Citrus clamping and shearing integrated end effector and control method thereof

Abstract

The invention relates to the field of citrus picking equipment, and discloses a citrus clamping and shearing integrated end effector and a control method thereof.The effector comprises a flexible clamping jaw and an electric push rod used for driving the flexible clamping jaw to be opened and closed, and the method comprises the steps that fruit images are recognized, and geometric morphology parameters of citrus are obtained; inputting the geometric morphology parameters into a peel thickness regression model to calculate the peel thickness; the maximum clamping force and the minimum clamping force of the flexible clamping jaw are determined according to the peel thickness and the geometrical morphology parameters; according to the maximum clamping force and the minimum clamping force, the maximum stroke and the minimum stroke of the electric push rod are determined; determining a stroke range according to the maximum stroke and the minimum stroke, and determining a safety margin value of the stroke range according to a bilateral retraction rule; and compensating the maximum stroke and the minimum stroke based on the safety margin value to obtain a stroke safety control range, and generating a control instruction of the electric push rod according to the stroke safety control range. According to the invention, the research and development and manufacturing cost can be greatly reduced.

Description

Technical Field

[0001] This invention relates to the field of citrus harvesting equipment, specifically to a citrus clamping and shearing integrated end effector and its control method. Background Technology

[0002] With the rapid development of agricultural automation technology, fruit and vegetable harvesting robots have become core equipment for solving labor shortages and improving harvesting efficiency. Among these, the end effector, as a key component that directly interacts with the fruit, directly determines the harvesting quality and efficiency. Citrus, a widely cultivated economic crop globally, especially high-quality varieties like Ehime 38, requires extremely high precision and non-destructive harvesting due to its plump flesh and thin, tender peel. However, citrus orchards are often unstructured environments with intertwined branches, scattered fruit distribution, and tightly connected fruit stalks, necessitating shearing operations for harvesting. This presents stringent challenges to the structural design, drive mechanism, and control precision of the end effector.

[0003] Existing harvesting end effectors are mainly divided into three categories: mechanical clamping type, adsorption type, and cutting composite type. Mechanical clamping type is widely used for fruits with weak stem connections, such as apples and kiwis, due to its simple structure and convenient control. However, it is not well adapted to scenarios where the stems of citrus fruits need to be cut. Adsorption type operates through negative pressure suction cups, which can reduce fruit damage, but it has poor adaptability to the roughness of citrus surfaces and has strict requirements for the sealing and vacuum degree of the adsorption system, making it difficult to operate stably in complex orchard environments. Cutting composite type can take into account both clamping and cutting functions, but its design is mostly for small fruits such as grapes and strawberries, and does not fully consider the morphological characteristics and fragile nature of citrus fruits.

[0004] Current end-effectors for citrus harvesting still have many technical shortcomings: Firstly, the drive system mostly uses planetary geared motors or servo motors, resulting in large size and heavy weight of the equipment. This not only increases the load on the robotic arm and the energy consumption of the system, but also makes it susceptible to interference from the dense branches of citrus trees, reducing the flexibility of harvesting. Secondly, in terms of structural design, some parts adopt a fully rigid clamping mechanism, which is very easy to cause damage to the fruit peel and affect the appearance of the fruit for sale; Third, in order to achieve non-destructive clamping, existing solutions often rely on precision components such as tactile sensors and force sensors, which significantly increases the research and development and manufacturing costs. Moreover, the temperature and humidity fluctuations and accidental contact with branches and leaves in complex orchard environments can seriously affect the sensitivity of sensors and reduce the reliability of control.

[0005] In summary, existing harvesting end effectors still have significant shortcomings in adapting to the unstructured environment of citrus orchards, achieving non-destructive clamping, controlling costs, and improving reliability, making it difficult to meet the precision harvesting needs of high-quality varieties such as Ehime citrus. Summary of the Invention

[0006] The purpose of this invention is to provide an integrated clamping and shearing end effector for citrus and its control method, which solves the problem that existing harvesting end effectors still have significant shortcomings in adapting to the unstructured environment of citrus orchards, achieving non-destructive clamping, controlling costs and improving reliability, and are unable to meet the refined harvesting needs of high-quality varieties such as Ehime citrus.

