Spherical motor and position estimation method

Abstract

The invention discloses a spherical motor and a position estimation method, and relates to the technical field of motors, the spherical motor comprises a stator mechanism and a rotor mechanism, the stator mechanism comprises a stator body, an electromagnetic coil group and a Hall sensor, the electromagnetic coil group and the Hall sensor are arranged on the stator body, and the stator body is provided with a spherical cavity. The rotor mechanism comprises a rotor body, a connecting block, an actuating magnet set and a positioning magnet, the rotor body is movably arranged in the spherical cavity through a spherical pair, the connecting block is arranged on the rotor body and extends out of the stator body, the actuating magnet set is arranged on the rotor body and corresponds to the electromagnetic coil set, and the positioning magnet is arranged on the rotor body. The Hall sensor interacts with the positioning magnet and obtains the magnetic induction intensity. According to the technical scheme provided by the invention, the technical problems of low control precision and poor integration level of the spherical motor can be solved.

Description

Technical Field

[0001] The present invention relates to the technical field of motors, and in particular to a spherical motor and a position estimation method. Background Art

[0002] Spherical motors, also known as spherical motors and spherical servo motors, are a special type of motor that allows the rotor to rotate in three-dimensional space for precise control in multiple degrees of freedom. However, traditional spherical motors have technical problems such as low control accuracy and poor integration.

[0003] Therefore, it is necessary to provide a new spherical motor and position estimation method to solve the above technical problems. Summary of the invention

[0004] The main purpose of the present invention is to provide a spherical motor and a position estimation method, aiming to solve the technical problems of low control accuracy and poor integration of the spherical motor.

[0005] To achieve the above object, the present invention provides a spherical motor, comprising:

[0006] A stator mechanism, the stator mechanism comprising a stator body, an electromagnetic coil group and a Hall sensor arranged on the stator body, and the stator body is provided with a spherical cavity;

[0007] The rotor mechanism comprises a rotor body, a connecting block, an actuating magnet group and a positioning magnet, wherein the rotor body is movably arranged in the spherical cavity through a spherical pair, the connecting block is arranged on the rotor body and extends out of the stator body, the actuating magnet group is arranged on the rotor body, and the positioning magnet is arranged on the rotor body; the electromagnetic coil group is used to drive the actuating magnet group to move so as to drive the rotor body to move, and the Hall sensor interacts with the positioning magnet and obtains magnetic induction intensity.

[0008] In one embodiment, the stator body includes a stator seat and a stator cover, the stator seat is hemispherical, the stator cover is arranged on the stator seat and is surrounded by the stator seat to form the spherical cavity, the electromagnetic coil group is arranged on the stator seat, and the Hall sensor is arranged on the stator cover.

[0009] In one embodiment, the stator cover is provided with a mounting boss extending in the circumferential direction of the stator seat, the mounting boss is provided with a circuit board, and the circuit board is provided with a plurality of the Hall sensors evenly spaced in the circumferential direction of the stator seat.

[0010] In one embodiment, the stator seat has a hemispherical mounting surface, and a plurality of mounting grooves are evenly arranged on the mounting surface, and the mounting grooves are truncated cone-shaped;

[0011] The electromagnetic coil assembly includes a plurality of coil units, the number of the coil units is equal to the number of the mounting slots, and the plurality of coil units are arranged in the plurality of mounting slots in a one-to-one correspondence.

[0012] In one embodiment, the rotor body has a first hemispherical surface and a second hemispherical surface, the first hemispherical surface is arranged toward the stator seat, the second hemispherical surface is arranged toward the stator cover, and the actuating magnet group is arranged on the first hemispherical surface, and the positioning magnet is arranged on the second hemispherical surface.

[0013] In one embodiment, the number of the actuating magnet groups is multiple, and the multiple actuating magnet groups are arranged at intervals along the central axis direction of the rotor body;

[0014] Each of the actuating magnet groups comprises a plurality of permanent magnets, the plurality of permanent magnets are evenly spaced along the circumferential direction of the rotor body, and the polarities of the plurality of permanent magnets are alternately arranged;

[0015] There are multiple positioning magnets, and the multiple positioning magnets are evenly spaced along the circumferential direction of the rotor body.

