A multi-degree-of-freedom robot control method and system

Abstract

The present invention discloses a multi-degree-of-freedom robot control method and system, which relates to the field of robot control, and obtains image processing results; determines the direction of an obstacle and the distance between the obstacle and the multi-degree-of-freedom robot; draws a three-dimensional obstacle map based on the direction and distance of the obstacle; determines whether the multi-degree-of-freedom robot can cross the obstacle; a controller schedules the operating parameters of each component of the multi-degree-of-freedom robot to complete the obstacle crossing; determines the size and shape of the object to be grasped; and based on the size and shape of the grasped object, a processor determines the grasping point and force of the multi-degree-of-freedom robot. The advantages of the present invention are: by setting an obstacle model building module, an obstacle crossing determination module, a route detour module, and an object grasping module, precise control of the multi-degree-of-freedom robot crossing obstacles and picking up objects is achieved, avoiding grasping errors or crossing errors caused by inconsistent calls to certain components.

Description

Technical Field

[0001] The present invention relates to the field of robot control, and in particular to a multi-degree-of-freedom robot control method and system. Background Art

[0002] A multi-degree-of-freedom robot is one with multiple independent degrees of freedom. These degrees of freedom typically refer to the number of axes or joints that can independently move in three-dimensional space. Degrees of freedom are a core measure of a robot's flexibility and range of functions, directly determining the complexity and precision of tasks it can perform.

[0003] Because multi-degree-of-freedom robots have many moving axes and joints, problems such as incoordination of related joints are prone to occur when controlling multi-degree-of-freedom robots, which can make it difficult to complete operations. Although multi-degree-of-freedom robots have more precise and flexible operations, there are higher requirements for controlling multi-degree-of-freedom robots, and a series of operating guidelines need to be set. Summary of the Invention

[0004] In order to solve the above technical problems, a multi-degree-of-freedom robot control method and system are proposed. This technical solution solves the problems raised in the above background technology.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0006] A multi-degree-of-freedom robot control method obtains image information from a visual sensor, and a processor analyzes the image information to obtain an image processing result;

[0007] Based on the image processing results, the direction of the obstacle is determined, and the distance between the obstacle and the multi-degree-of-freedom robot is determined using an ultrasonic sensor;

[0008] Draw a three-dimensional obstacle map based on the direction and distance of obstacles;

[0009] Based on the height of the obstacle and the robot's own parameters, determine whether the multi-degree-of-freedom robot can cross the obstacle;

[0010] If it can, the robot crosses the obstacle;

[0011] Based on the height of the obstacle, the controller adjusts the operating parameters of each component of the multi-degree-of-freedom robot to complete the obstacle crossing;

[0012] If not, the robot plans a detour route;

[0013] If there is a grasping task, the size and shape of the object to be grasped are determined based on the image information of the visual sensor and the distance judgment of the ultrasonic sensor;

[0014] Based on the size and shape of the object being grasped, the processor determines the grasping point and force of the multi-degree-of-freedom robot.

[0015] Preferably, the acquiring of the image information from the visual sensor, the processor analyzing the image information, and obtaining the image processing result include:

[0016] The multi-degree-of-freedom robot rotates its head vision sensor to obtain images from different perspectives;

[0017] Mark the horizontal line in the middle of the picture as the eye level;

[0018] Find at least two parallel lines in the real world, extend at least two parallel lines in the two-dimensional image, and record the intersection as the vanishing point, and the parallel lines as the vanishing lines;

[0019] If the first object on the same vanishing line is farther from the vanishing point than the second object, then the first object is behind the second object.

[0020] Get the height difference between the object and the horizon. If the height difference between the object and the horizon is negative, it means that the height of the object is lower than the object with a positive height difference.

[0021] The position relationship and height information of the objects are stored in the memory.

[0022] Preferably, the object with the shortest straight-line distance to the visual sensor is selected as the reference object;

[0023] Measure the distance from the vision sensor to the bottom and top of the reference object, and use the Pythagorean theorem to obtain the height of the reference object;

[0024] obtaining a height of the object by multiplying the height of the reference object by the relative height ratio based on a relative height ratio between at least one object in the picture and a reference object, wherein the relative height ratio is the height of the object in the picture from the horizon in the picture divided by the height of the reference object in the picture from the horizon in the picture;

[0025] Use the distance calculation formula to calculate the distance between objects;

[0026] The distance calculation formula is: ,

[0027] Where S is the distance between the first object and the second object, is the distance from the first reference object to the visual sensor, is the distance from the second reference object to the visual sensor, is the distance from the first object to the vanishing point, is the distance from the second object to the vanishing point, is the distance from the first reference object to the vanishing point, is the distance from the second reference object to the vanishing point.