[0007] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: In a first aspect, the present invention provides a control method for an integrated end effector for clamping and cutting citrus fruits, the actuator comprising a flexible gripper for clamping citrus fruits and an electric push rod for driving the flexible gripper to open and close, the method comprising: Acquire images of citrus fruits, identify the fruit images, and obtain the geometric morphological parameters of the citrus fruits; The geometric morphological parameters of citrus are input into a pre-constructed peel thickness regression model to calculate the peel thickness of citrus. Based on the peel thickness and geometric parameters of the citrus fruit, the maximum and minimum clamping forces of the flexible grippers are determined. The maximum and minimum strokes of the electric actuator are determined based on the maximum and minimum clamping forces of the flexible grippers. The stroke range is determined based on the maximum and minimum strokes of the electric linear actuator, and the safety margin of the stroke range is determined according to the double-sided inward contraction rule. The maximum and minimum strokes are compensated based on the safety margin value to obtain the safe control range of the electric linear actuator's stroke. Control commands for the electric linear actuator are then generated based on this safe control range.

[0008] Preferably, the geometric parameters include: transverse diameter and longitudinal diameter, and the step of determining the minimum clamping force of the flexible gripper includes: Obtain the density of the pulp and the density of the peel; The volume of the pulp and the volume of the peel are determined based on the thickness, longitudinal diameter, and transverse diameter of the peel. The pulp weight is determined based on the pulp volume and density, and the peel weight is determined based on the peel volume and density. The total weight of the citrus fruit is determined based on the weight of the pulp and peel, and the minimum clamping force of the flexible gripper is determined based on the total weight of the citrus fruit.

[0009] Preferably, the step of determining the maximum clamping force of the flexible gripper includes: Fruit images are input into a pre-built maturity prediction model to obtain citrus maturity indicators, including firmness and sugar content; wherein, the maturity prediction model is constructed using a random forest algorithm. The peel thickness, longitudinal diameter, transverse diameter, hardness, and sugar content are input into a pre-constructed critical equivalent stress prediction model to obtain the critical equivalent stress of the citrus. The critical equivalent stress of the citrus is used as the maximum clamping force of the flexible gripper. The critical equivalent stress prediction model is constructed using the finite element analysis method.

[0010] Preferably, the method further includes: Based on the maturity index of citrus, a maturity level of citrus is constructed, which includes: immature, mature and overripe; The safety margin value is dynamically adjusted based on the maturity level of the citrus fruit.

[0011] Preferably, the maximum and minimum strokes of the electric actuator are determined based on the maximum and minimum clamping forces of the flexible grippers, including: Using the maximum and minimum clamping forces as clamping force input values, the clamping force input values ​​are input into a pre-built stroke-clamping force regression model to obtain the stroke output. The stroke output includes: the maximum stroke of the electric actuator corresponding to the maximum clamping force and the minimum stroke of the electric actuator corresponding to the minimum clamping force.

[0012] Preferably, the method further includes: constructing a stroke-clamping force regression model, including: Obtain several sets of clamping force samples of citrus fruits, including minimum clamping force samples and maximum clamping force samples; Based on the finite element method, the stroke and clamping force samples of the electric actuator are simulated to obtain several stroke-clamping force sequences. Each stroke-clamping force sequence includes a set of clamping force samples and the stroke of the electric actuator corresponding to the clamping force sample. The initial regression model is obtained by iteratively fitting several stroke-clamping force sequences based on Newton's iteration method. The coefficients of the initial regression model are constrained, verified, and corrected using a sequential quadratic programming algorithm to obtain a corrected regression model, which is then used as the stroke-clamping force regression model.

[0013] Preferably, the method further includes: constructing a peel thickness regression model, including: Obtain morphological values ​​of several sets of citrus samples. Each set of morphological values ​​includes transverse diameter measurement, longitudinal diameter measurement, and peel thickness measurement. Based on the least squares method, linear fitting was performed on several groups of transverse diameter measurements, longitudinal diameter measurements, and peel thickness measurements to obtain a peel thickness regression model.

