A robot picking device and method based on proximity perception

Abstract

The present invention discloses a robot picking device and method based on proximity perception, which belongs to the field of robot sensing and control technology, including a supporting disc, a light emitting array, a light receiving device, a transparent rubber diaphragm, a pneumatic suction cup array and a signal acquisition board; the picking method includes the following steps: hand-eye calibration of a multi-degree-of-freedom robot and initialization of a picking device; collecting images through a monocular camera and uploading them to an intelligent terminal for processing to obtain the coordinates of a target object; the intelligent terminal issues a robot movement instruction until the distance threshold is maintained with the approaching object, and then issues an instruction to adjust the robot posture so that the pneumatic suction cup at the end of the robot arm remains parallel to the target object picking surface. The proximity picking measurement of the present invention has high precision and fast response speed, and can complete posture adjustment on the basis of approaching and not contacting the target object, which greatly reduces the difficulty of picking up thin, fragile and other valuable items in a cluttered environment, and increases the probability of successfully picking up unknown objects.

Description

Technical Field

[0001] The present invention relates to the field of robot sensing and control technology, and in particular to a robot picking device and method based on proximity perception. Background Art

[0002] With the development of science and technology, robot operation has become an important means to replace human labor. From parcel sorting in logistics centers to workpiece picking in manufacturing plants, from household messy items to military explosives dismantling, robot operation plays an important role. Among them, robot picking up objects is one of the most common tasks.

[0003] At present, the positioning technology based on vision guidance is the main means for robots to obtain information and execute grasping strategies. For example, CN112288819B discloses a multi-source data fusion vision-guided robot grasping method, which determines the initial area of ​​the object and calculates the object posture from the 3D point cloud to guide the robot body to grasp the target part; CN108491880B discloses a method for obtaining the position and posture of the target part by using a neural network method of multi-view images of the object for grasping by an intelligent robotic arm.

[0004] However, the first technology mentioned above guides the robotic arm to grasp the target part by collecting 3D point cloud information, which not only requires expensive deep vision equipment but also involves complex visual calibration and grasping algorithms; the second technology mentioned above learns the position and posture of the current target part through multi-view images of the object, which not only requires a large amount of training data but also makes it difficult to estimate the optimal posture of the new target part. Summary of the invention

[0005] In view of the shortcomings of existing grasping technologies, the present invention provides a robot picking device and method based on proximity perception. Specifically, the present invention discloses a picking device that does not rely on contact with an object to adjust the posture of the robot body and a method of using the picking device to effectively adjust the robot body and absorb the target object. The device and method reduce the difficulty of picking up thin, fragile and other valuable items in a cluttered environment, and increase the probability of successfully picking up unknown objects; it realizes the adjustment of the robot posture based on approaching but not contacting the object to complete the picking task.

[0006] The technical solution adopted by the present invention is: a picking method of a robot picking device based on proximity perception, comprising the following steps:

[0007] Step 1: Robot hand-eye calibration operation and picking device initialization;

[0008] Step 2: Obtain the approximate coordinate position of the target part relative to the robot base coordinate system;

[0009] Step 3: The robot picking device moves relative to the target part until the approach distance is equal to the distance threshold;

[0010] Step 4: Issue a command to adjust the robot picking device so that the pneumatic suction cup array at the end of the robot remains parallel to the picking surface of the target object, and continuously detect whether the approach distance is equal to the distance threshold. If the condition is not met, return to step 3;

[0011] Step 5: The robot moves slowly along the normal direction of the target object's picking surface until the pneumatic suction cup array fits the target object, completing the picking task;

[0012] The robot hand-eye calibration operation in step 1 uses the camera calibration toolkit in Opencv or matlab software to solve the position transformation relationship of the target object in the robot user coordinate system and the base coordinate system;