[0016] The present invention further proposes a position estimation method, which is applied to the above-mentioned spherical motor, wherein the spherical motor is mounted on a fixing seat, a detection rod with a visual mark is mounted on the connecting block, and a shooting camera is mounted on one side of the spherical motor, wherein the shooting camera is used to obtain position data of the visual mark; the position estimation method comprises:

[0017] Obtain the magnetic induction intensity of the Hall sensor and input it into the training model;

[0018] Acquiring position information of the detection rod;

[0019] The step of obtaining the training model includes:

[0020] Acquire detection data of the Hall sensor and position data of the visual mark, wherein the detection data of the Hall sensor includes the magnetic induction intensity of the Hall sensor at different time points, and the position data of the visual mark includes the position information of the detection rod at different time points; perform data matching based on UNIX timestamps to acquire data pairs including the magnetic induction intensity of the Hall sensor and the position information of the visual mark, wherein each data pair constitutes a data set;

[0021] Based on the data set, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed, and a neural network is constructed based on the mapping relationship to generate a training model.

[0022] In one embodiment, after acquiring data pairs including the magnetic induction intensity of the Hall sensor and the position information of the visual mark, and before each of the data pairs forms a data set, the method further includes:

[0023] Comparing the difference between the magnetic induction intensity of the Hall sensor and the UNIX timestamp of the position information of the visual mark with a preset threshold;

[0024] If the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the detection rod in the data pair is less than a preset threshold, the data pair is output; if the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the detection rod in the data pair is greater than the preset threshold, the data pair is discarded.

[0025] In one embodiment, based on the data set, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed, and a neural network is constructed based on the mapping relationship, and the step of generating a training model includes: selecting part of the data set, and based on part of the data set, constructing a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark, and constructing a neural network based on the mapping relationship to generate a training model;

[0026] Obtain the training model and import it into the computer, and output the mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark; compare the difference between the position information of the visual mark output by the computer and the position information of the visual mark of another part of the data set when the magnetic induction intensity of the Hall sensor is the same, and the size of the first preset error value;

[0027] If the difference between the position information of the visual mark output by the computer and the position information of the visual mark in another part of the data set is less than the first preset error value, it means that the training model meets the requirements; if the difference between the position information of the visual mark output by the computer and the position information of the visual mark in another part of the data set is greater than the first preset error value, it means that the training model does not meet the requirements, optimize the neural network, generate a training model, and execute the steps of obtaining the training model and importing it into the computer;

[0028] Obtain the training model and import it into the controller of the spherical motor, and compare the difference between the position information of the visual mark output by the controller and the position information of the visual mark acquired by the shooting camera at the same time point with the size of a second preset error value;

[0029] If the difference between the position information of the visual mark output by the controller and the position information of the visual mark obtained by the shooting camera is less than the second preset error value, it means that the training model meets the requirements; if the difference between the position information of the visual mark output by the controller and the position information of the visual mark obtained by the shooting camera is greater than the second preset error value, it means that the training model does not meet the requirements, optimize the neural network, generate a training model, and execute the steps of obtaining the training model and importing it into the computer.

[0030] In one embodiment, after the step of obtaining the training model, the method further includes:

[0031] The training model is obtained and imported into the controller of the spherical motor, the detection rod is removed, and the spherical motor is installed on a mechanical device capable of three-degree-of-freedom control, and the controller of the mechanical device is electrically signal-connected to the controller of the spherical motor.

[0032] The technical solution of the present invention realizes the relative movement of the rotor body and the stator body through a spherical pair, and obtains the magnetic induction intensity through the interaction between the Hall sensor and the positioning magnet, and then calculates the position information of the rotor body, which can reduce the friction of the spherical motor, improve the control accuracy of the spherical motor, and improve the integration of the spherical motor. In this embodiment, the rotor body is movably arranged in the spherical cavity of the stator body through a spherical pair, which can reduce the friction during the operation of the spherical motor and improve the operation accuracy of the spherical motor. The Hall sensor can interact with the positioning magnet and transmit the magnetic induction intensity obtained to the controller of the spherical motor. The controller can convert the magnetic induction intensity obtained by the Hall sensor into the position information of the detection rod through a training model including a neural network, and then calculate the position information of the rotor body; when obtaining the position information of the rotor body, it does not need to contact the rotor body, which can reduce the friction during the operation of the spherical motor and improve the operation accuracy of the spherical motor. The actuating magnet group can interact with the electromagnetic coil group to drive the rotor body to move; specifically, when the electromagnetic coil group is energized, the electromagnetic coil group will generate a magnetic field. By changing the magnitude and direction of the current, the magnetic field generated by the electromagnetic coil group will change, and then it will drive the actuating magnet group to move, and then drive the rotor body to move. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0034] Figure 1 A schematic diagram of the structure of a spherical motor in an embodiment of the present invention;

