CN114193488A - A Flexible Adaptive Tactile Sensor, Gripper Finger and Mechanical Gripper - Google Patents

Abstract

The invention discloses a flexible and adaptive tactile sensor, a clamping finger and a mechanical claw. The tactile sensor includes a pressure sensor and a flexible circuit board. The flexible circuit board includes a middle jaw and two side jaws. The two side claws are respectively arranged on both sides of the middle claws. The inner ends of the two side claws are connected with the inner ends of the middle claws. The middle claws are straight; the two side claws are curved, and the outer ends are both bent to the side away from the middle claws. A plurality of pressure sensors arranged in sequence are installed on the middle jaw and the two side jaws. In use, the middle jaw and the two side jaws wrap around the spherically shaped gripping finger end surfaces. The distribution mode of the three-dimensional fingertip pressure sensor array in the present invention can structurally detect the reaction force distribution of the end clamping structure. In addition, the present invention uses the number of pressure sensors whose detection value exceeds the threshold to detect the deformation of the object when it is squeezed, so as to identify the type of the object being clamped.

Description

A Flexible Adaptive Tactile Sensor, Gripper Finger and Mechanical Gripper

technical field

The invention belongs to the technical field of object type recognition, in particular to a flexible and adaptive tactile sensor, a clamping finger and a mechanical claw.

Design a structure and method for adaptively grasping soft objects with different degrees of deformation, specifically design a tactile sensor based on multiple pressure sensor arrays, and an adaptive grasping system based on multi-tactile information, It can adaptively grasp flexible and deformable objects and ensure that the applied force is the minimum force required, and the deformation of the object is minimized.

Background technique

The robot hand at the end of the manipulator is usually used to perform various operation tasks. When performing the operation task, the manipulator physically interacts with the object. That is, when the manipulator grasps the object, the end effector is the core mechanism of the grasping ability of the manipulator. When the robot hand performs different grasping tasks, the grasping planner of the system designs grasping strategies in advance to effectively grasp the target object. At present, most of the grasping robots use the visual information provided by the vision system to identify the target object and use it as an important sensory information source for identifying certain objects to grasp the objects. However, intangible intrinsic properties of objects, such as softness, weight, fragility, and surface roughness, cannot be correctly identified using visual information alone. Therefore, the present application proposes a grasping system and a tactile sensor structure thereof to adaptively grasp a flexible and deformable object and ensure that the applied force is the minimum required force and the deformation of the object is minimized.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a flexible and adaptive tactile sensor, a clamping finger and a mechanical claw.

In a first aspect, the present invention provides a flexible and adaptive tactile sensor, including a pressure sensor and a flexible circuit board. The flexible circuit board includes a middle jaw and two side jaws. The two side claws are respectively arranged on both sides of the middle claws. The inner ends of the two side claws are connected with the inner ends of the middle claws. The middle claws are linear; the two side claws are curved, and the outer ends are both bent to the side away from the middle claws. A plurality of pressure sensors arranged in sequence are installed on the middle jaw and the two side jaws. In use, the middle jaw and the two side jaws wrap around the spherically shaped gripping finger end surfaces.

Preferably, the inner end of the side claw is tangent to the inner end of the middle claw.

其中，r₁为夹持指末端表面的球半径；r₂为夹持指末端表面两侧圆弧边缘的半径。d为侧爪的宽度。Preferably, the end surface of the clamping finger is a partial spherical surface with symmetrical circular arc edges on both sides. The arc of the side claw is

Among them, r ₁ is the spherical radius of the end surface of the clamping finger; r ₂ is the radius of the arc edges on both sides of the end surface of the clamping finger. d is the width of the lateral claw.

The radii of the two side claws are R ₁ and R2 respectively. in,

Preferably, the flexible circuit board further includes a base. The base and the inner end of the middle claw are connected by a connecting segment.

Preferably, a control module is installed on the flexible circuit board; each pressure sensor is connected to the control module.

