectively because it cannot update its environmental perception and map information in real time.

[0004] Therefore, an optimized mobile robot navigation and positioning system and method based on an electronic compass is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a mobile robot navigation and positioning system and method based on an electronic compass, which uses artificial intelligence technology based on deep learning to analyze and process the robot's heading angle data, acceleration values ​​and angular velocity values ​​collected by the electronic compass, accelerometer and gyroscope, thereby mining the robot's direction and posture information, and using a laser scanner for real-time environmental perception and map construction, thereby combining the surrounding environment information with the robot's direction and posture information to plan the robot's navigation path. In this way, the navigation strategy can be adjusted in real time to improve the robot's navigation and positioning accuracy and adaptability in complex environments.

[0006] Accordingly, according to one aspect of the present application, a mobile robot navigation and positioning system based on an electronic compass is provided, comprising:

[0007] A mobile robot state perception module, used to obtain point cloud data of the surrounding environment collected by a laser scanner deployed on the mobile robot, and heading angle data, acceleration value and angular velocity value of the mobile robot collected by an electronic compass, accelerometer and gyroscope deployed on the mobile robot;

[0008] A surrounding environment three-dimensional modeling module, used for performing three-dimensional modeling on the surrounding environment based on the point cloud data of the surrounding environment to obtain a surrounding environment three-dimensional model;

[0009] A target position embedding module, used for performing three-dimensional spatial feature extraction after marking the target position information in the surrounding environment three-dimensional model to obtain a feature vector of the target position embedded in the surrounding environment three-dimensional model;

[0010] A robot posture feature extraction module, used to extract a current posture feature vector of the robot based on the heading angle data, the acceleration value and the angular velocity value;

[0011] The path planning module is used to generate control instructions based on the interactive fusion features of the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the current posture of the robot.

[0012] In the above-mentioned mobile robot navigation and positioning system based on electronic compass, the target position embedding module includes: a target position marking unit, which is used to mark the target position information in the surrounding environment three-dimensional model to obtain a surrounding environment three-dimensional model containing the target position; a three-dimensional spatial feature extraction unit, which is used to pass the surrounding environment three-dimensional model containing the target position through a three-dimensional spatial feature extractor based on a three-dimensional convolutional neural network model to obtain a feature vector of the target position embedded in the surrounding environment three-dimensional model.

[0013] In the above-mentioned mobile robot navigation and positioning system based on electronic compass, the three-dimensional spatial feature extractor based on the three-dimensional convolutional neural network model includes an input layer, a three-dimensional convolution layer, an activation layer based on ReLU activation function, a pooling layer, a batch normalization layer and an output layer.

[0014] In the above-mentioned mobile robot navigation and positioning system based on electronic compass, the robot posture feature extraction module is used to: arrange the heading angle data, the acceleration value and the angular velocity value as a robot posture parameter input vector and then pass it through a parameter encoder based on a fully connected layer to obtain the current posture feature vector of the robot.

[0015] In the above-mentioned mobile robot navigation and positioning system based on electronic compass, the path planning module includes: an association interaction fusion unit, which is used to associate and encode the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the robot's current posture to obtain a robot navigation and positioning characterization feature matrix; a feature distribution optimization unit, which is used to perform position scattering based on depth anchor features on the robot navigation and positioning characterization feature matrix to obtain an optimized robot navigation and positioning characterization feature matrix; a robot control unit, which is used to pass the optimized robot navigation and positioning characterization feature matrix through a classifier-based path planner to obtain the control instruction, and the control instruction is used to represent the movement direction of the robot at the current time point.

[0016] In the above-mentioned mobile robot navigation and positioning system based on electronic compass, the associated interactive fusion unit is used to: calculate the transposed vector of the feature vector of the target position embedded in the surrounding environment three-dimensional model and the vector product of the robot's current posture feature vector to obtain the robot navigation and positioning characterization feature matrix.