[0028] The heights of all objects in the picture, the distances between objects, and the front-to-back relationships between objects are counted, and the objects are abstracted into geometric bodies to obtain a three-dimensional obstacle map.

[0029] Preferably, the method comprises collecting statistics on the bending angles of the hip joint, knee joint and ankle joint of normal humans when crossing obstacles to establish a database;

[0030] Obtain the length from the hip joint to the knee joint and the length from the knee joint to the ankle joint when the multi-degree-of-freedom robot is standing normally;

[0031] Match the flexion angles of the hip, knee, and ankle joints of humans when crossing obstacles of the same height to the data related to the multi-degree-of-freedom robot legs in the database;

[0032] To analyze the physical center of gravity when the bending angle of each joint reaches the maximum, the steps are as follows:

[0033] The multi-degree-of-freedom robot is divided into two parts from the hip joint, and the physical center of gravity of the upper part and the geometric center of gravity of the upper part are on the same vertical line;

[0034] Divide the upper part into upper and lower halves along a horizontal line located in the middle of the vertical line of the upper part, and count the mass of the upper half and the mass of the lower half;

[0035] The physical center of gravity of the upper part is located at a preset distance from the middle point of the vertical line in the upper part. If the preset value is a positive number, it means that the physical center of gravity is above the middle point. If the preset value is a negative number, it means that the physical center of gravity is below the middle point.

[0036] The physical center of gravity of the thighs and calves on both sides of the multi-degree-of-freedom robot is obtained through the suspension method;

[0037] Obtain the three-dimensional coordinates of the physical center of gravity of the thighs and calves on both sides when the bending angle of each joint of the multi-degree-of-freedom robot reaches the maximum;

[0038] Use the center of gravity position formula to calculate the physical center of gravity coordinates of the lower part. The center of gravity position formula is as follows:

[0039] ,

[0040] Where, is the physical center of gravity of the lower part, is the mass of the left thigh, is the mass of the right thigh, is the mass of the left calf, is the mass of the right calf, are the coordinates of the left thigh, are the coordinates of the right thigh, The coordinates of the left calf is the coordinate of the right calf;

[0041] Get the coordinates of the upper physical center of gravity. The sub-coordinates of the physical center of gravity of the multi-DOF robot are the sum of the sub-coordinates of the upper physical center of gravity multiplied by the mass of the upper part plus the sub-coordinates of the lower physical center of gravity multiplied by the mass of the lower part, divided by the overall mass of the multi-DOF robot.

[0042] Determine whether the physical center of gravity and the geometric center of gravity of the multi-degree-of-freedom robot are on the same vertical line. If so, the multi-degree-of-freedom robot can cross the obstacle;

[0043] The steps for obtaining the preset value are as follows:

[0044] The preset value is obtained using a distance formula, which is as follows:

[0045] ,

[0046] Where A is the preset value, and the upper half mass is , the lower half mass is , the length of the vertical line in the upper part is 1.

[0047] Preferably, the method comprises collecting statistics on the bending angles of the hip joint, knee joint and ankle joint of normal humans when crossing obstacles to establish a database;

[0048] Obtain the length from the hip joint to the knee joint and the length from the knee joint to the ankle joint when the multi-degree-of-freedom robot is standing normally;

[0049] Match the flexion angles of the hip, knee, and ankle joints of humans when crossing obstacles of the same height to the data related to the multi-degree-of-freedom robot legs in the database;

[0050] The controller adjusts the bending angle of each joint of the multi-degree-of-freedom robot according to the matching data.

[0051] Preferably, the step of obtaining the height and maximum width of the multi-degree-of-freedom robot;

[0052] Based on the three-dimensional obstacle map, the entrances with a height greater than the height of the multi-degree-of-freedom robot and a width greater than the maximum width of the multi-degree-of-freedom robot are selected and recorded as candidate entrances;

[0053] The multi-degree-of-freedom robot preferentially chooses the rightmost alternative entrance to enter. After entering the entrance, it determines whether the multi-degree-of-freedom robot can cross the obstacle it encounters again. If so, it crosses the obstacle it encounters again. If not, it plans a detour route for the obstacle it encounters again until the multi-degree-of-freedom robot reaches the working position.

[0054] Preferably, the ultrasonic sensor is used to obtain the distance and direction between the ultrasonic sensor and at least one point on the grasped object;

[0055] In the simulation system, projecting the position of at least one point according to the distance direction of the at least one point;

[0056] Based on the distribution of the point graph, connect at least one point to construct a surface to obtain the basic model for grasping the object;

[0057] Based on the image information of the visual sensor, artificial intelligence is used to search and compare to determine the type of objects in the image;

[0058] Based on the type of grasped object, the surface roughness is refined on the basis of the basic model to obtain the surface friction coefficient of the grasped object;

[0059] Based on the type of grasped object, the density of the grasped object is determined, the volume of the object is determined through image recognition, and the gravity of the object is calculated based on the volume and density of the object.