In a second aspect, the present invention provides a gripping finger based on tactile detection, which includes a finger end base, an elastic filling layer and the aforementioned tactile sensor. The ends of the finger bases are partially spherical. The middle claw and the two side claws of the flexible circuit board cover the end of the base body of the finger end and are pasted and fixed; each pressure sensor is arranged outward. The end of the finger base is provided with an elastic filling layer covering each pressure sensor.

Preferably, the material of the elastic filling layer is silica gel.

Preferably, the outer side of the elastic filling layer is covered with an outer film.

In a third aspect, the present invention provides a mechanical gripper, which includes a gripper body, and gripping fingers mounted on the ends of two gripping parts of the gripper body. The mechanical gripper is the type of the object to be gripped during the gripping process. The specific process is as follows: when the pressure value detected by the pressure sensors on both gripping fingers is greater than the preset minimum gripping force; The number of sensors _Sn ; the number of clamping sensors _Sn is the average value of the number of pressure sensors whose detection pressure values on the two clamping fingers exceed the threshold, and is recorded as the number of clamping sensors _Sn . Three demarcation values n _a , n _b , and n _c that increase in sequence are set. If Sn < _{n a} , it is considered that the clamping is unreliable, and the clamping force of the mechanical clamping jaw is increased, so that the elastic filling layer on the two clamping fingers is deformed, and the number of clamping sensors _Sn increases. If n _a ≤S _n <n _b , it is determined that the target object is a rigid object; if n _b ≤S _n <n _c , it is determined that the target object is an overall deformable object; if S _n ≥n _c , it is determined that the target object is a local object Deformed objects.

Preferably, the values of n _a , n _b , and n _c are selected as follows: using mechanical grippers to clamp rigid objects, overall deformed objects, and locally deformed objects, respectively, and record the number of clamping sensors _Sn as p ₁ , p ₂ , p ₃ . Take n _a =a·p ₁ ; n _b =b·(p ₁ +p ₂ );n _c =c·(p ₂ +p ₃ ). The value ranges of a, b, and c are all 0.5 to 1.

The beneficial effects that the present invention has are:

1. The present invention uses a three-dimensional fingertip pressure sensor array composed of multiple pressure sensors as a tactile sensor. The distribution of the three-dimensional fingertip pressure sensor array can structurally detect the reaction force distribution of the end clamping structure. In addition, the present invention uses the number of pressure sensors whose detection value exceeds the threshold to detect the deformation of the object when it is squeezed, so as to identify the type of the object being clamped.

2. The present invention can be applied to the end effector of the grasping robot. By identifying the type of the object and matching the corresponding clamping force, the flexible object can be grasped with the smallest deformation.

4. The present invention uses silica gel casting to seal the pressure sensor and the flexible circuit board, and can utilize the elasticity of the silica gel to distribute the reaction force of the grasped object to a plurality of pressure sensors. Moreover, the distributed reaction force detected by each pressure sensor is different, and the tactile sensor and its adaptive system of the present invention can be effectively applied.

Description of drawings

1 is a schematic structural diagram of a tactile sensor provided in Embodiment 1 of the present invention;

2 is a schematic structural diagram of a clamping finger provided in Embodiment 2 of the present invention;

3 is a schematic cross-sectional view of a clamping finger provided in Embodiment 2 of the present invention;

4 is a schematic diagram of the mechanical gripper provided in Embodiment 3 of the present invention when different types of objects are clamped;

Fig. 5 is the working flow chart of the mechanical gripper in the present invention.