[0013] In step 2, the center coordinates P of the target part relative to the robot user coordinate system are obtained by using a monocular camera. u , and then through P b =T b2u ·P u The formula calculates the coordinate position of the target part relative to the robot base coordinate system, where T b2u The transformation matrix of the established base coordinate system relative to the robot user coordinate system;

[0014] In step 3, the signal acquisition board detects the voltage ratio of the light receiving device on the inner circle with a radius of r1 and the light emitting device on the outer circle with a radius of r2, which is offset by 90° phase. To detect the approach distance value d:

[0015]

[0016] Where K is the light intensity ratio of light emitting devices with different phases;

[0017] In step 4, the signal acquisition board detects the voltage ratio of the light receiving device collected by the light emitting device on the left and right sides with a radius of r2 and a phase shift of 90°. To detect the approach tilt angle θ:

[0018]

[0019] Where K′ is the light intensity ratio of light emitting devices with different phases;

[0020] The pickup device includes a supporting disc, a light receiving device, a light emitting array, a transparent rubber diaphragm, a pneumatic suction cup array and a signal acquisition board;

[0021] The support disk serves as the base of the pickup device. Above it, there are a light receiving device, a light emitting array, and a pneumatic suction cup array being towed, and a signal acquisition board is embedded in the middle.

[0022] The light receiving device is mounted at the center point above the support disk and can be used to receive the light intensity signal reflected by the light emitting array.

[0023] The light emitting array is centered on the center point of the disk, with radii r1 and r2 respectively (where r1 < r2), and a total of m light emitters are evenly placed in the horizontal and vertical directions.

[0024] The light emitter is driven by a modulated square wave signal.

[0025] The phase of the modulation signal of the light emitter is adjustable from 0° to 90°.

[0026] The transparent rubber diaphragm is located on the surfaces of the light receiving device and the light emitting array and is flush with the lower end of the micro suction nozzle.

[0027] The pneumatic suction cup array includes a vacuum pump and micro suction nozzles.

[0028] The pneumatic suction cup array is mounted above the support disk, centered on the center point of the disk, with a radius of r3 (where r3 > r2), and a total of n micro suction nozzles are placed in the horizontal and vertical directions.

[0029] Preferably, the support disk is made of a hard insulating material, and hanging hooks are provided on both sides of the surface.

[0030] Preferably, the transparent rubber diaphragm has light transmittance and stretchability.

[0031] Preferably, the signal acquisition board includes a lock-in amplifier unit, a low-pass filter unit, a high-speed and high-precision AD sampling unit, a signal calculation unit, and a signal transmission interface.

[0032] Preferably, in steps three and four, the inclination angles θ x and θ y of the target object in the x-axis and y-axis directions in the robot user coordinate system are calculated respectively on the x-axis and y-axis of the pickup device, and then the required rotational angle increments Δθ′ x and Δθ′ y around the x and y directions relative to the base coordinate system in the robot user coordinate system are calculated;

[0033] In step five, the offset of the target object relative to the robot user coordinate system is calculated, so as to obtain the motion increment of the robot relative to the base coordinate system, and make the pneumatic suction cup array fit the surface of the target object.

[0034] Compared with the existing devices and technologies, the beneficial effects of the present invention are as follows:

[0035] (1) The pickup device of the present invention is light and cheap, and can be easily mounted on a robot; and has high measurement accuracy for the distance and tilt angle of the target object, fast response speed, and is not affected by the reflectivity of the object surface.

[0036] (2) The present invention ignores the depth information of the object and the reasoning of the three-dimensional object, and can determine the precise position and posture of the target object with the help of the proximity sensor; thus reducing the required hardware cost and the complexity of the object's spatial position estimation.

[0037] (3) The present invention uses a pneumatic suction cup array to pick up the target object, allowing the object to be scattered and tilted in a cluttered environment. The target object picking process is simple and efficient, and can increase the success rate of picking up unknown objects.