[0060] Preferably, at least one sampling point is uniformly taken on the surface of the object, a tangent plane of the object surface is made at the sampling point, and the angle between the tangent plane corresponding to the sampling point and the vertical direction is taken as the relative angle;

[0061] The number of fingers of the multi-degree-of-freedom robot is used as the eigenvalue;

[0062] The object's gravity is divided by the friction coefficient of the grasping point and then divided by the characteristic value to obtain the finger grasping force;

[0063] According to the type of object, the tolerable pressure of the sampling point of the object is obtained;

[0064] The sampling point that can withstand a pressure greater than the finger grasping force is used as the target sampling point;

[0065] Randomly combining target sampling points to form at least one target sampling point combination, where the number of sampling points included in the target sampling point combination is equal to the eigenvalue;

[0066] The target sampling points that are coplanar with the target sampling points are combined as the calibration sampling point combination;

[0067] The closed plane area formed by sequentially connecting the adjacent target sampling points in the calibration sampling point combination is used as the plane sampling area;

[0068] The area through which the plane sampling area moves in the vertical direction is used as the stereo projection range of the calibration sampling point combination;

[0069] Obtain the center of gravity of the object, and use the calibration sampling point combination corresponding to the stereoscopic projection range containing the center of gravity of the object as the preliminary sampling point combination;

[0070] The surface of the object located above the plane where the plane sampling area is located is taken as the feature surface;

[0071] Selecting a characteristic sampling point combination from at least one calibration sampling point combination, so that the area of ​​the characteristic surface generated by the plane sampling region corresponding to the characteristic sampling point combination is the smallest;

[0072] Use the sampling points in the feature sampling point combination as the grabbing points.

[0073] Compared with the existing technology, the beneficial effects of the present invention are: determining three-dimensional obstacles based on visual sensors and ultrasonic sensors, providing a basis for multi-degree-of-freedom robots to cross obstacles, judging in advance whether the multi-degree-of-freedom robot can cross obstacles, avoiding energy waste caused by finding that it cannot cross during the crossing process, and an important basis for judging whether the multi-degree-of-freedom robot can cross obstacles is whether the center of gravity is unstable and causes falls during the crossing process, avoiding damage to the robot caused by falls during the crossing process, and preventing the grasped objects from slipping by locating the grasping points. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 Schematic diagram of a flow chart of a multi-degree-of-freedom robot control method of the present invention;

[0075] Figure 2 This is a flow chart of the process of obtaining image information from a visual sensor, analyzing the image information by a processor, and obtaining image processing results according to the present invention;

[0076] Figure 3 A schematic diagram of the process of drawing a three-dimensional obstacle map based on the direction and distance of obstacles in the present invention;

[0077] Figure 4 This is a flow chart of the present invention for determining whether a multi-degree-of-freedom robot can cross an obstacle based on the height of the obstacle and various parameters of the robot itself;

[0078] Figure 5 This is a flow chart of the controller scheduling the operating parameters of various components of the multi-degree-of-freedom robot to complete obstacle crossing based on the height of the obstacle in the present invention;

[0079] Figure 6 A schematic diagram of a flow chart of a detour route planned by the robot of the present invention;

[0080] Figure 7 A schematic diagram of the process of determining the size and shape of an object to be grasped based on image information from a visual sensor and distance judgment from an ultrasonic sensor according to the present invention;

[0081] Figure 8This is a flow chart of the processor determining the grasping point and force of the multi-degree-of-freedom robot based on the size and shape of the grasped object according to the present invention. DETAILED DESCRIPTION

[0082] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0083] Reference Figure 1 As shown, a multi-degree-of-freedom robot control method obtains image information from a visual sensor, and a processor analyzes the image information to obtain an image processing result;

[0084] Based on the image processing results, the direction of the obstacle is determined, and the distance between the obstacle and the multi-degree-of-freedom robot is determined using an ultrasonic sensor;

[0085] Draw a three-dimensional obstacle map based on the direction and distance of obstacles;

[0086] Based on the height of the obstacle and the robot's own parameters, determine whether the multi-degree-of-freedom robot can cross the obstacle;

[0087] If it can, the robot crosses the obstacle;

[0088] Based on the height of the obstacle, the controller adjusts the operating parameters of each component of the multi-degree-of-freedom robot to complete the obstacle crossing;

[0089] If not, the robot plans a detour route;

[0090] If there is a grasping task, the size and shape of the object to be grasped are determined based on the image information of the visual sensor and the distance judgment of the ultrasonic sensor;

[0091] Based on the size and shape of the object being grasped, the processor determines the grasping point and force of the multi-degree-of-freedom robot.