[0035] Figure 2 A schematic diagram of the structure of a stator mechanism in an embodiment provided by the present invention;

[0036] Figure 3 for Figure 2 Schematic diagram from another perspective;

[0037] Figure 4 A schematic structural diagram of a rotor mechanism in an embodiment of the present invention;

[0038] Figure 5 for Figure 4 Schematic diagram from another perspective;

[0039] Figure 6 A schematic diagram of the installation of a spherical motor when obtaining a training model in an embodiment provided by the present invention;

[0040] Figure 7 A flow chart of a position estimation method in an embodiment of the present invention;

[0041] Figure 8 A flowchart of obtaining a training model in an embodiment provided by the present invention.

[0042] Description of Figure Numbers:

[0043] 100, stator mechanism; 110, stator body; 111, spherical cavity; 112, stator seat; 1121, mounting surface; 1122, mounting groove; 113, stator cover; 1131, mounting boss; 200, rotor mechanism; 210, rotor body; 211, first hemispherical surface; 212, second hemispherical surface; 220, connecting block; 230, actuating magnet group; 231, permanent magnet; 240, positioning magnet; 310, detection rod; 320, fixing seat; 330, shooting camera.

[0044] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0047] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, if "and / or" or "and / or" appears in the full text, its meaning includes three parallel schemes. Taking "A and / or B" as an example, it includes scheme A, or scheme B, or a scheme that satisfies both A and B.

[0048] In addition, the technical solutions between the various embodiments of the present invention can be combined with each other, but it must be based on the fact that ordinary technicians in the field can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0049] Spherical motors are also called spherical motors and spherical servo motors. Spherical motors allow the rotor to rotate in three-dimensional space to achieve precise control in multiple degrees of freedom. Spherical motors include rotors and stators. In actual use, researchers found that the rotors of traditional spherical motors are mostly supported by ball bearings, which increases the friction between the rotor and the stator and affects the operating accuracy of the spherical motor. In addition, during the operation of the spherical motor, the position of the rotor needs to be sensed in real time. Traditional spherical motors mostly use a contact structure to achieve position sensing. During the position sensing process, they need to contact the rotor, which increases friction and affects the operating accuracy of the spherical motor.

[0050] The present invention provides a spherical motor and a position estimation method, aiming to solve the technical problems of low control accuracy and poor integration of the spherical motor.

[0051] See also Figure 1 , Figure 2 and Figure 4In one embodiment of the present invention, the spherical motor includes a stator mechanism 100 and a rotor mechanism 200. The stator mechanism 100 includes a stator body 110, an electromagnetic coil group and a Hall sensor arranged on the stator body 110. The stator body 110 is provided with a spherical cavity 111. The rotor mechanism 200 includes a rotor body 210, a connecting block 220, an actuating magnet group 230 and a positioning magnet 240. The rotor body 210 is movably arranged in the spherical cavity 111 through a spherical pair. The connecting block 220 is arranged on the rotor body 210 and extends out of the stator body 110. The actuating magnet group 230 is arranged on the rotor body 210, and the positioning magnet 240 is arranged on the rotor body 210. The electromagnetic coil group is used to drive the actuating magnet group 230 to move, so as to drive the rotor body 210 to move. The Hall sensor interacts with the positioning magnet and obtains magnetic induction intensity.

[0052] The technical solution of the present invention realizes the relative movement of the rotor body 210 and the stator body 110 through the spherical pair, and obtains the magnetic induction intensity through the interaction between the Hall sensor and the positioning magnet, and then calculates the position information of the rotor body 210, which can reduce the friction of the spherical motor, improve the control accuracy of the spherical motor, and improve the integration of the spherical motor. In this embodiment, the rotor body 210 is movably arranged in the spherical cavity 111 of the stator body 110 through the spherical pair, which can reduce the friction when the spherical motor is running and improve the operation accuracy of the spherical motor. The Hall sensor can interact with the positioning magnet 240 and transmit the magnetic induction intensity obtained by it to the controller of the spherical motor. The controller can convert the magnetic induction intensity obtained by the Hall sensor into the position information of the detection rod 310 through the training model containing the neural network, and then calculate the position information of the rotor body 210; when obtaining the position information of the rotor body 210, it does not need to contact the rotor body 210, which can reduce the friction when the spherical motor is running and improve the operation accuracy of the spherical motor. The actuating magnet group 230 can interact with the electromagnetic coil group to drive the rotor body 210 to move; specifically, when the electromagnetic coil group is energized, the electromagnetic coil group will generate a magnetic field. By changing the magnitude and direction of the current, the magnetic field generated by the electromagnetic coil group will change, and then the actuating magnet group 230 will be driven to move, thereby driving the rotor body 210 to move.