[0017] In the above-mentioned mobile robot navigation and positioning system based on electronic compass, the feature distribution optimization unit includes: a feature flattening subunit, which is used to perform feature flattening on the robot navigation and positioning characterization feature matrix to obtain a robot navigation and positioning characterization feature vector; a position scattering subunit, which is used to perform position scattering on the robot navigation and positioning characterization feature vector based on depth anchoring features to obtain an optimized robot navigation and positioning characterization feature vector; and a reverse feature aggregation subunit, which is used to perform reverse feature aggregation on the optimized robot navigation and positioning characterization feature vector in accordance with the feature flattening method to obtain the optimized robot navigation and positioning characterization feature matrix.

[0018] In the above-mentioned mobile robot navigation and positioning system based on electronic compass, the position scattering subunit is used to: determine the mean of the eigenvalue set of the robot navigation and positioning characterization feature vector; accumulate and sum the eigenvalues ​​of the robot navigation and positioning characterization feature vector to obtain the sum value of the robot navigation and positioning characterization feature vector; subtract the mean of the robot navigation and positioning characterization feature vector from the sum value of the robot navigation and positioning characterization feature vector, divide it by a predetermined hyperparameter, and then divide it by the length of the robot navigation and positioning characterization feature vector to obtain a first quotient value; calculate the exponential function value with a natural constant as the base for the first quotient value to obtain a first natural exponential function value; multiply the first natural exponential function value by the robot navigation and positioning characterization feature vector according to the position point to obtain the optimized robot navigation and positioning characterization feature vector.

[0019] According to another aspect of the present application, a mobile robot navigation and positioning method based on an electronic compass is provided, comprising:

[0020] Acquire point cloud data of the surrounding environment collected by a laser scanner deployed on the mobile robot, and heading angle data, acceleration value and angular velocity value of the mobile robot collected by an electronic compass, accelerometer and gyroscope deployed on the mobile robot;

[0021] Based on the point cloud data of the surrounding environment, three-dimensional modeling is performed on the surrounding environment to obtain a three-dimensional model of the surrounding environment;

[0022] After marking the target position information in the surrounding environment three-dimensional model, three-dimensional space feature extraction is performed to obtain a feature vector of the target position embedded in the surrounding environment three-dimensional model;

[0023] Extracting a current posture feature vector of the robot based on the heading angle data, the acceleration value and the angular velocity value;

[0024] Based on the interactive fusion features of the target position embedded in the surrounding environment three-dimensional model feature vector and the robot's current posture feature vector, a control instruction is generated.

[0025] Compared with the prior art, the mobile robot navigation and positioning system and method based on electronic compass provided by the present application uses artificial intelligence technology based on deep learning to analyze and process the robot heading angle data, acceleration value and angular velocity value collected by the electronic compass, accelerometer and gyroscope, and mines the robot's direction and posture information, and uses a laser scanner for real-time environmental perception and map construction, thereby combining the surrounding environment information with the robot's direction and posture information to plan the robot's navigation path. In this way, the navigation strategy can be adjusted in real time to improve the robot's navigation and positioning accuracy and adaptability in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0027] Figure 1 4 is a block diagram of a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application.

[0028] Figure 2 Schematic diagram of the architecture of a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application.

[0029] Figure 34 is a block diagram of a target position embedding module in a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application.

[0030] Figure 4 4 is a block diagram of a path planning module in a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application.

[0031] Figure 5 It is a block diagram of a feature distribution optimization unit in a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application.