[0092] It can be explained that the determination of three-dimensional obstacles based on visual sensors and ultrasonic sensors provides a basis for the multi-degree-of-freedom robot to cross obstacles, and judges in advance whether the multi-degree-of-freedom robot can cross the obstacle to avoid energy waste caused by finding that it cannot cross the obstacle during the crossing process. The important basis for judging whether the multi-degree-of-freedom robot can cross the obstacle is whether the center of gravity will be unstable and cause it to fall during the crossing process, so as to avoid damage to the robot caused by falling during the crossing process, and prevent the grasped object from slipping by locating the grasping point.

[0093] Reference Figure 2 As shown, the image information of the visual sensor is obtained, and the processor analyzes the image information to obtain the image processing results including:

[0094] The multi-degree-of-freedom robot rotates its head vision sensor to obtain images from different perspectives;

[0095] Mark the horizontal line in the middle of the picture as the eye level;

[0096] Find at least two parallel lines in the real world, extend at least two parallel lines in the two-dimensional image, and record the intersection as the vanishing point, and the parallel lines as the vanishing lines;

[0097] If the first object on the same vanishing line is farther from the vanishing point than the second object, then the first object is behind the second object.

[0098] Get the height difference between the object and the horizon. If the height difference between the object and the horizon is negative, it means that the height of the object is lower than the object with a positive height difference.

[0099] The position relationship and height information of the objects are stored in the memory.

[0100] What can be explained is that the perspective method is used to determine the vanishing point in the picture. The so-called vanishing point refers to the point where parallel straight lines in the three-dimensional world intersect in the two-dimensional picture. The distance between the object and the vanishing point can be used to determine the distance relationship between the objects. Then, the characteristic that the horizon is parallel to the visual sensor and the height difference between the object and the horizon can be used to determine the height difference between the objects.

[0101] Reference Figure 3 As shown, drawing a 3D obstacle map based on the direction and distance of obstacles includes:

[0102] Select the object closest to the visual sensor in a straight line as the reference object;

[0103] Measure the distance from the vision sensor to the bottom and top of the reference object, and use the Pythagorean theorem to obtain the height of the reference object;

[0104] obtaining a height of the object by multiplying the height of the reference object by the relative height ratio based on a relative height ratio between at least one object in the picture and a reference object, wherein the relative height ratio is the height of the object in the picture from the horizon in the picture divided by the height of the reference object in the picture from the horizon in the picture;

[0105] Use the distance calculation formula to calculate the distance between objects;

[0106] The distance calculation formula is: ,

[0107] Where S is the distance between the first object and the second object, is the distance from the first reference object to the visual sensor, is the distance from the second reference object to the visual sensor, is the distance from the first object to the vanishing point, is the distance from the second object to the vanishing point, is the distance from the first reference object to the vanishing point, is the distance from the second reference object to the vanishing point.

[0108] The heights of all objects in the picture, the distances between objects, and the front-to-back relationships between objects are counted, and the objects are abstracted into geometric bodies to obtain a three-dimensional obstacle map.

[0109] It can be explained that the ultrasonic sensor and the top and bottom of the reference object form a right triangle. The distance from the ultrasonic sensor to the top and bottom of the reference object is known, and the height of the reference object can be calculated. Then, the height of the reference object and the relative height ratio of other objects to the reference object can be used to calculate the height of other objects. The distance between the objects can be calculated by using the fixed ratio of the distance difference between the two objects and the distance between the two objects on the screen. The distance between the objects can be calculated by the distance of the ultrasonic sensor, and the distance difference between the two objects on the screen can be calculated by the distance difference between the objects and the vanishing point.

[0110] Reference Figure 4 As shown in the figure, based on the height of the obstacle and the robot's own parameters, the following steps are used to determine whether the multi-degree-of-freedom robot can cross the obstacle:

[0111] Count the bending angles of the hip, knee, and ankle joints of normal humans when crossing obstacles and build a database;

[0112] Obtain the length from the hip joint to the knee joint and the length from the knee joint to the ankle joint when the multi-degree-of-freedom robot is standing normally;

[0113] Match the flexion angles of the hip, knee, and ankle joints of humans when crossing obstacles of the same height to the data related to the multi-degree-of-freedom robot legs in the database;

[0114] To analyze the physical center of gravity when the bending angle of each joint reaches the maximum, the steps are as follows:

[0115] The multi-degree-of-freedom robot is divided into two parts from the hip joint, and the physical center of gravity of the upper part and the geometric center of gravity of the upper part are on the same vertical line;

[0116] Divide the upper part into upper and lower halves along a horizontal line located in the middle of the vertical line of the upper part, and count the mass of the upper half and the mass of the lower half;

[0117] The physical center of gravity of the upper part is located at a preset distance from the middle point of the vertical line in the upper part. If the preset value is a positive number, it means that the physical center of gravity is above the middle point. If the preset value is a negative number, it means that the physical center of gravity is below the middle point.