[0014] Secondly, the present invention provides an integrated end effector for clamping and cutting citrus fruits, the actuator comprising: a frame, on which are mounted: A controller for performing a control method for the citrus clamping and shearing end effector as described in any one of claims 1-7; A clamping device, comprising a flexible gripper and an electric push rod for driving the flexible gripper to open and close, the electric push rod being electrically connected to a controller; An image acquisition device, electrically connected to a controller, is used to acquire images of citrus fruits; A cutting device, located above the clamping device and electrically connected to the controller, is used to cut the stems of citrus fruits on the flexible grippers.

[0015] Preferably, the shearing device includes: a mounting plate mounted on a frame, the front end of the mounting plate being provided with a fixed blade and a movable blade, the movable blade being rotatably connected to the mounting plate; a servo motor being provided on the mounting plate, a connecting arm being mounted on the drive end of the servo motor, a first connecting rod being hinged to the connecting arm, the end of the first connecting rod away from the connecting arm being hinged to the movable blade.

[0016] Preferably, the clamping device further includes: a push rod frame deployed in front of the electric push rod, the telescopic end of the electric push rod being connected to the middle of the push rod frame, slide rails being deployed on the frame located on both sides of the electric push rod, sliders being slidably connected on the slide rails, and the sliders being fixedly connected to the push rod frame; Both ends of the push rod frame are hinged with a second connecting rod, and a finger clamping mounting frame is provided in front of the push rod frame, which is mounted on the frame. The flexible gripper includes two flexible gripping fingers, each with a gripping finger fixing member at its end. One end of each gripping finger fixing member is hinged to the gripping finger mounting frame, and the other end of each gripping finger fixing member is hinged to the corresponding second connecting rod.

[0017] The beneficial effects of this invention are mainly reflected in: 1. This invention adopts a compact clamping and shearing integrated upper and lower composite structure, which effectively reduces the size and weight of the end effector, reduces the interference of dense branches, and improves the flexibility of the robotic arm; at the same time, it can also reduce the load on the robotic arm, extend the equipment's battery life, and reduce the system's operating energy consumption; 2. The clamping device of the present invention uses flexible grippers, which can avoid damage to the fruit peel caused by rigid contact; 3. The control method of the present invention identifies the geometric morphological parameters of citrus, calculates the peel thickness of citrus using a peel thickness regression model, and then calculates the maximum and minimum stroke of the electric push rod using the geometric morphological parameters and peel thickness parameters. Finally, the stroke safety control range of the electric push rod is determined based on the maximum and minimum stroke of the electric push rod. Within the stroke safety control range, both clamping stability (preventing fruit slippage) and plastic deformation caused by excessive clamping can be guaranteed. 4. This invention abandons expensive tactile and force sensors and achieves clamping force control through control algorithms, which greatly reduces research and development and manufacturing costs. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of the overall structure of an integrated end effector for clamping and shearing citrus fruits provided in one embodiment of the present invention; Figure 2 This is a schematic diagram of the clamping device provided in one embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a shearing device provided in one embodiment of the present invention; Figure 4 This is a flowchart of a control method for an integrated end effector for citrus fruit clamping and cutting, provided in one embodiment of the present invention.

[0019] Legend: 1. Frame; 2. Controller; 3. Clamping device; 4. Shearing device; 5. RGB-D camera; 6. Camera mounting bracket; 301. Electric push rod; 302. Slide rail; 303. Slider; 304. Second connecting rod; 305. Finger clamping mounting bracket; 306. Flexible finger clamp; 307. Finger clamping fixing component; 308. Push rod frame; 401. Mounting plate; 402. Servo motor; 403. Connecting arm; 404. First connecting rod; 405. Fixed blade; 406. Moving blade. Detailed Implementation

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in conjunction with the accompanying drawings and descriptions of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0021] Example 1 Figure 1 This is a schematic diagram of the overall structure of an integrated end effector for clamping and cutting citrus fruits provided in one embodiment of the present invention. Figure 1As shown, this embodiment provides an integrated end effector for citrus fruit clamping and shearing. The integrated end effector for citrus fruit clamping and shearing is mounted on a citrus picking robot, which moves while carrying the integrated end effector. The actuator in this embodiment includes: a frame 1, on which a controller 2, a clamping device 3, an image acquisition device, and a shearing device 4 are mounted.