[0032] Figure 6 The present invention is a flowchart of a mobile robot navigation and positioning method based on an electronic compass according to an embodiment of the present application. DETAILED DESCRIPTION

[0033] Below, the embodiments of the present application will be described in more detail in conjunction with the accompanying drawings, and the above and other purposes, features and advantages of the present application will become more apparent. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

[0034] Figure 1 4 is a block diagram of a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, the mobile robot navigation and positioning system 100 based on the electronic compass includes: a mobile robot state perception module 110, which is used to obtain point cloud data of the surrounding environment collected by a laser scanner deployed on the mobile robot, and the heading angle data, acceleration value and angular velocity value of the mobile robot collected by the electronic compass, accelerometer and gyroscope deployed on the mobile robot; a surrounding environment three-dimensional modeling module 120, which is used to perform three-dimensional modeling of the surrounding environment based on the point cloud data of the surrounding environment to obtain a surrounding environment three-dimensional model; a target position embedding module 130, which is used to perform three-dimensional spatial feature extraction after marking the target position information in the surrounding environment three-dimensional model to obtain a target position embedded in the surrounding environment three-dimensional model feature vector; a robot posture feature extraction module 140, which is used to extract the robot's current posture feature vector based on the heading angle data, the acceleration value and the angular velocity value; a path planning module 150, which is used to generate a control instruction based on the interactive fusion feature of the target position embedded in the surrounding environment three-dimensional model feature vector and the robot's current posture feature vector.

[0035] As mentioned in the background technology above, traditional mobile robot navigation and positioning technologies mostly rely on pre-built maps or environmental models. However, when the environment changes, the robot may not be able to effectively perform its work tasks because it cannot update its environmental perception data and map information in real time. In response to the above technical problems, the technical concept of the present application is to use artificial intelligence technology based on deep learning to analyze and process the robot's heading angle data, acceleration values ​​and angular velocity values ​​collected by the electronic compass, accelerometer and gyroscope, thereby mining the robot's direction and posture information, and using a laser scanner for real-time environmental perception and map construction, thereby combining the surrounding environment information with the robot's direction and posture information to plan the robot's navigation path. In this way, the navigation strategy can be adjusted in real time to improve the robot's navigation and positioning accuracy and adaptability in complex environments.

[0036] In the above-mentioned mobile robot navigation and positioning system 100 based on electronic compass, the mobile robot state perception module 110 is used to obtain the point cloud data of the surrounding environment collected by the laser scanner deployed on the mobile robot, as well as the heading angle data, acceleration value and angular velocity value of the mobile robot collected by the electronic compass, accelerometer and gyroscope deployed on the mobile robot. It should be understood that the laser scanner can quickly scan the surrounding environment to form point cloud data, accurately capture the geometric information of the robot's surrounding environment, such as obstacle distribution, road width, terrain undulation, etc., provide a real-time, high-precision three-dimensional representation of the robot's environment, and can update the point cloud data in time when the environment changes, so that the robot can perceive obstacles, terrain changes, etc., and understand its current position. The electronic compass can provide the robot's absolute direction (heading angle), so that the robot can determine its orientation relative to the earth's magnetic field. The accelerometer can measure the acceleration of the robot in all directions, reflect the speed change of the robot, and perceive the robot's motion state. The gyroscope can measure the robot's angular velocity, that is, the rate and direction of the robot's rotational motion, so as to determine the robot's posture change. Therefore, through the collaborative work of laser scanners, electronic compasses, accelerometers and gyroscopes, comprehensive perception of the mobile robot can be provided, including its environment, motion status and posture information, providing key data support for subsequent navigation path planning.

[0037] In the above-mentioned mobile robot navigation and positioning system 100 based on electronic compass, the surrounding environment three-dimensional modeling module 120 is used to perform three-dimensional modeling of the surrounding environment based on the point cloud data of the surrounding environment to obtain a three-dimensional model of the surrounding environment. It should be understood that in order to convert the collected point cloud data of the surrounding environment into a practical and applicable navigation map, it is necessary to further process the point cloud data of the surrounding environment to construct a three-dimensional model. In a specific implementation, the point cloud data of the surrounding environment can be processed using point cloud processing algorithms, such as point-based cloud registration, surface reconstruction and other technologies, to construct a three-dimensional model of the surrounding environment. In this way, based on the three-dimensional model of the surrounding environment, it is possible to more intuitively and accurately reflect the geometric structure of the robot's environment, obstacle distribution and other information, and provide accurate spatial information reference for subsequent navigation path planning. In addition, the three-dimensional model of the surrounding environment can be dynamically updated according to the point cloud data collected by the laser scanner to adapt to changes in the environment.