[0038] (4) The present invention can complete the posture adjustment of the picking device while approaching but not contacting the object; for picking up thin, fragile and valuable objects, it has unique advantages compared with the traditional contact posture adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the structure of the pickup device of the present invention;

[0040] Figure 2 This is a schematic diagram of the proximity sensing and pneumatic suction cup picking up a target object of the present invention;

[0041] Figure 3 This is a flow chart of the picking control method of the present invention;

[0042] Figure 4 This is a schematic diagram of a specific implementation example of the present invention.

[0043] Among them, 1-fixed multi-degree-of-freedom robot, 2-monocular camera, 3-proximity sensing component, 31-support disc, 32-light receiving device, 33-gasket, 34-light emitting array (341-348 are light receiving devices 1 to 8), 35-transparent rubber diaphragm, 4-pneumatic suction cup, 41-vacuum pump, 42-micro suction nozzle (421-424 are micro suction nozzle 1 to micro suction nozzle 4), 5-signal acquisition board, 6-network port, 7-intelligent terminal, 8-target object DETAILED DESCRIPTION

[0044] In order to make the technical solutions, methods and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0045] In view of the problems in the prior art that robot grasping involves complex visual calibration and object pose reasoning, and has poor adaptability to objects placed in different ways, the present invention provides a picking device that does not rely on contacting the object to adjust the pose of the robot body, and a method for effectively adjusting the robot body and sucking the target part by using the picking device. The device and method reduce the difficulty of picking light, fragile and other valuable items in a messy environment, and enhance the success probability of picking unknown objects; realize the function of adjusting the robot pose on the basis of approaching but not contacting the object to complete the picking task.

[0046] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0047] As Figure 1 、 2 、shown in FIG. 4, a robot picking device based on proximity sensing includes: a support disk 31, a light receiving device 32, a light emitting array 34, a transparent rubber diaphragm 35, a pneumatic suction cup array 4, and a signal acquisition board 5; the support disk 31 serves as the base of the picking device, and the light emitting array 34, the light receiving device 32, and the pneumatic suction cup array 4 are carried above it, and the signal acquisition board 5 is embedded in the middle; the transparent rubber diaphragm 35 is located on the surfaces of the light receiving device 32 and the light emitting array 34, and is flush with the lower end of the micro suction nozzle 42;

[0048] The support disk 31 is made of a hard insulating material, and hanging buckles are designed on both sides of the surface for upper and lower fixed connection;

[0049] The light receiving device 32 is mounted at the center point above the support disk 31 and can be used to receive the reflected light intensity of the light emitting array 34;

[0050] The light emitting array 34 is mounted above the support disk 31. Taking the center point of the disk as the center of the circle, m (m = 8) light emitters (341-348) are placed horizontally and vertically in sequence with radii of r1 and r2 (r1 < r2); the light emitters (341-348) are driven by a 30K hz modulated square wave signal; the phase of the modulation signal of the light emitters (341-348) is adjustable from 0° to 90°.

[0051] The transparent rubber diaphragm 35 has good light transmittance and stretchability and can be used to protect the proximity sensing component;

[0052] The pneumatic suction cup array 4 is mounted above the support disk 31. Taking the center point of the disk as the center of the circle, with a radius of r3 (r3 > r2), n (n = 4) micro suction nozzles (421-424) are placed horizontally and vertically; the pneumatic suction cup array 4 includes a vacuum pump 41 and micro suction nozzles (421-424);

[0053] The signal acquisition board 5 includes a phase-locked amplifier unit, a low-pass filter unit, a high-speed and high-precision AD sampling unit, a signal calculation unit, and a signal transmission interface;

[0054] Reference Figure 4 , shows a schematic diagram of an embodiment of the present invention. Figure 3 This is a flowchart of a robot picking process based on proximity perception of the present invention, comprising the following steps:

[0055] Step 1: Robot hand-eye calibration operation and picking device initialization