[0053] It should be noted that the reason why the relative motion between the rotor body 210 and the stator body 110 can be realized by using a spherical pair and the friction during operation of the spherical motor can be reduced is that the spherical pair makes the rotor body 210 and the stator body 110 in spherical contact, and the spherical contact produces an arc-shaped contact area, which helps to disperse the contact stress and reduce the friction between the rotor body 210 and the stator body 110. At the same time, the use of a spherical pair connection can also improve the load-bearing capacity of the spherical motor, simplify the structure of the spherical motor, and reduce the manufacturing cost.

[0054] See also Figure 2 and Figure 3 In one embodiment of the present invention, the stator body 110 includes a stator seat 112 and a stator cover 113. The stator seat 112 is hemispherical. The stator cover 113 is arranged on the stator seat 112 and is surrounded by the stator seat 112 to form a spherical cavity 111. The electromagnetic coil group is arranged on the stator seat 112, and the Hall sensor is arranged on the stator cover 113. In this embodiment, the stator seat 112 is designed to be hemispherical, which can reduce the volume of the spherical motor and facilitate the installation of the spherical motor. In a specific embodiment, the stator seat 112 is provided with an extension plate along its circumference, and the extension plate is connected to the stator cover 113 by bolts. The stator seat 112 is a plastic part or a metal part made by 3D printing technology, wherein the metal part has better thermal conductivity. The use of metal to make the stator seat 112 can improve the stability of the spherical motor during operation.

[0055] In one embodiment of the present invention, the stator cover 113 is provided with a mounting boss 1131 extending along the circumferential direction of the stator seat 112, and the mounting boss 1131 is provided with a circuit board, and the circuit board is evenly spaced along the circumferential direction of the stator seat 112. Multiple Hall sensors are arranged. In this embodiment, the Hall sensor can interact with the positioning magnet 240, and transmit the magnetic induction intensity obtained to the controller of the spherical motor, and the controller can convert the magnetic induction intensity obtained by the Hall sensor into the position information of the detection rod 310 through a training model including a neural network, and then calculate the position information of the rotor body 210. The position information of the rotor body 210 is obtained by using the Hall sensor and the positioning magnet 240 to cooperate, which can more accurately determine the position of the rotor body 210 and realize high-precision position recognition. In addition, when the Hall sensor obtains the position of the positioning magnet 240, it does not need to contact the rotor body 210, which can reduce the friction of the spherical motor during operation and improve the operation accuracy of the spherical motor. At the same time, the Hall sensor also has the advantages of small size and low cost, which can reduce the size of the spherical motor and reduce the manufacturing cost of the spherical motor. In a specific embodiment, the circuit board is circular or C-shaped, and the number of the Hall sensors is four. The four Hall sensors are evenly spaced and arranged on the circuit board along the circumferential direction of the stator seat 112 .

[0056] In one embodiment of the present invention, the stator seat 112 has a hemispherical mounting surface 1121, and a plurality of mounting grooves 1122 are evenly arranged on the mounting surface 1121, and the mounting grooves 1122 are truncated cone-shaped; the electromagnetic coil group includes a plurality of coil units, and the number of coil units and mounting grooves 1122 is equal, and the plurality of coil units are arranged in a one-to-one correspondence in the plurality of mounting grooves 1122. In this embodiment, each coil unit is truncated cone-shaped, and by providing a plurality of truncated cone-shaped mounting grooves 1122 on the stator seat 112, and arranging the plurality of coil units in a one-to-one correspondence in the plurality of mounting grooves 1122, the arrangement density of the coil units on the mounting surface 1121 of the stator seat 112 can be increased, thereby increasing the output torque of the spherical motor. The reason is that installing the truncated cone-shaped coil unit into the truncated cone-shaped installation groove 1122 can effectively avoid interference between the coil unit and the stator base 112; if the installation groove 1122 and the coil unit are both cylindrical, since the diameter of the stator base 112 gradually decreases from the outside to the inside, when the coil unit is installed into the stator base 112, the ends of two adjacent coil units inserted into the interior of the stator base 112 are very likely to interfere. In a specific embodiment, the number of the installation grooves 1122 and the number of the coil units are both eight, and among the eight installation grooves 1122, two of the installation grooves 1122 are opened on the side of the stator base 112 away from the stator cover 113, and the other six installation grooves 1122 are evenly spaced and arranged on the outside of two of the installation grooves 1122, and among the eight installation grooves 1122, four of the installation grooves 1122 are arranged in an arc shape. In this embodiment, the controller of the spherical motor includes a driving board, which is connected to the coil unit through a wire or the like, and the position of the rotor body 210 of the spherical motor can be adjusted by running a driving control program of the spherical motor.