[0038] In the above-mentioned mobile robot navigation and positioning system 100 based on electronic compass, the target position embedding module 130 is used to extract three-dimensional spatial features after marking the target position information in the surrounding environment three-dimensional model to obtain a feature vector of the target position embedded in the surrounding environment three-dimensional model. Figure 3 As shown, the target position embedding module 130 includes: a target position marking unit 131, used to mark the target position information in the surrounding environment three-dimensional model to obtain a surrounding environment three-dimensional model containing the target position; a three-dimensional space feature extraction unit 132, used to pass the surrounding environment three-dimensional model containing the target position through a three-dimensional space feature extractor based on a three-dimensional convolutional neural network model to obtain a feature vector of the target position embedded in the surrounding environment three-dimensional model.

[0039] Specifically, the target position marking unit 131 is used to mark the target position information in the surrounding environment three-dimensional model to obtain the surrounding environment three-dimensional model containing the target position. It should be understood that the target position information is a specific position that the robot is expected to reach input by the user. When marking the target position, the coordinate information in the surrounding environment three-dimensional model can be used to mark the target position at the corresponding position in the surrounding environment three-dimensional model, so that the robot can clearly know the specific position it needs to reach, so that the navigation target of the robot is determined according to the target position information during subsequent path planning, and the corresponding navigation path planning is performed.

[0040] Specifically, the three-dimensional space feature extraction unit 132 is used to pass the three-dimensional model of the surrounding environment containing the target position through a three-dimensional space feature extractor based on a three-dimensional convolutional neural network model to obtain a feature vector of the three-dimensional model of the target position embedded in the surrounding environment. A person of ordinary skill in the art should know that a three-dimensional convolutional neural network model is a neural network model suitable for processing three-dimensional data. Compared with a traditional two-dimensional convolutional neural network, a three-dimensional convolutional neural network can better process three-dimensional data with depth information and effectively extract rich spatial feature information in three-dimensional space. In the present application, the three-dimensional space feature extractor based on the three-dimensional convolutional neural network model performs a convolution operation on the three-dimensional model of the surrounding environment containing the target position in three dimensions (length, width and height), and gradually extracts higher-level feature representations through connections and information transfer between network layers, effectively captures the spatial features of the surrounding environment closely related to the target position, and comprehensively understands the target position and environmental information, thereby providing guidance for subsequent navigation path planning. In a specific example of the present application, the three-dimensional spatial feature extractor based on the three-dimensional convolutional neural network model includes an input layer, a three-dimensional convolution layer, an activation layer based on a ReLU activation function, a pooling layer, a batch normalization layer and an output layer.

[0041] In the above-mentioned mobile robot navigation and positioning system 100 based on electronic compass, the robot posture feature extraction module 140 is used to extract the robot's current posture feature vector based on the heading angle data, the acceleration value and the angular velocity value. In a specific example of the present application, the encoding method for extracting the robot's current posture feature vector based on the heading angle data, the acceleration value and the angular velocity value is to arrange the heading angle data, the acceleration value and the angular velocity value as a robot posture parameter input vector and then pass through a parameter encoder based on a fully connected layer to obtain the robot's current posture feature vector. It should be understood that the heading angle, the acceleration value and the angular velocity value are key parameters for describing the robot's current posture and motion state, and the three together reflect the robot's real-time state in the environment. By combining the heading angle data, the acceleration value and the angular velocity value into the form of a robot posture parameter input vector for full connection encoding, the correlation and information complementarity between the three can be fully utilized to more comprehensively describe the robot's posture and motion characteristics. Specifically, the fully connected layer (also called a dense layer or a fully connected network) has powerful feature extraction and feature integration capabilities. Here, the parameter encoder based on the fully connected layer introduces nonlinear feature extraction capability through a nonlinear activation function, performs high-level feature abstraction and integration on the robot posture parameter input vector to generate the robot's current posture feature vector, thereby effectively characterizing the robot's motion state in the current environment.

