A method for sensing the clamping force of mechanically picking spherical fruits and vegetables

Abstract

This application discloses a method for sensing the gripping force of a spherical fruit and vegetable mechanical picking machine, which relates to the technical field of agricultural fruit and vegetable mechanical picking. The method involves selecting real fruits and vegetables for three-dimensional modeling and 3D printing the modeled model. A pressure sensor is placed in the 3D-printed fruit and vegetable model so that the contact position between the spherical fruit and vegetable mechanical picking flexible hand and the grasped object is at the center of the pressure sensor. The stepper motor driving the flexible hand is controlled to step through a preset stroke in succession, and data from the pressure sensor is acquired after each step through a preset distance. The gripping force of the spherical fruit and vegetable mechanical picking flexible hand under different states is determined based on the changing trends of the pressure sensor data. Real fruits and vegetables are modeled, and sensors are placed in the fruit and vegetable model to capture in real time the force changes at different gripping levels when the flexible hand grasps the fruit and vegetable model. This allows the spherical fruit and vegetable picking robot to determine the appropriate gripping force based on the different fruits and vegetables when actually picking.

Description

Technical Field

[0001] The present application relates to the technical field of agricultural fruit and vegetable mechanical picking, and in particular to a method for sensing the clamping force of spherical fruit and vegetable mechanical picking. Background Art

[0002] As the field of fruit and vegetable harvesting robots continues to expand, they require more complex and dexterous force sensing and control capabilities to ensure efficient operation in unstructured agricultural environments. Harvesting end-effectors are key components of fruit and vegetable harvesting robots, and flexible picking arms are the most widely used end-effectors for spherical fruit and vegetable harvesting. The interaction force between the gripping arm and the spherical fruit and vegetables during the harvesting process determines the stability and reliability of the harvesting process. Excessive gripping force can damage the fruit and vegetables.

[0003] Accurately capturing the interaction forces exerted by a robotic arm during fruit picking in real time is a key component of force sensing and intelligent control for fruit and vegetable harvesting machines. It is also a core issue for efficient and low-loss mechanical harvesting of fruits and vegetables. Because flexible robotic arms undergo significant deformation during the picking process, traditional rigid pressure sensors are difficult to attach. Furthermore, due to size limitations, sensors cannot be simply placed on or embedded within the manipulator's fingers. Therefore, many force sensing solutions for fruit and vegetable harvesting robots have explored the use of flexible thin-film pressure sensors.

[0004] However, flexible film pressure sensors can measure surface pressure and interface contact force on flat surfaces or surfaces with small curvature changes, but are not suitable for spherical fruits and vegetables with large curvature changes. When the robot wraps around spherical fruits and vegetables for picking, the robot deforms significantly, and the flexible film pressure sensor cannot accurately measure the clamping force of each finger of the robot when picking fruits and vegetables. Summary of the Invention

[0005] In order to solve the above technical problems, this application proposes the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a method for sensing the gripping force of spherical fruits and vegetables during mechanical picking, comprising:

[0007] By selecting real fruits and vegetables for 3D modeling and then 3D printing the modeled models;

[0008] A pressure sensor is placed in the 3D-printed fruit and vegetable model so that the contact point between the spherical fruit and vegetable mechanical picking flexible hand and the grasped object is at the center of the pressure sensor;

[0009] Controlling the stepping motor driving the flexible hand to step through a preset stroke one by one, and acquiring data from the pressure sensor after each stepping through the preset distance;

[0010] The clamping force of the spherical fruit and vegetable mechanical picking flexible hand in different states is determined according to the changing trend of the pressure sensor data for multiple times.

[0011] In a possible implementation, selecting real fruits and vegetables for three-dimensional modeling and then 3D printing the modeled models includes:

[0012] Spherical fruit and vegetable samples with uniform quality are selected, and the fruits and vegetables to be tested are scanned in three dimensions and modeled. The fruit and vegetable models have the shape and texture of real fruits and vegetables.

[0013] The model is segmented according to the number of fingers of the flexible hand using three-dimensional mapping software and divided into modules for 3D printing.