[0057] See also Figure 4 and Figure 5In one embodiment of the present invention, the rotor body 210 has a first hemispherical surface 211 and a second hemispherical surface 212, the first hemispherical surface 211 is arranged toward the stator seat 112, and the second hemispherical surface 212 is arranged toward the stator cover 113, and the actuating magnet group 230 is arranged on the first hemispherical surface 211, and the positioning magnet 240 is arranged on the second hemispherical surface 212. In this embodiment, the actuating magnet group 230 and the positioning magnet 240 are respectively arranged on two opposite hemispherical surfaces of the rotor body 210, which can increase the distance between the actuating magnet group 230 and the positioning magnet 240, and further increase the distance between the actuating magnet group 230 and the Hall sensor, so as to reduce the influence of the actuating magnet group 230 on the Hall sensor, and further make the detection data of the Hall sensor more accurate. Correspondingly, a mounting boss 1131 is arranged on the stator cover 113, and a circuit board with the Hall sensor mounted thereon is mounted on the mounting boss 1131, which can also increase the distance between the actuating magnet group 230 and the Hall sensor, and make the detection data of the Hall sensor more accurate. In a specific embodiment, there are multiple positioning magnets 240, and the multiple positioning magnets 240 are evenly spaced along the circumferential direction of the rotor body 210. In a more specific embodiment, there are two positioning magnets 240, and the two positioning magnets 240 are symmetrically arranged, and the polarities of the two positioning magnets 240 are opposite.

[0058] In one embodiment of the present invention, there are multiple actuating magnet groups 230, and the multiple actuating magnet groups 230 are arranged at intervals along the central axis direction of the rotor body 210, and each actuating magnet group 230 includes multiple permanent magnets 231, and the multiple permanent magnets 231 are evenly arranged along the circumferential direction of the rotor body 210, and the polarities of the multiple permanent magnets 231 are arranged alternately. In this embodiment, the polarities of the permanent magnets 231 are arranged alternately, which can generate a more uniform and continuous magnetic field, thereby generating a greater torque on the rotor body 210, thereby increasing the motion angle of the spherical motor and improving the output torque of the spherical motor; at the same time, it can also provide more flexible magnetic field control and enhance the dynamic performance of the spherical motor. In a specific embodiment, there are two actuating magnet groups 230, and each actuating magnet group 230 includes four permanent magnets 231, that is, 8 permanent magnets 231 are installed on the rotor body of the spherical motor.

[0059] See also Figure 7 The present invention also proposes a position estimation method, which is applied to the above-mentioned spherical motor. The position estimation method comprises:

[0060] S100, obtaining the magnetic induction intensity of the Hall sensor and inputting it into the training model;

[0061] S200 , obtaining position information of the detection rod 310 ; then, the controller can obtain the relative posture of the rotor body 210 of the spherical motor according to the obtained position information of the detection rod 310 .

[0062] See also Figure 6 When using the position estimation method, it is necessary to first obtain a training model containing a neural network. Prior to this, the spherical motor needs to be installed on the fixing seat 320, a detection rod 310 with a visual mark is installed on the connecting block 220, and a shooting camera 330 for obtaining the position data of the visual mark is installed on one side of the spherical motor; wherein, the detection rod 310 with a visual mark can be a detection rod 310 with a CylinderTag code, and in this embodiment, a card slot is provided on the connecting block 220, and a card block is provided on the detection rod 310. The detection rod 310 can be installed on the connecting block 220 by inserting the card block into the card slot. It should be noted that the CylinderTag code is a visual mark specially designed for cylindrical objects, aiming to achieve high-precision posture estimation; it encodes information in the direction of zero curvature of the surface by utilizing the cross ratio in the projection invariance, thereby achieving accurate tracking and positioning of cylindrical objects.