[0042] In the above-mentioned mobile robot navigation and positioning system 100 based on electronic compass, the path planning module 150 is used to generate control instructions based on the interactive fusion features of the target position embedded in the surrounding environment three-dimensional model feature vector and the robot's current posture feature vector. Figure 4 FIG. 1 is a block diagram of a path planning module in a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application. Figure 4 As shown, the path planning module 150 includes: an associated interactive fusion unit 151, which is used to associate and encode the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the robot's current posture to obtain a robot navigation and positioning characterization feature matrix; a feature distribution optimization unit 152, which is used to perform position scattering based on depth anchor features on the robot navigation and positioning characterization feature matrix to obtain an optimized robot navigation and positioning characterization feature matrix; a robot control unit 153, which is used to pass the optimized robot navigation and positioning characterization feature matrix through a classifier-based path planner to obtain the control instruction, and the control instruction is used to represent the movement direction of the robot at the current time point.

[0043] Specifically, the associated interactive fusion unit 151 is used to associate and encode the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the current posture of the robot to obtain the robot navigation and positioning characterization feature matrix. It should be understood that the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the current posture of the robot represent the target position information in the navigation task and the motion state information of the robot itself, respectively. By associating and encoding the two, the target position environment information and the real-time posture information of the robot can be integrated, which helps to more comprehensively understand the target position, the surrounding environment and the position and posture of the robot itself in the environment, so as to more accurately determine the relative relationship between the robot's own position and the target, and improve the accuracy and reliability of navigation. In a specific example of the present application, the encoding method for associating the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the current posture of the robot is to calculate the transposed vector of the feature vector of the target position embedded in the surrounding environment three-dimensional model and the vector product of the feature vector of the current posture of the robot to obtain the robot navigation and positioning characterization feature matrix. That is, the target position is embedded in the feature vector of the surrounding environment three-dimensional model and the feature vector of the robot's current posture by means of vector multiplication, and the target position information and the robot posture information are integrated into a unified feature representation, thereby establishing the association relationship between the target position, the surrounding environment and the robot's motion state in the navigation task, so as to enable the robot to make navigation and positioning decisions.

[0044] Specifically, the feature distribution optimization unit 152 is used to perform position scattering based on the depth anchor feature on the robot navigation and positioning characterization feature matrix to obtain an optimized robot navigation and positioning characterization feature matrix. In particular, in the technical solution of the present application, the robot navigation and positioning characterization feature matrix is ​​a key information carrier that combines the three-dimensional model features of the surrounding environment and the current posture of the robot, and is used to guide the path planner to generate control instructions. Therefore, improving the feature significance of the robot navigation and positioning characterization feature matrix and avoiding feature redundancy are of great significance for ensuring the accuracy and efficiency of robot navigation and positioning. Specifically, significant robot navigation and positioning characterization features help to more accurately identify the target position and the structure of the surrounding environment, thereby improving the accuracy of navigation. In addition, significant robot navigation and positioning characterization features help the robot to better understand its environment, including the location of obstacles and possible movement paths, provide richer guidance information, and help the path planner make more reasonable movement decisions. At the same time, removing redundant features can reduce the computational burden and storage requirements of the model, reduce the system's dependence on hardware resources, speed up the path planning speed, and improve the robot's reaction speed. In addition, redundant features may cause the model to overfit, affecting its performance in new or changing environments. Removing feature redundancy helps to improve the generalization ability of the model. That is to say, significant robot navigation and positioning characterization features help the robot quickly adjust its path when encountering unknown obstacles or environmental changes, identify potential collision risks earlier, and take evasive measures to reduce navigation errors caused by misidentification or insufficient information. Based on this, in the technical solution of the present application, the robot navigation and positioning characterization feature matrix is ​​further subjected to position scattering based on the depth anchor feature to obtain an optimized robot navigation and positioning characterization feature matrix.