[0022] In this embodiment, the frame 1 has an overall C-shaped structure, including a C-shaped back plate, a top plate, and a bottom plate. The clamping device 3, the shearing device 4, and the processor are deployed between the top plate and the bottom plate, and the image acquisition device is deployed above the top plate.

[0023] The image acquisition device in this embodiment is electrically connected to the controller 2 and is used to acquire images of citrus fruits. The image acquisition device in this embodiment uses an RGB-D camera 5, which is mounted on a C-shaped backplate via a camera mounting bracket 6.

[0024] The clamping device 3 in this embodiment includes a flexible gripper and an electric push rod 301 that drives the flexible gripper to open and close. The electric push rod 301 is electrically connected to the controller 2.

[0025] As a further optimization of this embodiment, such as Figure 2 As shown, the clamping device 3 further includes: a push rod frame 308 deployed in front of the electric push rod 301, the telescopic end of the electric push rod 301 is connected to the middle of the push rod frame 308, and slide rails 302 are deployed on the frame 1 located on both sides of the electric push rod 301. Slider 303 is slidably connected on the slide rails 302, and the slider 303 is fixedly connected to the push rod frame 308. The push rod frame 308 is hinged to two ends with a second connecting rod 304, and a finger clamping mounting bracket 305 is provided in front of the push rod frame 308. The finger clamping mounting bracket 305 is mounted on the frame 1. The flexible gripper includes two flexible gripping fingers 306. Each flexible gripping finger 306 has a gripping finger fixing member 307 at its end. One end of any gripping fixing member is hinged to the gripping finger mounting frame 305, and the other end of any gripping fixing member is hinged to the corresponding second connecting rod 304.

[0026] In this embodiment, the two flexible gripping fingers 306 are made of TPU material and adopt a biomimetic fish fin design. The internal structure is supported by a fish fin-like structure, which allows for passive shape changes to adapt to the size of the citrus fruit and avoids damage to the peel caused by rigid contact. The push rod frame 308, the second connecting rod 304, and the gripping fasteners in the gripping device 3 are made of PLA material. The gripping device 3 in this embodiment avoids increasing the weight of the overall device while achieving stable gripping of the fruit.

[0027] In this embodiment, the controller 2 preferably uses an STM32 series microcontroller. The controller 2 in this embodiment uses a control algorithm to control the electric actuator 301. The control algorithm in this embodiment includes the following steps: acquiring an image of a citrus fruit, recognizing the fruit image to obtain the geometric morphological parameters of the citrus fruit; inputting the geometric morphological parameters of the citrus fruit into a pre-constructed peel thickness regression model to calculate the peel thickness of the citrus fruit; determining the maximum and minimum clamping forces of the flexible gripper based on the peel thickness and geometric morphological parameters of the citrus fruit; determining the maximum and minimum stroke of the electric actuator 301 based on the maximum and minimum stroke of the electric actuator 301; determining the stroke range based on the maximum and minimum stroke of the electric actuator 301, and determining the safety margin value of the stroke range according to the bilateral inward contraction rule; compensating for the maximum and minimum stroke based on the safety margin value to obtain the safe control range of the electric actuator 301's stroke, and generating control commands for the electric actuator 301 based on the safe control range of the electric actuator 301's stroke.

[0028] In this embodiment, the cutting device 4 is located above the clamping device 3. The cutting device 4 is electrically connected to the controller 2 and is used to cut the fruit stems of citrus fruits on the flexible grippers.

[0029] As a further optimization of this embodiment, such as Figure 3 As shown, the shearing device 4 includes: a mounting plate 401 mounted on a frame 1, the front end of the mounting plate 401 being provided with a fixed blade 405 and a movable blade 406, the movable blade 406 being rotatably connected to the mounting plate 401; a servo motor 402 being provided on the mounting plate 401, the servo motor 402 being electrically connected to the controller 2, a connecting arm 403 being mounted on the drive end of the servo motor 402, a first connecting rod 404 being hinged to the connecting arm 403, the end of the first connecting rod 404 away from the connecting arm 403 being hinged to the movable blade 406.