[0063] See also Figure 8 , the steps to obtain the training model include:

[0064] The detection data of the Hall sensor and the position data of the visual mark are obtained, wherein the detection data of the Hall sensor includes the magnetic induction intensity of the Hall sensor at different time points, and the position data of the visual mark includes the position information of the visual mark at different time points; data matching is performed based on the UNIX timestamp to obtain data pairs including the magnetic induction intensity of the Hall sensor and the position information of the visual mark, and each data pair constitutes a data set; based on the data set, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed, and a neural network is constructed based on the mapping relationship to generate a training model. Among them, the position information of the visual mark can be the quaternion data of the CylinderTag code.

[0065] Specifically, the controller of the spherical motor can obtain the magnetic induction intensity of the Hall sensor, and use serial communication to send the data to the computer in real time to obtain the detection data of the Hall sensor; at the same time, the shooting data of the camera 330 can be transmitted to the computer, and the video data shot by the camera 330 is processed by the computer, and the CylinderTag posture estimation algorithm is used to calculate the posture of the detection rod 310 in real time (when recording the posture of the detection rod 310, a posture zero point can be set manually, and the relative posture between other positions and the posture zero point is calculated) to obtain the position data of the visual mark. Subsequently, based on the UNIX timestamp (also called POSIX timestamp or Epoch time, which refers to the number of seconds from January 1, 1970 to the current time.) Data matching is performed, and data pairs that are close in time are selected, and each data pair is formed into a data set. Based on the data set, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed. The computer will use the matched data set to train a neural network through the back propagation algorithm to generate a training model. During the training process, the training model will learn how to predict the position information of the corresponding visual marker according to the magnetic induction intensity of the input Hall sensor, and will minimize the difference between the predicted output and the actual output by adjusting the internal weights and biases. In the embodiment of the present invention, the input layer is the 12 magnetic induction intensities of the Hall sensor, and the output layer is the 4 quaternion data of the visual marker.

[0066] In one embodiment of the present invention, after obtaining the data pairs including the magnetic induction intensity of the Hall sensor and the position information of the visual mark, before each data pair forms a data set, it also includes: comparing the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the visual mark with the preset threshold; if the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the detection rod 310 of the data pair is less than the preset threshold, then the data pair is output; if the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the detection rod 310 of the data pair is greater than the preset threshold, then the data pair is discarded. The part of the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the visual mark that is less than the preset threshold is retained to ensure that the magnetic induction intensity of the Hall sensor of the data pair and the position information of the visual mark are collected at a similar time point, so as to more accurately analyze the relationship between the two. In this embodiment, the preset threshold can be 5ms.

[0067] In one embodiment of the present invention, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed based on the data set, and a neural network is constructed based on the mapping relationship, and the steps of generating a training model include: selecting a part of the data set, constructing a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark based on the part of the data set, constructing a neural network based on the mapping relationship, and generating a training model;

[0068] Obtain the training model and import it into the computer, output the mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark; compare the difference between the position information of the visual mark output by the computer and the position information of the visual mark in another part of the data set with the first preset error value when the magnetic induction intensity of the Hall sensor is the same; if the difference between the position information of the visual mark output by the computer and the position information of the visual mark in another part of the data set is less than the first preset error value, it means that the training model meets the requirements; if the difference between the position information of the visual mark output by the computer and the position information of the visual mark in another part of the data set is greater than the first preset error value, it means that the training model does not meet the requirements, optimize the neural network, generate the training model, and execute the steps of obtaining the training model and importing it into the computer;

[0069] Obtain the training model and import it into the controller of the spherical motor, compare the difference between the position information of the visual mark output by the controller and the position information of the visual mark acquired by the shooting camera 330 at the same time point with the second preset error value; if the difference between the position information of the visual mark output by the controller and the position information of the visual mark acquired by the shooting camera 330 is less than the second preset error value, it means that the training model meets the requirements; if the difference between the position information of the visual mark output by the controller and the position information of the visual mark acquired by the shooting camera 330 is greater than the second preset error value, it means that the training model does not meet the requirements, optimize the neural network, generate the training model, and execute the steps of obtaining the training model and importing it into the computer.

[0070] Specifically, when generating a training model, 70%-80% of the data set is selected to generate the training model, and after the training model is generated, the initial performance of the training model can be verified based on the remaining 20%-30% of the data set. Specifically: obtain the training model and import it into the computer, and then compare the difference between the position information of the visual mark output by the computer and the position information of the visual mark of the remaining 20%-30% of the data set when the magnetic induction intensity of the Hall sensor is the same with the first preset error value to verify the initial performance of the training model. If the difference between the two is less than the first preset error value, it means that the training model meets the requirements; if the difference between the two is greater than the first preset error value, it means that the training model does not meet the requirements, and then further adjust the neural network (for example, increase or decrease the number of layers, adjust the number of neurons, etc.) to generate a new training model, and then execute the steps of obtaining the training model and importing it into the computer until the training model meets the requirements.