[0014] In one possible implementation, the pressure sensor is arranged in the 3D-printed fruit and vegetable model so that the contact position between the spherical fruit and vegetable mechanical picking flexible hand and the grasped object is at the center of the pressure sensor, including:

[0015] determining the force points of each finger when the flexible hand grasps the 3D fruit and vegetable model using different grasping postures;

[0016] Reserving pressure sensor installation positions at the divided module positions corresponding to the fulcrum points;

[0017] The pressure sensors are arranged at the pressure sensor installation positions so that a line connecting the center point of each pressure sensor and the corresponding finger force point is perpendicular to the detection end surface of the pressure sensor.

[0018] In a possible implementation, determining the force point of each finger when the flexible hand grasps the 3D fruit and vegetable model using different grasping postures includes:

[0019] Determining the number of joints of the fingers and different grasping postures of the flexible hand in actual picking;

[0020] The position of the force point is determined according to the contact point between each joint and the 3D fruit and vegetable model under different grasping postures.

[0021] In a possible implementation, each joint may include one or more fulcrum points, and the number of the fulcrum points is dynamically adjusted according to the volume of the 3D fruit and vegetable model and the size of the flexible hand.

[0022] In a possible implementation, each module on which the pressure sensor is mounted has the same width, and an anti-collision gap is provided between adjacent modules.

[0023] In a possible implementation, controlling the stepping motor that drives the flexible hand to step through a preset stroke one by one and acquiring data from the pressure sensor after each stepping through the preset distance includes:

[0024] Determine the forward distance of the stepping motor outputting a forward step of 1 mm as the preset stroke;

[0025] Dividing the preset step into a plurality of sub-strokes, and controlling the stepping motor to stop after each sub-stroke;

[0026] The data of the pressure sensor is collected at each stopping point and after each stepping of a preset distance.

[0027] In a possible implementation, collecting data from the pressure sensor at each stopping point and after each stepping of a preset distance includes:

[0028] Real-time acquisition of voltage signals from each channel of the pressure sensor;

[0029] The collected voltage signal is transmitted to the host computer via the data acquisition card;

[0030] The data processing system of the host computer receives the voltage signal from the signal acquisition card, performs linear transformation and feature extraction on the voltage signal received corresponding to the acquisition channel, and obtains the feature vector of the voltage of each acquisition channel. The feature extraction includes extracting the average value, maximum value and minimum value of the voltage.

[0031] In an embodiment of the present application, real fruits and vegetables are modeled, and sensors are placed in the fruit and vegetable model to collect in real time the changes in force at different grasping degrees when the flexible hand grasps the fruit and vegetable model. This can then determine the appropriate grasping force for the spherical fruit and vegetable picking robot according to the different fruits and vegetables when actually picking. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A schematic diagram of a flow chart of a method for sensing the gripping force of mechanically picking spherical fruits and vegetables provided in an embodiment of the present application;

[0033] Figure 2 Schematic diagram of the 3D printing model of fruits and vegetables provided in the embodiment of this application;

[0034] Figure 3 A schematic diagram of placing pressure sensors in a 3D printed model of fruits and vegetables provided in an embodiment of the present application;

[0035] Figure 4 A schematic diagram of a pressure sensor provided in an embodiment of the present application;

[0036] Figure 5 Schematic diagram of the three-finger flexible hand picking gripping force sensing device provided in an embodiment of the present application;

[0037] Figure 6 Graphs showing the changes in motor stroke and pressure at different heights of 30mm, 40mm, and 50mm from the flexible palm provided in an embodiment of the present application;

[0038] Figure 7 A diagram showing the variation in the gripping force of a single finger of the three-finger flexible hand under different motor strokes provided in an embodiment of the present application. DETAILED DESCRIPTION

[0039] The present solution will be described below with reference to the accompanying drawings and specific implementation methods.

[0040] See also Figure 1 The present embodiment provides a method for sensing the gripping force of mechanically picking spherical fruits and vegetables, including:

[0041] S101, performing three-dimensional modeling by selecting real fruits and vegetables and performing 3D printing on the modeled models.