Detailed ways

Example 1

其中，r₁为夹持指末端表面的球半径；r₂为夹持指末端表面两侧圆弧边缘的半径；d为侧爪的宽度。两个侧爪的半径分别为R₁、R2。其中，

因此，在该条件下，侧爪能够更好贴合半球状指端基体1的弧面，从而提高传感器检测效率。As shown in FIG. 1 , a flexible and adaptive tactile sensor includes a pressure sensor 3 and a flexible circuit board 6 . The flexible circuit board 6 is in the shape of three claws, including a base, a middle claws 6-1 and two side claws 6-2. The inner end of the middle claw 6-1 is connected with the base through a connecting section. The two side claws 6-2 are respectively arranged on both sides of the middle claw 6-1. The inner ends of the two side claws 6-2 are connected with the inner ends of the middle claws 6-1. The middle claw 6-1 adopts a rectangular flexible circuit board, on which five pressure sensors 3 are arranged at intervals. The two side claws 6-2 are both curved, and the outer ends are both bent to the side away from the middle claws 6-1. The end surface of the clamping finger is a partial spherical surface with symmetrical arc edges on both sides. The planes of the arc edges on both sides of the gripping finger end are parallel to each other. The inner end of the side claw 6-2 is tangent to the inner end of the middle claw 6-1, and the radian of the side claw 6-2 is

Among them, r ₁ is the spherical radius of the end surface of the clamping finger; r ₂ is the radius of the arc edges on both sides of the end surface of the clamping finger; d is the width of the side claw. The radii of the two side claws are R ₁ and R2 respectively. in,

Therefore, under this condition, the side claw can better fit the arc surface of the hemispherical finger end base 1, thereby improving the detection efficiency of the sensor.

A control module 7 is installed on the base; the control module 7 is a signal transmission and processing circuit built by commercial components, and the controller in the control module 7 in this embodiment adopts a single-chip microcomputer. The two side claws 6-2 are provided with three pressure sensors 3 which are arranged at intervals in sequence. The control module 7 is connected to the pressure sensors 3 on the middle jaw 6-1 and the side jaws 6-2 respectively through flexible circuits. When the flexible circuit board 6 is covered on the hemispherical holder, the outwardly bent side claws 6-2 make the pressure sensors 3 evenly arranged on the hemispherical holder. Among the eleven pressure sensors 3, five pressure sensors 3 are distributed in the middle area of the curved surface of the hemispherical clamp, and the remaining six are distributed in the two sides of the curved surface respectively, forming a three-dimensional fingertip pressure sensor array structure to realize the clamping Omni-directional detection of the compression conditions at different positions of the holder.

Example 2

As shown in FIGS. 2 and 3 , a gripping finger based on tactile detection is used to be mounted on a mechanical gripper, which includes a finger end base 2 , an elastic filling layer 4 and the tactile sensor described in Embodiment 1. The end of the finger base body 2 is hemispherical, and the whole is like a finger end. The middle claw 6-1 and the two side claws 6-2 of the flexible circuit board 6 cover the end of the finger end base 2 and are pasted and fixed; each pressure sensor is arranged outward. The distal end of the finger base 2 is provided with an elastic filling layer 4 covering each pressure sensor. The material of the elastic filling layer 4 is silica gel, and its manufacturing method is mold casting. In this embodiment, the elastic filling layer 4 covers the flexible circuit board 6 , the pressure sensor 3 and the finger end base 2 without gaps, and forms a certain thickness. The elastic filling layer 4 has the characteristics of compressibility and isotropy; thus, the reaction force generated during the grasping process is transmitted to each pressure sensor 3 at different positions inside. The outer side of the elastic filling layer 4 is covered with an outer film 5 to protect the elastic filling layer 4 .

The clamping refers to that when pressing objects with different characteristics, the pressure sensors at different positions detect different pressure values, and according to the pressure values detected by each pressure sensor, the characteristics of the pressed object can be analyzed to simulate tactile sensation. Effect.

Example 3

A mechanical gripper includes a gripper body and gripping fingers mounted on the ends of two gripping parts of the gripper body. The structure of the clamping fingers is as described in Embodiment 2. The main body of the gripper adopts the existing gripper, which can perform the gripping action under the driving of the power element.

During the working process of the mechanical gripper, the tactile sensor as described in Embodiment 1 is used to collect pre-identification information; the controller preprocesses the collected information. The principle of using the tactile sensor to recognize the object to be clamped is: using the manipulator equipped with the tactile sensor mechanism of the present invention to clamp three different objects, which are the rigid object 8 (such as a glass) whose overall shape is easily deformed but the surface is hard. Deformable objects 9 (such as paper cups), locally deformed objects 10 (sponges, tomatoes) whose entire shape and surface are easily deformed.