[0118] The physical center of gravity of the thighs and calves on both sides of the multi-degree-of-freedom robot is obtained through the suspension method;

[0119] Obtain the three-dimensional coordinates of the physical center of gravity of the thighs and calves on both sides when the bending angle of each joint of the multi-degree-of-freedom robot reaches the maximum;

[0120] Use the center of gravity position formula to calculate the physical center of gravity coordinates of the lower part. The center of gravity position formula is as follows: ,

[0121] Where, is the physical center of gravity of the lower part, is the mass of the left thigh, is the mass of the right thigh, is the mass of the left calf, is the mass of the right calf, are the coordinates of the left thigh, are the coordinates of the right thigh, The coordinates of the left calf is the coordinate of the right calf;

[0122] Get the coordinates of the upper physical center of gravity. The sub-coordinates of the physical center of gravity of the multi-DOF robot are the sum of the sub-coordinates of the upper physical center of gravity multiplied by the mass of the upper part plus the sub-coordinates of the lower physical center of gravity multiplied by the mass of the lower part, divided by the overall mass of the multi-DOF robot.

[0123] Determine whether the physical center of gravity and the geometric center of gravity of the multi-degree-of-freedom robot are on the same vertical line. If so, the multi-degree-of-freedom robot can cross the obstacle;

[0124] The steps for obtaining the preset value are as follows:

[0125] The preset value is obtained using a distance formula, which is as follows: ,

[0126] Where A is the preset value, and the upper half mass is , the lower half mass is , the length of the vertical line in the upper part is 1.

[0127] It can be explained that the multi-degree-of-freedom robot is made with reference to the human body structure. When crossing obstacles, the multi-degree-of-freedom robot refers to the rotation angle of human joints, which helps to better cross obstacles. Whether the multi-degree-of-freedom robot can cross obstacles is not only determined by the maximum limit height being greater than the obstacle height, but also by whether the multi-degree-of-freedom robot will overturn during the crossing process. We judge the stability of the multi-degree-of-freedom robot by determining whether the physical center and geometric center of the multi-degree-of-freedom robot are on the same vertical line when each joint rotates to the maximum angle.

[0128] Reference Figure 5 As shown, based on the height of the obstacle, the controller schedules the operating parameters of each component of the multi-degree-of-freedom robot to complete the obstacle crossing, including:

[0129] Count the bending angles of the hip, knee, and ankle joints of normal humans when crossing obstacles and build a database;

[0130] Obtain the length from the hip joint to the knee joint and the length from the knee joint to the ankle joint when the multi-degree-of-freedom robot is standing normally;

[0131] Match the flexion angles of the hip, knee, and ankle joints of humans when crossing obstacles of the same height to the data related to the multi-degree-of-freedom robot legs in the database;

[0132] The controller adjusts the bending angle of each joint of the multi-degree-of-freedom robot according to the matching data.

[0133] What can be explained is that a database is established to count the bending angles of each joint of humans when crossing obstacles. The multi-degree-of-freedom robot is developed with reference to the human structure. When crossing obstacles, the movement of humans crossing obstacles is referred to to improve the success rate of crossing. By matching the length of the thigh and calf, the accuracy of the rotational joint is improved.

[0134] Reference Figure 6 As shown in the figure, the robot's planned detour route includes:

[0135] Get the height and maximum width of the multi-degree-of-freedom robot;

[0136] Based on the three-dimensional obstacle map, the entrances with a height greater than the height of the multi-degree-of-freedom robot and a width greater than the maximum width of the multi-degree-of-freedom robot are selected and recorded as candidate entrances;

[0137] The multi-degree-of-freedom robot preferentially chooses the rightmost alternative entrance to enter. After entering the entrance, it determines whether the multi-degree-of-freedom robot can cross the obstacle it encounters again. If so, it crosses the obstacle it encounters again. If not, it plans a detour route for the obstacle it encounters again until the multi-degree-of-freedom robot reaches the working position.

[0138] What can be explained is that first, the accessible entrances are screened out as the candidate entrances, and then the right entrance is taken regularly. If the next route cannot be passed, return to the previous intersection and try the next entrance again. Regularly try each possibility to ensure that no possibility is missed, realize the traversal of all possible options, and finally find the correct route.