[0045] Figure 5 FIG. 1 is a block diagram of a feature distribution optimization unit in a mobile robot navigation and positioning system based on an electronic compass according to an embodiment of the present application. Figure 5 As shown, the feature distribution optimization unit 152 includes: a feature flattening subunit 1521, used for performing feature flattening on the robot navigation and positioning characterization feature matrix to obtain a robot navigation and positioning characterization feature vector; a position scattering subunit 1522, used for performing position scattering on the robot navigation and positioning characterization feature vector based on a depth anchor feature to obtain an optimized robot navigation and positioning characterization feature vector; and a reverse feature aggregation subunit 1523, used for performing reverse feature aggregation on the optimized robot navigation and positioning characterization feature vector in accordance with the feature flattening method to obtain the optimized robot navigation and positioning characterization feature matrix.

[0046] More specifically, the position scattering subunit 1522 is used to: determine the mean of the eigenvalue set of the robot navigation and positioning characterization feature vector; accumulate and sum the eigenvalues ​​of the robot navigation and positioning characterization feature vector to obtain the sum of the robot navigation and positioning characterization feature vector; subtract the mean of the robot navigation and positioning characterization feature vector from the sum of the robot navigation and positioning characterization feature vector, divide it by a predetermined hyperparameter, and then divide it by the length of the robot navigation and positioning characterization feature vector to obtain a first quotient value; calculate the exponential function value with a natural constant as the base for the first quotient value to obtain a first natural exponential function value; multiply the first natural exponential function value by the robot navigation and positioning characterization feature vector by position point to obtain the optimized robot navigation and positioning characterization feature vector. That is, the robot navigation and positioning characterization feature vector is subjected to position scattering based on the depth anchor feature using the following optimization formula to obtain the optimized robot navigation and positioning characterization feature vector, wherein the optimization formula is: ;in, represents the robot navigation and positioning characterization feature vector, The first characterization vector representing the robot navigation and positioning The eigenvalues ​​at the positions, represents the mean value of the characteristic vector representing the robot navigation and positioning, represents the length of the robot navigation and positioning characteristic vector, represents the predetermined hyperparameters, represents the natural exponential function operation, It means point multiplication by position. Represents the characteristic vector representing the optimized robot navigation and positioning.

[0047] Specifically, in order to improve the feature significance of the robot navigation and positioning characterization feature matrix and avoid feature redundancy, in the technical solution of the present application, the robot navigation and positioning characterization feature matrix is ​​subjected to position scattering based on deep anchoring features, and the global mean of the robot navigation and positioning characterization feature matrix is ​​used as its dynamically formed deep anchoring feature. A coherence attention response mechanism based on the deep anchoring feature is applied to the original feature distribution to perform position scattering response depth correlation correction in a class divergence space with depth difference, so that the original feature distribution has a highly significant depth correlation distribution characteristic, so as to improve the feature significance of the robot navigation and positioning characterization feature matrix and avoid feature redundancy.

[0048] Specifically, the robot control unit 153 is used to pass the optimized robot navigation and positioning characterization feature matrix through a path planner based on a classifier to obtain the control instruction, and the control instruction is used to represent the moving direction of the robot at the current time point. It should be understood that the optimized robot navigation and positioning characterization feature matrix integrates the target position, the surrounding environment and the current posture information of the robot, and provides a comprehensive and accurate information basis for path planning. Furthermore, in order to convert the optimized robot navigation and positioning characterization feature matrix into a specific robot movement path guidance, in the technical solution of the present application, a classification algorithm is used to classify and decide on the optimized robot navigation and positioning characterization feature matrix. Specifically, the classifier-based path planner learns a large amount of training data so that it can identify the best path under different environmental conditions. Furthermore, when the trained path planner receives the optimized robot navigation and positioning characterization feature matrix, the path planner can make a classification decision based on its feature distribution and output the corresponding control instruction, that is, the moving direction of the robot at the current time point, to guide the robot to move according to the best path, thereby realizing efficient robot navigation control.