[0042] Spherical fruit and vegetable samples with uniform quality are selected, and the fruits and vegetables to be tested are scanned in three dimensions and modeled. The fruit and vegetable models have the shape and texture of real fruits and vegetables. The model is divided into modules according to the number of fingers of the flexible hand through three-dimensional mapping software for 3D printing, such as Figure 2 In this embodiment, the 3D printing material is white resin with a precision of 0.1 mm.

[0043] S102 , placing a pressure sensor in the 3D printed fruit and vegetable model so that the contact position between the spherical fruit and vegetable mechanical picking flexible hand and the grasped object is at the center of the pressure sensor.

[0044] Since the spherical fruit and vegetable robot will have different grasping postures according to the different growth postures of fruits and vegetables in actual operation, in this embodiment, the force point of each finger when the flexible hand adopts different grasping postures to grasp the 3D fruit and vegetable model is first determined; then, pressure sensor installation positions are reserved at the divided module positions according to the force points; the pressure sensors are arranged at the pressure sensor installation positions, see Figure 3 , so that the line connecting the center point of each pressure sensor and the corresponding finger's force point is perpendicular to the detection end surface of the pressure sensor. Figure 4 The pressure sensor in this embodiment is a cylindrical resistive sensor capable of detecting tension and pressure in the normal direction of the sensor. One end of the sensor is fixed, and the other end is a pressure detection end. The pressure sensor has a range of 0-20N and an accuracy of 0.1%.

[0045] Because the flexible hand includes multiple fingers, and each finger includes multiple joints, the fulcrums of different joints are different under different grasping postures, so it is necessary to consider the different force points of different joints under a variety of different grasping postures. In this embodiment, in order to determine the fulcrum of each finger when the flexible hand adopts different grasping postures to grasp the 3D fruit and vegetable model, the number of joints of the finger and the different grasping postures of the flexible hand in actual picking are first determined, and then the position of the fulcrum is determined according to the contact point between each joint and the 3D fruit and vegetable model under different grasping postures. In this embodiment, the fulcrum determined for each joint includes one or more, and the number of the fulcrums is dynamically adjusted according to the volume of the 3D fruit and vegetable model and the size of the flexible hand.

[0046] In this embodiment, each module for mounting the pressure sensor has the same width, set to 20-22 mm, to prevent mutual interference during the clamping process. An anti-collision gap is set between adjacent modules, and the anti-collision gap is set to 4-5 mm. In this embodiment, the modules and pressure sensors are assembled and fixed using an adhesive method.

[0047] S103, controlling the stepping motor that drives the flexible hand to step a preset stroke in sequence, and acquiring data from the pressure sensor after each step of the preset distance.

[0048] The flexible hand's grasping action relies on motor control. A controller sends commands to the stepper motor driver, which then drives the motor. In this embodiment, the preset stroke is determined as the forward distance of the stepper motor's output in a 1mm forward step. To prevent deformation delays in the flexible picking manipulator from affecting experimental accuracy, the preset stroke is divided into multiple substrokes, and the stepper motor is stopped after each substroke. Data from the pressure sensor is collected at each stop point and after each step of the preset stroke.

[0049] The pressure data acquisition system in this embodiment includes a host computer, a data processing system, and a data acquisition card. The data acquisition card is connected to the signal lines of the pressure sensor via wires and is used to collect voltage signals from each channel of the pressure sensor in real time. The collected voltage signals are then transmitted to the host computer via the data acquisition card. The data processing system receives the voltage signals from the signal acquisition card and performs linear transformation and feature extraction on the received voltage signals corresponding to the acquisition channels to obtain a feature vector for the voltage of each acquisition channel. Feature extraction includes extracting features such as the average, maximum, and minimum values of the voltage, providing a basis for subsequent fusion analysis of the pressure signals.

[0050] S104: Determine the clamping force of the spherical fruit and vegetable mechanical picking flexible hand in different states according to the changing trends of the pressure sensor data for multiple times.