[0071] Similarly, after verifying the preliminary performance of the training model on the computer side, the training model can be obtained and imported into the controller of the spherical motor to evaluate its performance in the hardware environment. Specifically: import the training model into the controller of the spherical motor, and compare the difference between the position information of the visual mark output by the controller and the position information of the visual mark obtained by the shooting camera 330 at the same time point with the size of the second preset error value. If the difference between the two is less than the second preset error value, it means that the training model meets the requirements, that is, the training model performs well in the hardware environment; if the difference between the two is greater than the second preset error value, it means that the training model does not meet the requirements, that is, the training model performs poorly in the hardware environment, and then further adjust the neural network (for example, increase or decrease the number of layers, adjust the number of neurons, etc.) to generate a new training model, and then execute the steps of obtaining the training model and importing it into the computer side until the training model meets the requirements.

[0072] After the step of obtaining the training model, it also includes: obtaining the training model and importing the controller of the spherical motor, removing the detection rod, and installing the spherical motor on a mechanical device capable of three-degree-of-freedom control, and the controller of the mechanical device is electrically connected to the controller of the spherical motor. Specifically, after completing the preliminary calibration of the training model, the spherical motor is installed on a mechanical device capable of three-degree-of-freedom control, and by obtaining the magnetic induction intensity of the Hall sensor, the position information of the rotor can be calculated by the training model including the neural network, so as to realize autonomous position perception without the detection rod. In a specific embodiment, the mechanical device capable of three-degree-of-freedom control can be a cross-medium robot, which includes a fuselage, a wing and a controller, a spherical motor is installed on the fuselage, a rotor body of the spherical motor is connected to the wing, and the spherical motor is electrically connected to the controller. After completing the preliminary calibration of the training model, the spherical motor can be installed on the cross-medium robot, the wing is driven to swing by the spherical motor, and the controller of the cross-medium robot is electrically connected to the spherical motor, and the controller of the cross-medium robot can obtain the position information of the rotor body in real time through the spherical motor, and then obtain the relative posture of the wing.

[0073] In this embodiment, in order to prevent the training model from overfitting during the training process, an early stopping mechanism is introduced into the training model to ensure that the training model has good generalization ability. Overfitting refers to the situation where the model performs well on the training data but performs poorly on the unseen data.

[0074] The above description is only an exemplary embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the technical concept of the present invention, or directly / indirectly applied in other related technical fields are included in the patent protection scope of the present invention.

Claims

1. A spherical motor, characterized in that: include: A stator mechanism, the stator mechanism comprising a stator body, an electromagnetic coil group and a Hall sensor arranged on the stator body, and the stator body is provided with a spherical cavity; The rotor mechanism comprises a rotor body, a connecting block, an actuating magnet group and a positioning magnet, wherein the rotor body is movably arranged in the spherical cavity through a spherical pair, the connecting block is arranged on the rotor body and extends out of the stator body, the actuating magnet group is arranged on the rotor body, and the positioning magnet is arranged on the rotor body; the electromagnetic coil group is used to drive the actuating magnet group to move so as to drive the rotor body to move, and the Hall sensor interacts with the positioning magnet and obtains magnetic induction intensity.

2. The spherical motor according to claim 1, characterized in that: The stator body comprises a stator seat and a stator cover, the stator seat is hemispherical, the stator cover is arranged on the stator seat and surrounds the stator seat to form the spherical cavity, the electromagnetic coil group is arranged on the stator seat, and the Hall sensor is arranged on the stator cover.

3. The spherical motor according to claim 2, characterized in that: The stator cover is provided with a mounting boss extending in the circumferential direction of the stator seat, the mounting boss is provided with a circuit board, and the circuit board is provided with a plurality of the Hall sensors evenly spaced in the circumferential direction of the stator seat.

4. The spherical motor according to claim 2, characterized in that: The stator seat has a hemispherical mounting surface, on which a plurality of mounting grooves are evenly arranged, and the mounting grooves are truncated cone-shaped; The electromagnetic coil assembly includes a plurality of coil units, the number of the coil units is equal to the number of the mounting slots, and the plurality of coil units are arranged in the plurality of mounting slots in a one-to-one correspondence.