[0049] In summary, the mobile robot navigation and positioning system based on the electronic compass according to the embodiment of the present application is explained, which uses artificial intelligence technology based on deep learning to analyze and process the robot heading angle data, acceleration value and angular velocity value collected by the electronic compass, accelerometer and gyroscope, and mines the robot's direction and posture information, and uses a laser scanner for real-time environmental perception and map construction, so as to plan the robot's navigation path by combining the surrounding environment information with the robot's direction and posture information. In this way, the navigation strategy can be adjusted in real time to improve the navigation and positioning accuracy and adaptability of the robot in complex environments.

[0050] Figure 6 FIG. 1 is a flow chart of a mobile robot navigation and positioning method based on an electronic compass according to an embodiment of the present application. Figure 6As shown, according to the embodiment of the present application, the navigation and positioning method of a mobile robot based on an electronic compass includes the following steps: S110, acquiring point cloud data of the surrounding environment collected by a laser scanner deployed on the mobile robot, and heading angle data, acceleration value and angular velocity value of the mobile robot collected by an electronic compass, accelerometer and gyroscope deployed on the mobile robot; S120, based on the point cloud data of the surrounding environment, performing three-dimensional modeling of the surrounding environment to obtain a three-dimensional model of the surrounding environment; S130, after marking the target position information in the three-dimensional model of the surrounding environment, performing three-dimensional spatial feature extraction to obtain a feature vector of the target position embedded in the three-dimensional model of the surrounding environment; S140, extracting the feature vector of the robot's current posture based on the heading angle data, the acceleration value and the angular velocity value; S150, generating a control instruction based on the interactive fusion features of the feature vector of the target position embedded in the three-dimensional model of the surrounding environment and the feature vector of the robot's current posture.

[0051] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned mobile robot navigation and positioning method based on the electronic compass have been described in the above reference. Figures 1 to 5 The description of the mobile robot navigation and positioning system based on electronic compass has been introduced in detail, and therefore, its repeated description will be omitted.

[0052] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0053] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the module division is only a logical function division, and there may be other division methods in actual implementation. The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0054] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0055] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0056] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0057] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0058] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A mobile robot navigation and positioning system based on an electronic compass, characterized in that: include: A mobile robot state perception module, used to obtain point cloud data of the surrounding environment collected by a laser scanner deployed on the mobile robot, and heading angle data, acceleration value and angular velocity value of the mobile robot collected by an electronic compass, accelerometer and gyroscope deployed on the mobile robot; A surrounding environment three-dimensional modeling module, used for performing three-dimensional modeling on the surrounding environment based on the point cloud data of the surrounding environment to obtain a surrounding environment three-dimensional model; A target position embedding module, used for performing three-dimensional spatial feature extraction after marking the target position information in the surrounding environment three-dimensional model to obtain a feature vector of the target position embedded in the surrounding environment three-dimensional model; A robot posture feature extraction module, used to extract a current posture feature vector of the robot based on the heading angle data, the acceleration value and the angular velocity value; A path planning module, used to generate control instructions based on the interactive fusion features of the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the current posture of the robot; Wherein, the target position embedding module includes: A target position marking unit, used for marking target position information in the surrounding environment three-dimensional model to obtain a surrounding environment three-dimensional model including the target position; A three-dimensional space feature extraction unit, used for passing the three-dimensional model of the surrounding environment including the target position through a three-dimensional space feature extractor based on a three-dimensional convolutional neural network model to obtain a feature vector of the three-dimensional model of the target position embedded in the surrounding environment; The three-dimensional spatial feature extractor based on the three-dimensional convolutional neural network model includes an input layer, a three-dimensional convolution layer, an activation layer based on a ReLU activation function, a pooling layer, a batch normalization layer and an output layer; Wherein, the robot posture feature extraction module is used to: Arranging the heading angle data, the acceleration value and the angular velocity value as a robot posture parameter input vector and then passing through a parameter encoder based on a fully connected layer to obtain a current posture feature vector of the robot; Wherein, the path planning module includes: An associative interaction fusion unit, used for associatively encoding the feature vector of the target position embedded in the surrounding environment three-dimensional model and the feature vector of the robot's current posture to obtain a robot navigation positioning representation feature matrix; A feature distribution optimization unit, used for performing position scattering based on the depth anchor feature on the robot navigation and positioning characterization feature matrix to obtain an optimized robot navigation and positioning characterization feature matrix; A robot control unit is used to pass the optimized robot navigation and positioning characterization feature matrix through a classifier-based path planner to obtain the control instruction, and the control instruction is used to indicate the movement direction of the robot at a current time point.