[0051] This application uses an apple as an example, but is not limited to apples. It also uses a three-finger flexible hand as an example, but is not limited to a three-finger robotic hand. This embodiment uses the three-finger flexible hand as an example for modular design. This method is universally applicable to flexible hands of varying structures (different flexible hands and fruits and vegetables have different modular designs). In this embodiment, the midline of the adjacent longitudinal front modules and the sensor is 120°, coinciding with the midline of the three-finger flexible hand.

[0052] See also Figure 5 The pressure sensors include: a first pressure sensor 1, a second pressure sensor 2, a third pressure sensor 3, a fourth pressure sensor 4, a fifth pressure sensor 5, a sixth pressure sensor 6, a seventh pressure sensor 7, an eighth pressure sensor 8, and a ninth pressure sensor 9. The pressure sensors are all resistive sensors capable of detecting tension and pressure in the direction of the sensor normal. The combined force of the first pressure sensor 1, the fourth pressure sensor 4, and the seventh pressure sensor 7 is used as the clamping force of the first mobile phone, the combined force of the second pressure sensor 2, the fifth pressure sensor 5, and the eighth pressure sensor 8 is used as the clamping force of the second finger, and the combined force of the third pressure sensor 3, the sixth pressure sensor 6, and the ninth pressure sensor 9 is used as the clamping force of the third finger.

[0053] In this example, three sets of tests were conducted at 30mm, 40mm, and 50mm from the bottom of the fruit and vegetable model to the palm. To ensure data accuracy, three tests were performed for each set of tests and the average value was taken to simulate the gripping force at different gripping positions during the actual picking process. The test was conducted as follows:

[0054] First, calibrate the pressure sensor using a standard mass weight according to the instructions. Adjust the device to the appropriate height, and send a command to the stepper motor driver through the controller. The driver drives the motor to rotate, causing the stepper motor to advance 1mm. When the motor stops rotating, the pressure sensor data of the data acquisition card is recorded by the host computer. Continue to repeat the above steps, each time making the motor step forward 1mm until the maximum stroke is reached. During the experiment, the data acquisition card is connected to the signal line of the pressure sensor through a wire, and these signals are transmitted to the host computer for analysis. In order to prevent the deformation delay of the flexible picking robot from affecting the accuracy of the experiment, the motor is stepped forward 1mm each time to complete the experiment in multiple sections, and the pressure sensor data is collected by the host computer at each pause point to record the clamping force changes under different strokes.

[0055] After one set of tests was completed, the flexible hand was reset and the height of the apple model was adjusted, and another set of tests was performed at different heights. Figure 6The figure shows the force-motor travel relationship detected by different pressure sensors at heights of 30mm, 40mm, and 50mm. At a height of 30mm from the palm, the force recorded by the sensor gradually increases with increasing displacement. The responses of different sensors vary significantly, with the highest detected force approaching 20N, indicating that the clamping force is relatively high at this height. At a height of 40mm, the force increase trend is more stable than at a height of 30mm, but the response distribution among different sensors is still relatively wide, with the maximum detected force being approximately 10N. This indicates that the clamping force generally decreases with increasing height. At a height of 50mm, the force response decreases further, with all sensors recording force values below 10N, indicating a decreasing clamping force trend with increasing height. The overall force-displacement curve is closer to smooth, indicating that the system's force response is relatively consistent at higher positions. The experiment demonstrates that the clamping force detection device can clearly display the clamping force changes at different pressure sensor positions and facilitate analysis of the clamping force distribution at different positions.

[0056] In this embodiment, the combined force of the three sensors in the longitudinal direction of the sensor is used as the gripping force of the flexible hand. Although the direction of the gripping force changes with the movement of the finger, for the fin-shaped gripper, when the gripper's opening angle reaches its maximum, the maximum angle between the normal direction and the sensor axis direction is approximately 12.5°. Therefore, assuming that the gripping force direction is consistent with the sensor axis direction, the maximum error caused by this is only about 1-cos(12.5°) = 2.4%, which can be ignored. The gripping force of the finger at different distances from the palm is as follows: Figure 7 As shown, the finger clamping force at different heights shows different growth trends, while the finger clamping force at the same height has almost the same growth trend, which proves that the clamping force detection device of the present invention can quickly and accurately reflect the clamping force of the flexible hand.