[0139] Reference Figure 7 As shown, if there is a grasping task, the size and shape of the object to be grasped are determined based on the image information of the visual sensor and the distance judgment of the ultrasonic sensor, including:

[0140] Using an ultrasonic sensor to obtain a distance and a direction between the ultrasonic sensor and at least one point on the grasped object;

[0141] In the simulation system, projecting the position of at least one point according to the distance direction of the at least one point;

[0142] Based on the distribution of the point graph, connect at least one point to construct a surface to obtain the basic model for grasping the object;

[0143] Based on the image information of the visual sensor, artificial intelligence is used to search and compare to determine the type of objects in the image;

[0144] Based on the type of grasped object, the surface roughness is refined on the basis of the basic model to obtain the surface friction coefficient of the grasped object;

[0145] Based on the type of grasped object, the density of the grasped object is determined, the volume of the object is determined through image recognition, and the gravity of the object is calculated based on the volume and density of the object.

[0146] What can be explained is that the direction and distance of the points on the grasped object are determined by the ultrasonic sensor, a point diagram is drawn, and the basic model of the grasped object is determined. The type of grasped object is determined by the visual sensor, and the surface friction coefficient and density of the grasped object are further determined. The gravity of the grasped object is calculated, which facilitates the determination of the grasping force.

[0147] Reference Figure 8 As shown, based on the size and shape of the grasped object, the processor determines the grasping point and force of the multi-degree-of-freedom robot including:

[0148] At least one sampling point is uniformly selected on the surface of the object, a tangent plane is made on the surface of the object at the sampling point, and the angle between the tangent plane corresponding to the sampling point and the vertical direction is taken as the relative angle;

[0149] The number of fingers of the multi-degree-of-freedom robot is used as the eigenvalue;

[0150] The object's gravity is divided by the friction coefficient of the grasping point and then divided by the characteristic value to obtain the finger grasping force;

[0151] According to the type of object, the tolerable pressure of the sampling point of the object is obtained;

[0152] The sampling point that can withstand a pressure greater than the finger grasping force is used as the target sampling point;

[0153] Randomly combining target sampling points to form at least one target sampling point combination, where the number of sampling points included in the target sampling point combination is equal to the eigenvalue;

[0154] The target sampling points that are coplanar with the target sampling points are combined as the calibration sampling point combination;

[0155] The closed plane area formed by sequentially connecting the adjacent target sampling points in the calibration sampling point combination is used as the plane sampling area;

[0156] The area through which the plane sampling area moves in the vertical direction is used as the stereo projection range of the calibration sampling point combination;

[0157] Obtain the center of gravity of the object, and use the calibration sampling point combination corresponding to the stereoscopic projection range containing the center of gravity of the object as the preliminary sampling point combination;

[0158] The surface of the object located above the plane where the plane sampling area is located is taken as the feature surface;

[0159] Selecting a characteristic sampling point combination from at least one calibration sampling point combination, so that the area of ​​the characteristic surface generated by the plane sampling region corresponding to the characteristic sampling point combination is the smallest;

[0160] Use the sampling points in the feature sampling point combination as the grabbing points.

[0161] It can be explained that the friction force formed by the pressure applied by each finger jointly supports the grasped object and maintains it in the hand. The bearing capacity of the sampling point needs to be greater than the grasping force to prevent damage to the grasped object. The smaller the angle between the tangent plane corresponding to the sampling point and the vertical direction, the greater the vertical component of the friction force caused by the pressure applied by the finger, and the smaller the force required to be applied, which is convenient for saving energy and ensuring that the center is within the range of the three-dimensional projection to ensure that the grasped object is not easy to slip.

[0162] Furthermore, the present solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is called, the above-mentioned multi-degree-of-freedom robot control method and method are executed.

[0163] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state disk (SSD).

[0164] To sum up, the advantages of the present invention are: determining three-dimensional obstacles based on visual sensors and ultrasonic sensors, providing a basis for multi-degree-of-freedom robots to cross obstacles, judging in advance whether the multi-degree-of-freedom robot can cross obstacles, avoiding energy waste caused by finding that it cannot cross during the crossing process, and an important basis for judging whether the multi-degree-of-freedom robot can cross obstacles is whether the center of gravity is unstable and causes falls during the crossing process, avoiding damage to the robot caused by falls during the crossing process, and preventing the grasped objects from slipping by locating the grasping points.