When grasping the rigid object 8, the position of the tactile sensor 1 of the present invention subjected to the reaction force is only at the contact point; when grasping the overall deformed object 9, the position of the reaction force received by the tactile sensor 10 is concentrated in the area around the contact point. When grasping the locally deformed object 10 , the position of the reaction force received by the tactile sensor 1 is dispersed in the entire area or most of the area of the tactile sensor 1 .

As shown in FIG. 4 , according to the grasping principle, a threshold parameter is firstly set for the eleven distributed pressure sensors 3 , that is, when the pressure sensor reaches the threshold, it can be deemed that the pressure sensor has detected Reaction force during grasping. Then the tactile sensor 1 of the present invention is collected for information. When grasping the rigid object 8 (shown in part a in FIG. 4 ), the number of the distributed pressure sensors 3 reaching the threshold is counted as n _a . When the object 9 is deformed but its contact surface is relatively hard (shown in part b in Figure 4 ), the number of distributed pressure sensors 3 reaching the threshold is counted as n _b , and it is very easy to grasp the whole and the contact position. When the deformed flexible object 10 (shown in part c in FIG. 4 ), the number of the distributed pressure sensors 3 reaching the threshold is counted as n _c .

As shown in Figure 5, the specific process of grasping the target object and identifying it is as follows:

Step 1. The robot drives the mechanical gripper on the end effector to approach the target object and clamp it.

Step 2: The adaptive grasping control system detects whether there is a pressure value detected by the pressure sensor 3 that is greater than the preset minimum clamping force. If both gripping fingers on the mechanical gripper have pressure sensors 3 with a pressure value greater than the minimum gripping force, proceed to step 3; otherwise, the mechanical gripper only needs to be tightened for gripping.

Step 3: Take the average value of the number of pressure sensors 3 whose detected pressure values on the two clamping fingers exceed the threshold, and denote it as the number of clamping sensors S _n . The threshold value is smaller than the minimum clamping force, and is used for whether the corresponding position of each pressure sensor 3 participates in the clamping of the target object; the adaptive grasping control system records the number _Sn of the clamping sensors.

Step 4: The adaptive grasping control system is set with three boundary values na , n _b , and n _c that increase sequentially _. If _Sn < _na , the clamping is considered unreliable, the clamping force of the mechanical clamping jaws is increased, the elastic filling layer 4 on the two clamping fingers is deformed, and the number of clamping sensors _Sn increases. If n _a ≤S _n <n _b , it is determined that the target object belongs to the rigid object 8; if n _b ≤S _n <n _c , it is determined that the target object belongs to the overall deformable object 9; if S _n ≥n _c , it is determined that the target object Belonging to Local Deformation Object 10.

The specific values of n _a , n _b , and n _c use one of the following two value methods:

Method 1. Use the mechanical gripper to clamp the rigid object 8, the overall deformed object 9, and the specific example of the locally deformed object 10 (glass cups, paper cups, tomatoes), and record the number of clamping sensors _Sn as p ₁ , p ₂ , p ₃ . Take n _a =p ₁ ; n _b =p ₂ ; n _c =p ₃ .

Method 1. Use the mechanical gripper to clamp the rigid object 8, the overall deformed object 9, and the specific example of the locally deformed object 10 (glass cups, paper cups, tomatoes), and record the number of clamping sensors _Sn as p ₁ , p ₂ , p ₃ . Take n _a = 0.5·p ₁ ; n _b =0.5·(p ₁ +p ₂ ); n _c =0.5·(p ₂ +p ₃ ).

Step 5: The mechanical gripper moves the gripped target object to the target position and then releases it. In the process of clamping and handling, the mechanical gripper adjusts the clamping force and transportation speed according to the type of target object (rigid object 8, overall deformable object 9 or locally deformed object 10) identified by itself, thereby reducing the possibility of damage to the target object sex.

In addition, after pre-setting, the robot can also classify the target objects according to the type of the identified target objects, and transport different types of target objects to the corresponding designated positions.