[0057] In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0058] The above description is merely a specific embodiment of the present application. Any person skilled in the art may easily conceive of variations or substitutions within the technical scope disclosed in this application, and such variations or substitutions shall be within the scope of protection of this application. The scope of protection of this application shall be subject to the scope of protection of the claims.

Claims

1. A method for sensing the gripping force of mechanically picked spherical fruits and vegetables, characterized in that: include: By selecting real fruits and vegetables for 3D modeling and then 3D printing the modeled models; A pressure sensor is arranged in the 3D printed fruit and vegetable model so that the contact position between the spherical fruit and vegetable mechanical picking flexible hand and the grasped object is at the center of the pressure sensor, including: determining the force points of each finger when the flexible hand grasps the 3D fruit and vegetable model using different grasping postures; Reserving pressure sensor installation positions at the divided module positions corresponding to the fulcrum points; Arrange the pressure sensors at the pressure sensor installation positions so that the line connecting the center point of each pressure sensor and the corresponding finger force point is perpendicular to the detection end surface of the pressure sensor; Determining the force point of each finger when the flexible hand grasps the 3D fruit and vegetable model using different grasping postures includes: Determining the number of joints of the fingers and different grasping postures of the flexible hand in actual picking; determining the position of the force point according to the contact point between each joint and the 3D fruit and vegetable model in different grasping postures; Controlling the stepping motor driving the flexible hand to step through a preset stroke one by one, and acquiring data from the pressure sensor after each stepping through the preset distance; The clamping force of the spherical fruit and vegetable mechanical picking flexible hand in different states is determined according to the changing trend of the pressure sensor data for multiple times.

2. The method for sensing the gripping force of mechanically picking spherical fruits and vegetables according to claim 1, characterized in that: The method of selecting real fruits and vegetables for three-dimensional modeling and 3D printing the modeled models includes: Spherical fruit and vegetable samples with uniform quality are selected, and the fruits and vegetables to be tested are scanned in three dimensions and modeled. The fruit and vegetable models have the shape and texture of real fruits and vegetables. The model is segmented according to the number of fingers of the flexible hand using three-dimensional mapping software and divided into modules for 3D printing.

3. The method for sensing the gripping force of mechanically picking spherical fruits and vegetables according to claim 1, characterized in that: Each joint includes one or more fulcrum points, and the number of the fulcrum points is dynamically adjusted according to the volume of the 3D fruit and vegetable model and the size of the flexible hand.

4. The method for sensing the gripping force of mechanically picking spherical fruits and vegetables according to claim 1, characterized in that: The width of each module on which the pressure sensor is installed is the same, and anti-collision gaps are set between adjacent modules.

5. The method for sensing the gripping force of mechanically picking spherical fruits and vegetables according to claim 1, characterized in that: The stepping motor that controls and drives the flexible hand to step through a preset stroke one by one, and obtains data from the pressure sensor after each stepping through the preset distance, including: Determine the forward distance of the stepping motor outputting a forward step of 1 mm as the preset stroke; Dividing the preset stroke into a plurality of sub-strokes, and controlling the stepping motor to stop after each sub-stroke; The data of the pressure sensor is collected at each stopping point and after each stepping of a preset distance.

6. The method for sensing the gripping force of mechanically picking spherical fruits and vegetables according to claim 5, characterized in that: The step of collecting data from the pressure sensor at each stopping point and each time the preset distance is stepped includes: Real-time acquisition of voltage signals from each channel of the pressure sensor; The collected voltage signal is transmitted to the host computer via the data acquisition card; The data processing system of the host computer receives the voltage signal from the signal acquisition card, performs linear transformation and feature extraction on the voltage signal received corresponding to the acquisition channel, and obtains the feature vector of the voltage of each acquisition channel. The feature extraction includes extracting the average value, maximum value and minimum value of the voltage.

Images