[0056] Step 2: Obtain the approximate coordinate position of the target part relative to the robot base coordinate system;

[0057] Step 3: The robot picking device moves relative to the target part until the approach distance is equal to the distance threshold;

[0058] Step 4: Issue a command to adjust the robot picking device so that the pneumatic suction cup array at the end of the robot remains parallel to the picking surface of the target object; at the same time, continuously detect whether the approach distance is equal to the distance threshold condition, and if the condition is not met, return to step 3;

[0059] Step 5: The robot moves slowly along the normal direction of the target object's picking surface until the pneumatic suction cup array fits the target object, completing the picking task;

[0060] Preferably, the robot hand-eye calibration operation in step 1 mainly utilizes the camera calibration toolkit in Opencv or matlab software to solve the position transformation relationship of the target object in the robot user coordinate system and the base coordinate system.

[0061] Preferably, the step 2 can obtain the target part center coordinate P relative to the robot user coordinate system through a monocular camera. u , and then through P b =T b2u ·P u The formula calculates the approximate coordinate position of the target part relative to the robot base coordinate system.

[0062] Where T b2u It is the transformation matrix of the established base coordinate system relative to the robot user coordinate system.

[0063] Preferably, in step 3, the signal acquisition board can detect the voltage ratio of the light receiving device on the inner circle with a radius of r1 and the light emitting device on the outer circle with a radius of r2, which is offset by 90° phase. To detect the distance value d:

[0064]

[0065] Where K is the light intensity ratio of light emitting devices with different phases;

[0066] Preferably, in step 4, the signal acquisition board can detect the voltage ratio of the left and right light emitting devices with a radius of r2 and a phase shift of 90° collected by the light receiving device. To detect the approach tilt angle θ:

[0067]

[0068] Where K′ is the light intensity ratio of light emitting devices with different phases;

[0069] Preferably, in order to simplify calculation and take loss into consideration, the light intensity ratio K and K′ is designed to be 1.

[0070] Preferably, by performing step 3 and step 4 on the x-axis and y-axis of the picking device respectively, the inclination angle θ of the target object in the x-axis and y-axis directions in the robot user coordinate system can be calculated. x and θ y ; Furthermore, the required rotation angle increment Δθ′ of the robot user coordinate system relative to the base coordinate system around the x and y directions can be calculated x and Δθ′ y ;

[0071] Preferably, the step five calculates the offset of the target object relative to the robot user coordinate system, thereby obtaining the motion increment of the robot relative to the base coordinate system, so that the pneumatic suction cup array fits the surface of the target object.

[0072] Compared with existing devices and technologies, the present invention has the following beneficial effects:

[0073] The pickup device of the present invention is light and cheap, and can be easily mounted on a robot; and has high measurement accuracy for the distance and tilt angle of the target object, fast response speed, and is not affected by the reflectivity of the object surface.

[0074] The present invention ignores the depth information of the object and the reasoning of the three-dimensional object, and can determine the precise position and posture of the target object with the help of a proximity sensor; thereby reducing the required hardware cost and the complexity of estimating the spatial position of the object.

[0075] The present invention uses a pneumatic suction cup array to pick up the target object, allowing the object to be scattered and tilted in a cluttered environment. The target object picking process is simple and efficient, and can increase the success rate of picking up unknown objects.

[0076] The present invention can complete the posture adjustment of the picking device under the condition of approaching but not contacting the object; for picking up thin, fragile and valuable objects, it has unique advantages compared with the traditional contact posture adjustment.

[0077] The specific embodiments of the present invention are described in detail above, but they are only examples, and the present invention is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications and substitutions made to the present invention are also within the scope of the present invention. Therefore, the equalization changes and modifications made without departing from the spirit and scope of the present invention are all within the scope of the present invention.