2. The mobile robot navigation and positioning system based on electronic compass according to claim 1, characterized in that: The association interaction fusion unit is used to: The vector product of the transposed vector of the feature vector of the three-dimensional model of the target position embedded in the surrounding environment and the feature vector of the current posture of the robot is calculated to obtain the robot navigation and positioning characterization feature matrix.

3. The mobile robot navigation and positioning system based on electronic compass according to claim 2, characterized in that: The feature distribution optimization unit comprises: A feature flattening subunit, used for flattening the robot navigation and positioning characterization feature matrix to obtain a robot navigation and positioning characterization feature vector; A position scattering subunit, used for performing position scattering on the robot navigation and positioning characterization feature vector based on the depth anchor feature to obtain an optimized robot navigation and positioning characterization feature vector; The reverse feature aggregation subunit is used to perform reverse feature aggregation on the optimized robot navigation and positioning characterization feature vector in the feature flattening manner to obtain the optimized robot navigation and positioning characterization feature matrix.

4. The mobile robot navigation and positioning system based on electronic compass according to claim 3 is characterized in that: The position scattering subunit is used for: Determining the mean of a set of eigenvalues ​​of the robot navigation and positioning characterization eigenvector; Accumulating and summing each eigenvalue of the robot navigation and positioning characterization feature vector to obtain a total value of the robot navigation and positioning characterization feature vector; Subtracting the mean value of the robot navigation and positioning characterization feature vector from the sum of the robot navigation and positioning characterization feature vectors and dividing the sum by a predetermined hyperparameter and then by the length of the robot navigation and positioning characterization feature vector to obtain a first quotient value; Calculating an exponential function value with a natural constant as a base for the first quotient value to obtain a first natural exponential function value; The first natural exponential function value is multiplied by the robot navigation and positioning characterization feature vector at position points to obtain the optimized robot navigation and positioning characterization feature vector.

5. A mobile robot navigation and positioning method based on an electronic compass, using the mobile robot navigation and positioning system based on an electronic compass according to claim 1, characterized in that: include: Acquire point cloud data of the surrounding environment collected by a laser scanner deployed on the mobile robot, and heading angle data, acceleration value and angular velocity value of the mobile robot collected by an electronic compass, accelerometer and gyroscope deployed on the mobile robot; Based on the point cloud data of the surrounding environment, three-dimensional modeling is performed on the surrounding environment to obtain a three-dimensional model of the surrounding environment; After marking the target position information in the surrounding environment three-dimensional model, three-dimensional space feature extraction is performed to obtain a feature vector of the target position embedded in the surrounding environment three-dimensional model; Extracting a current posture feature vector of the robot based on the heading angle data, the acceleration value and the angular velocity value; Based on the interactive fusion features of the target position embedded in the surrounding environment three-dimensional model feature vector and the robot's current posture feature vector, a control instruction is generated.

Images