[0165] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-degree-of-freedom robot control method, characterized in that: include: Acquire the image information from the visual sensor, and the processor analyzes the image information to obtain the image processing results; Based on the image processing results, the direction of the obstacle is determined, and the distance between the obstacle and the multi-degree-of-freedom robot is determined using an ultrasonic sensor; Draw a three-dimensional obstacle map based on the direction and distance of obstacles; Based on the height of the obstacle and the robot's own parameters, determine whether the multi-degree-of-freedom robot can cross the obstacle; If it can, the robot crosses the obstacle; Based on the height of the obstacle, the controller adjusts the operating parameters of each component of the multi-degree-of-freedom robot to complete the obstacle crossing; If not, the robot plans a detour route; If there is a grasping task, the size and shape of the object to be grasped are determined based on the image information of the visual sensor and the distance judgment of the ultrasonic sensor; Based on the size and shape of the object being grasped, the processor determines the grasping point and force of the multi-degree-of-freedom robot; The processor determines the grasping point and force of the multi-degree-of-freedom robot based on the size and shape of the grasped object, including: At least one sampling point is uniformly selected on the surface of the object, a tangent plane is made on the surface of the object at the sampling point, and the angle between the tangent plane corresponding to the sampling point and the vertical direction is taken as the relative angle; The number of fingers of the multi-degree-of-freedom robot is used as the eigenvalue; The object's gravity is divided by the friction coefficient of the grasping point and then divided by the characteristic value to obtain the finger grasping force; According to the type of object, the tolerable pressure of the sampling point of the object is obtained; The sampling point that can withstand a pressure greater than the finger grasping force is used as the target sampling point; Randomly combining target sampling points to form at least one target sampling point combination, where the number of sampling points included in the target sampling point combination is equal to the eigenvalue; The target sampling points that are coplanar with the target sampling points are combined as the calibration sampling point combination; The closed plane area formed by sequentially connecting the adjacent target sampling points in the calibration sampling point combination is used as the plane sampling area; The area through which the plane sampling area moves in the vertical direction is used as the stereo projection range of the calibration sampling point combination; Obtain the center of gravity of the object, and use the calibration sampling point combination corresponding to the stereoscopic projection range containing the center of gravity of the object as the preliminary sampling point combination; The surface of the object located above the plane where the plane sampling area is located is taken as the feature surface; Selecting a characteristic sampling point combination from at least one calibration sampling point combination, so that the area of ​​the characteristic surface generated by the plane sampling region corresponding to the characteristic sampling point combination is the smallest; Use the sampling points in the characteristic sampling point combination as the grasping points; The determination of whether the multi-degree-of-freedom robot can cross the obstacle based on the height of the obstacle and various parameters of the robot itself includes: Count the bending angles of the hip, knee, and ankle joints of normal humans when crossing obstacles and build a database; Obtain the length from the hip joint to the knee joint and the length from the knee joint to the ankle joint when the multi-degree-of-freedom robot is standing normally; Match the flexion angles of the hip, knee, and ankle joints of humans when crossing obstacles of the same height to the data related to the multi-degree-of-freedom robot legs in the database; To analyze the physical center of gravity when the bending angle of each joint reaches the maximum, the steps are as follows: The multi-degree-of-freedom robot is divided into two parts from the hip joint, and the physical center of gravity of the upper part and the geometric center of gravity of the upper part are on the same vertical line; Divide the upper part into upper and lower halves along a horizontal line located in the middle of the vertical line of the upper part, and count the mass of the upper half and the mass of the lower half; The physical center of gravity of the upper part is located at a preset distance from the middle point of the vertical line in the upper part. If the preset value is a positive number, it means that the physical center of gravity is above the middle point. If the preset value is a negative number, it means that the physical center of gravity is below the middle point. The physical center of gravity of the thighs and calves on both sides of the multi-degree-of-freedom robot is obtained through the suspension method; Obtain the three-dimensional coordinates of the physical center of gravity of the thighs and calves on both sides when the bending angle of each joint of the multi-degree-of-freedom robot reaches the maximum; Use the center of gravity position formula to calculate the physical center of gravity coordinates of the lower part. The center of gravity position formula is as follows: ， Where, is the physical center of gravity of the lower part, is the mass of the left thigh, is the mass of the right thigh, is the mass of the left calf, is the mass of the right calf, are the coordinates of the left thigh, are the coordinates of the right thigh, The coordinates of the left calf is the coordinate of the right calf; Get the coordinates of the upper physical center of gravity. The sub-coordinates of the physical center of gravity of the multi-DOF robot are the sum of the sub-coordinates of the upper physical center of gravity multiplied by the mass of the upper part plus the sub-coordinates of the lower physical center of gravity multiplied by the mass of the lower part, divided by the overall mass of the multi-DOF robot. Determine whether the physical center of gravity and the geometric center of gravity of the multi-degree-of-freedom robot are on the same vertical line. If so, the multi-degree-of-freedom robot can cross the obstacle; The steps for obtaining the preset value are as follows: The preset value is obtained using a distance formula, which is as follows: ， Where A is the preset value, and the upper half mass is , the lower half mass is , the length of the vertical line in the upper part is 1.