Claims

1. A robot picking method based on proximity perception, characterized in that: The steps include: Step 1: Robot hand-eye calibration operation and picking device initialization; Step 2: Obtain the coordinate position of the target part relative to the robot base coordinate system; Step 3: The robot picking device moves relative to the target part until the approach distance is equal to the distance threshold; Step 4: Issue a command to adjust the robot picking device so that the pneumatic suction cup array at the end of the robot remains parallel to the picking surface of the target object, and continuously detect whether the approach distance is equal to the distance threshold. If the condition is not met, return to step 3; Step 5: The robot moves slowly along the normal direction of the target object's picking surface until the pneumatic suction cup array fits the target object, completing the picking task; The robot hand-eye calibration operation in step 1 uses the camera calibration toolkit in Opencv or matlab software to solve the position transformation relationship of the target object in the robot user coordinate system and the base coordinate system; In step 2, the center coordinates P of the target part relative to the robot user coordinate system are obtained by using a monocular camera. u , and then through P b =T b2u ·P u The formula calculates the coordinate position of the target part relative to the robot base coordinate system, where T b2u The transformation matrix of the established base coordinate system relative to the robot user coordinate system; In step 3, the signal acquisition board detects the voltage ratio of the light receiving device on the inner circle with a radius of r1 and the light emitting device on the outer circle with a radius of r2, which is offset by 90° phase. To detect the approach distance value d: Where K is the light intensity ratio of light emitting devices with different phases; In step 4, the signal acquisition board detects the voltage ratio of the light receiving device collected by the light emitting device on the left and right sides with a radius of r2 and a phase shift of 90°. To detect the approaching tilt angle θ: Where K' is the light intensity ratio of light emitting devices with different phases; The pickup device includes a supporting disc, a light receiving device, a light emitting array, a transparent rubber diaphragm, a pneumatic suction cup array and a signal acquisition board; The support disc serves as the base of the pickup device, on which the light receiving device, the light emitting array, and the pneumatic suction cup array are towed, and a signal collection board is embedded in the middle; The light receiving device is mounted at the center point above the supporting disk and can be used to receive the light intensity signal reflected by the light emitting array; The light emitting array takes the center point of the disk as the center of the circle, takes r1 and r2 as the radius respectively, and r1<r2, and m light emitters are evenly placed in the horizontal and vertical directions; The optical transmitter is driven by a modulated square wave signal; The modulation signal phase of the optical transmitter is adjustable from 0° to 90°; The transparent rubber diaphragm is located on the surface of the light receiving device and the light emitting array, and is flush with the lower end of the micro nozzle; The pneumatic suction cup array includes a vacuum pump and a micro suction nozzle; The pneumatic suction cup array is mounted above the supporting disc, with the center point of the disc as the center of the circle, r3 as the radius, and r3>r2, and n micro suction nozzles are placed in the horizontal and vertical directions.

2. A robot picking method based on proximity perception according to claim 1, characterized in that: The supporting disc is made of hard insulating material, and hooks are arranged on both sides of the surface.

3. The robot picking method based on proximity perception according to claim 1 is characterized in that: The transparent rubber diaphragm has light transmittance and stretchability.

4. The robot picking method based on proximity perception according to claim 1 is characterized in that: The signal acquisition board includes a phase-locked amplifier unit, a low-pass filter unit, a high-speed and high-precision AD sampling unit, a signal calculation unit, and a signal transmission interface.

5. The robot picking method based on proximity perception according to claim 1 is characterized in that: In step 3 and step 4, the inclination angle θ of the target object in the x-axis and y-axis directions in the robot user coordinate system is calculated on the x-axis and y-axis of the picking device respectively. x and θ y , and then calculate the rotation angle increment Δθ′ required by the robot user coordinate system relative to the base coordinate system around the x and y directions x and Δθ′ y ; The step five calculates the offset of the target object relative to the robot user coordinate system, thereby obtaining the motion increment of the robot relative to the base coordinate system, so that the pneumatic suction cup array fits the surface of the target object.

Images