5. The spherical motor according to claim 2, characterized in that: The rotor body has a first hemispherical surface and a second hemispherical surface, the first hemispherical surface is arranged toward the stator seat, the second hemispherical surface is arranged toward the stator cover, the actuating magnet group is arranged on the first hemispherical surface, and the positioning magnet is arranged on the second hemispherical surface.

6. The spherical motor according to claim 5, characterized in that: There are multiple actuating magnet groups, and the multiple actuating magnet groups are arranged at intervals along the central axis direction of the rotor body; Each of the actuating magnet groups comprises a plurality of permanent magnets, the plurality of permanent magnets are evenly spaced along the circumferential direction of the rotor body, and the polarities of the plurality of permanent magnets are alternately arranged; There are multiple positioning magnets, and the multiple positioning magnets are evenly spaced along the circumferential direction of the rotor body.

7. A position estimation method, applied to a spherical motor as claimed in any one of claims 1 to 6, characterized in that: The spherical motor is mounted on a fixing seat, a detection rod with a visual mark is mounted on the connecting block, and a shooting camera is mounted on one side of the spherical motor, the shooting camera is used to obtain position data of the visual mark; The position estimation method comprises: Obtain the magnetic induction intensity of the Hall sensor and input it into the training model; Acquiring position information of the detection rod; The step of obtaining the training model includes: Acquire detection data of the Hall sensor and position data of the visual mark, wherein the detection data of the Hall sensor includes the magnetic induction intensity of the Hall sensor at different time points, and the position data of the visual mark includes the position information of the detection rod at different time points; perform data matching based on UNIX timestamps to acquire data pairs including the magnetic induction intensity of the Hall sensor and the position information of the visual mark, wherein each data pair constitutes a data set; Based on the data set, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed, and a neural network is constructed based on the mapping relationship to generate a training model.

8. The position estimation method according to claim 7, characterized in that: After acquiring data pairs including the magnetic induction intensity of the Hall sensor and the position information of the visual mark, and before each of the data pairs forms a data set, the method further includes: Comparing the difference between the magnetic induction intensity of the Hall sensor and the UNIX timestamp of the position information of the visual mark with a preset threshold; If the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the detection rod in the data pair is less than a preset threshold, the data pair is output; if the UNIX timestamp difference between the magnetic induction intensity of the Hall sensor and the position information of the detection rod in the data pair is greater than the preset threshold, the data pair is discarded.

9. The position estimation method according to claim 7, characterized in that: Based on the data set, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed, and a neural network is constructed based on the mapping relationship. The step of generating a training model includes: selecting a part of the data set, and based on the part of the data set, a mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark is constructed, and a neural network is constructed based on the mapping relationship to generate a training model; Obtain the training model and import it into the computer, and output the mapping relationship between the magnetic induction intensity of the Hall sensor and the position information of the visual mark; compare the difference between the position information of the visual mark output by the computer and the position information of the visual mark of another part of the data set when the magnetic induction intensity of the Hall sensor is the same, and the size of the first preset error value; If the difference between the position information of the visual mark output by the computer and the position information of the visual mark in another part of the data set is less than the first preset error value, it means that the training model meets the requirements; if the difference between the position information of the visual mark output by the computer and the position information of the visual mark in another part of the data set is greater than the first preset error value, it means that the training model does not meet the requirements, optimize the neural network, generate a training model, and execute the steps of obtaining the training model and importing it into the computer; Obtain the training model and import it into the controller of the spherical motor, and compare the difference between the position information of the visual mark output by the controller and the position information of the visual mark acquired by the shooting camera at the same time point with the size of a second preset error value; If the difference between the position information of the visual mark output by the controller and the position information of the visual mark obtained by the shooting camera is less than the second preset error value, it means that the training model meets the requirements; if the difference between the position information of the visual mark output by the controller and the position information of the visual mark obtained by the shooting camera is greater than the second preset error value, it means that the training model does not meet the requirements, optimize the neural network, generate a training model, and execute the steps of obtaining the training model and importing it into the computer.

10. The position estimation method according to claim 7, characterized in that: After the step of obtaining the training model, the method further includes: The training model is obtained and imported into the controller of the spherical motor, the detection rod is removed, and the spherical motor is installed on a mechanical device capable of three-degree-of-freedom control, and the controller of the mechanical device is electrically signal-connected to the controller of the spherical motor.

Images