2. A multi-degree-of-freedom robot control method according to claim 1, characterized in that: The acquiring of image information from the visual sensor, the processor analyzing the image information, and obtaining an image processing result include: The multi-degree-of-freedom robot rotates its head vision sensor to obtain images from different perspectives; Mark the horizontal line in the middle of the picture as the eye level; Find at least two parallel lines in the real world, extend at least two parallel lines in the two-dimensional image, and record the intersection as the vanishing point, and the parallel lines as the vanishing lines; If the first object on the same vanishing line is farther from the vanishing point than the second object, then the first object is behind the second object. Get the height difference between the object and the horizon. If the height difference between the object and the horizon is negative, it means that the height of the object is lower than the object with a positive height difference. The position relationship and height information of the objects are stored in the memory.

3. A multi-degree-of-freedom robot control method according to claim 2, characterized in that: Drawing a three-dimensional obstacle map based on the direction and distance of the obstacle includes: Select the object closest to the visual sensor in a straight line as the reference object; Measure the distance from the vision sensor to the bottom and top of the reference object, and use the Pythagorean theorem to obtain the height of the reference object; obtaining a height of the object by multiplying the height of the reference object by the relative height ratio based on a relative height ratio between at least one object in the picture and a reference object, wherein the relative height ratio is the height of the object in the picture from the horizon in the picture divided by the height of the reference object in the picture from the horizon in the picture; Use the distance calculation formula to calculate the distance between objects; The distance calculation formula is: ， Where S is the distance between the first object and the second object, is the distance from the first reference object to the visual sensor, is the distance from the second reference object to the visual sensor, is the distance from the first object to the vanishing point, is the distance from the second object to the vanishing point, is the distance from the first reference object to the vanishing point, is the distance from the second reference object to the vanishing point; The heights of all objects in the picture, the distances between objects, and the front-to-back relationships between objects are counted, and the objects are abstracted into geometric bodies to obtain a three-dimensional obstacle map.

4. A multi-degree-of-freedom robot control method according to claim 3, characterized in that: Based on the height of the obstacle, the controller schedules the operating parameters of each component of the multi-degree-of-freedom robot to complete the obstacle crossing, including: Count the bending angles of the hip, knee, and ankle joints of normal humans when crossing obstacles and build a database; Obtain the length from the hip joint to the knee joint and the length from the knee joint to the ankle joint when the multi-degree-of-freedom robot is standing normally; Match the flexion angles of the hip, knee, and ankle joints of humans when crossing obstacles of the same height to the data related to the multi-degree-of-freedom robot legs in the database; The controller adjusts the bending angle of each joint of the multi-degree-of-freedom robot according to the matching data.

5. A multi-degree-of-freedom robot control method according to claim 4, characterized in that: The robot's planned detour route includes: Get the height and maximum width of the multi-degree-of-freedom robot; Based on the three-dimensional obstacle map, the entrances with a height greater than the height of the multi-degree-of-freedom robot and a width greater than the maximum width of the multi-degree-of-freedom robot are selected and recorded as candidate entrances; The multi-degree-of-freedom robot preferentially chooses the rightmost alternative entrance to enter. After entering the entrance, it determines whether the multi-degree-of-freedom robot can cross the obstacle it encounters again. If so, it crosses the obstacle it encounters again. If not, it plans a detour route for the obstacle it encounters again until the multi-degree-of-freedom robot reaches the working position.

6. A multi-degree-of-freedom robot control method according to claim 5, characterized in that: If there is a grasping task, the size and shape of the object to be grasped are determined based on the image information of the visual sensor and the distance judgment of the ultrasonic sensor, including: Using an ultrasonic sensor to obtain a distance and a direction between the ultrasonic sensor and at least one point on the grasped object; In the simulation system, projecting the position of at least one point according to the distance direction of the at least one point; Based on the distribution of the point graph, connect at least one point to construct a surface to obtain the basic model for grasping the object; Based on the image information of the visual sensor, artificial intelligence is used to search and compare to determine the type of objects in the image; Based on the type of grasped object, the surface roughness is refined on the basis of the basic model to obtain the surface friction coefficient of the grasped object; Based on the type of grasped object, the density of the grasped object is determined, the volume of the object is determined through image recognition, and the gravity of the object is calculated based on the volume and density of the object.

7. A multi-degree-of-freedom robot control system, used to implement a multi-degree-of-freedom robot control method according to any one of claims 1 to 6, characterized in that: include: An obstacle model building module, which determines the height and distance of obstacles based on visual sensors and ultrasonic sensors and builds a three-dimensional obstacle map; an obstacle crossing determination module, which determines whether the multi-degree-of-freedom robot can cross the obstacle based on the height of the obstacle and the multi-degree-of-freedom robot's own data; A route detour module, wherein the route detour module provides a detour route for the multi-degree-of-freedom robot; The object grasping module determines the bending degree and grasping force of the multi-degree-of-freedom robot's finger joints when grasping the object based on the size and shape of the object.

Images