Claims (10)

1. A flexible adaptive tactile sensor comprising a pressure sensor (3) and a flexible circuit board (6); the method is characterized in that: the flexible circuit board (6) comprises a middle claw (6-1) and two side claws (6-2); the two side claws (6-2) are respectively arranged at two sides of the middle claw (6-1); the inner ends of the two side claws (6-2) are connected with the inner end of the middle claw (6-1); the middle claw (6-1) is linear; the two side claws (6-2) are bent, and the outer ends of the two side claws are bent to one side far away from the middle claw (6-1); a plurality of pressure sensors (3) which are arranged in sequence are arranged on the middle claw (6-1) and the two side claws (6-2); in use, the middle claw (6-1) and the two side claws (6-2) are wrapped on the end surface of the spherical clamping finger.

2. A flexible adaptive tactile sensor according to claim 1, wherein: the inner end of the side claw (6-2) is tangent to the inner end of the middle claw (6-1).

3. A flexible adaptive tactile sensor according to claim 1, wherein: the surface of the tail end of the clamping finger is a circular arc with two symmetrical sidesA partial spherical surface of the edge; the radian of the side claw (6-2) is

Wherein r is₁Is the spherical radius of the end surface of the clamping finger; r is₂The radius of the arc edges at the two sides of the surface of the tail end of the clamping finger; d is the width of the side claw;

the radius of two side claws is R respectively₁R2; wherein,

4. a flexible adaptive tactile sensor according to claim 1, wherein: the flexible circuit board further comprises a base part; the base part is connected with the inner end of the middle claw (6-1) through a connecting section.

5. A flexible adaptive tactile sensor according to claim 1, wherein: a control module (7) is arranged on the flexible circuit board (6); each pressure sensor (3) is connected with the control module (7).

6. A gripping finger comprising a finger end base (2); the method is characterized in that: further comprising an elastic filling layer (4) and a tactile sensor according to any of claims 1-5; the tail end of the finger tip base body (2) is partially spherical; a middle claw (6-1) and two side claws (6-2) of the flexible circuit board (6) are covered at the tail end of the finger end base body (2) and are stuck and fixed; each pressure sensor is arranged outwards; the tail end of the finger end base body (2) is provided with an elastic filling layer (4) covering each pressure sensor.

7. A gripping finger according to claim 1, characterised in that: the elastic filling layer (4) is made of silica gel.

8. A gripping finger according to claim 1, characterised in that: the outer side of the elastic filling layer (4) is covered with an outer film (5).

9. A gripper comprising a jaw body; the method is characterized in that: further comprising the gripping fingers of claim 7 mounted on the ends of the two gripping portions of the jaw body; the mechanical clamping jaw is the type of an object to be clamped in the clamping process, and the specific process is as follows: when the pressure values detected by the pressure sensors (3) on the two clamping fingers are larger than the preset lowest clamping force; detecting the number S of grip Sensors_n(ii) a Number of clamping sensors S_nThe average value of the number of the pressure sensors (3) with the detected pressure values exceeding the threshold value on the two clamping fingers is recorded as the number S of the clamping sensors_n(ii) a Is provided with three boundary values n which are increased in sequence_a、n_b、n_c(ii) a If S_n＜n_aThe clamping is not reliable, the clamping force of the mechanical clamping jaw is increased, the elastic filling layers (4) on the two clamping fingers are deformed, and the number S of the clamping sensors is increased_nIncreasing; if n is_a≤S_n＜n_bJudging that the target object belongs to a rigid object; if n is_b≤S_n＜n_cJudging that the target object belongs to the overall deformed object; if S_n≥n_cThen the target object is determined to belong to the locally deformed object.

10. A gripper according to claim 9, wherein: n is_a、n_b、n_cThe value taking method comprises the following steps: respectively clamping a rigid object, an integrally deformed object and a locally deformed object by using a mechanical clamping jaw, and respectively recording the quantity S of clamping sensors_nIs p₁、p₂、p₃(ii) a Take n_a＝a·p₁；n_b＝b·(p₁+p₂)；n_c＝c·(p₂+p₃) (ii) a a. The value ranges of b and c are both 0.5-1.

Images