[0030] In this embodiment, the front end of the mounting plate 401 is provided with a C-shaped notch, and the top plate also has a notch with the same shape as the front end of the mounting plate 401. The fixed blade 405 is directly fixed to the edge of the notch at the front end of the mounting plate 401 by bolts, and the movable blade 406 is rotatably connected to the mounting plate 401. In this embodiment, the connecting arm 403 is fixed to the drive end of the servo motor 402. At this time, the servo motor 402 drives the connecting arm 403 to rotate, and then the first connecting rod 404 drives the movable blade 406 to rotate. After the movable blade 406 rotates, it cuts the fruit stalk of the citrus.

[0031] In this embodiment, the fixed cutting edge 405 and the moving cutting edge 406 are made of SK5 tool steel, while the mounting plate 401, the connecting arm 403, and the first connecting rod 404 are made of aluminum alloy to reduce the weight of the overall device.

[0032] This embodiment adopts a compact clamping and shearing integrated upper and lower composite structure, which effectively reduces the size and weight of the end effector, reduces interference from dense branches, and improves the flexibility of the robotic arm's movement; at the same time, it can also reduce the load on the robotic arm, extend the equipment's battery life, and reduce the system's operating energy consumption.

[0033] Example 2 Figure 4 This invention provides a control method for an integrated end effector for clamping and shearing citrus fruits, according to one embodiment of the present invention. Figure 4 As shown, this embodiment provides a control method for an integrated end effector for citrus fruit clamping and shearing, the method comprising: Step S10: Obtain an image of the citrus fruit, identify the fruit image, and obtain the geometric morphological parameters of the citrus fruit. In this embodiment, the geometric morphological parameters include the transverse diameter and the longitudinal diameter.

[0034] In this embodiment, the fruit image is captured by an RGB-D camera. Priority is given to capturing the RGB image and depth image of the front of the fruit (equatorial side facing the camera). If the posture is tilted, it is finely adjusted by a robotic arm (patented "visual positioning" function) to ensure that the visible area of ​​the fruit is ≥70%. The RGB-D camera then sends the fruit image to the controller, which recognizes the fruit image to obtain the geometric morphological parameters of the citrus.

[0035] The fruit image recognition steps in this embodiment are as follows: First, the fruit image is preprocessed, for example: white balance correction: eliminate the influence of orchard lighting fluctuations (such as strong light and shadow) on color and unify the image color space; noise reduction: smooth the image with Gaussian filtering (kernel size 3×3), remove random noise, and preserve the fruit outline; background segmentation: use "color thresholding + morphological operation" to separate the fruit from the background of branches and leaves.

[0036] Among them, the color threshold is based on the HSV space. When the Ehime mandarin orange is ripe, H∈25-35°, S∈0.6-0.8, and V∈0.5-0.9. The initial mask of the fruit is obtained by threshold filtering. Among them, morphological optimization involves performing "dilation → erosion" operations on the mask (to eliminate small twig and leaf residues), and then using "hole filling" to fill the shadow gaps inside the fruit to obtain a complete fruit outline mask.

[0037] Then, the fruit outline is extracted and fitted with the minimum bounding rectangle: the fruit outline is extracted from the preprocessed mask, and the longitudinal and transverse diameters are located by the "minimum bounding rectangle (MER)" to avoid interference from the irregular shape of the fruit.

[0038] Finally, size calibration and error correction: the pixel length in the image needs to be converted to the actual physical length (mm), and measurement errors need to be corrected.

[0039] Step S20: Input the geometric morphological parameters of the citrus into the pre-constructed peel thickness regression model to calculate the peel thickness of the citrus.

[0040] As a further optimization of this embodiment, the construction steps of the peel thickness regression model are as follows: obtain the morphological values ​​of several groups of sample citrus fruits, each group of morphological values ​​including transverse diameter measurement value, longitudinal diameter measurement value and peel thickness measurement value; perform linear fitting on the transverse diameter measurement value, longitudinal diameter measurement value and peel thickness measurement value of several groups based on the least squares method to obtain the peel thickness regression model.

[0041] In this embodiment, a digital vernier caliper was used to measure the peel thickness, transverse diameter, and longitudinal diameter of the citrus samples. For peel thickness measurement, an experimental sample of the citrus peel from the equator was taken. For each sample, the peel thickness, transverse diameter, and longitudinal diameter were measured three times consecutively, and the average value was taken to obtain the final transverse diameter, longitudinal diameter, and peel thickness measurements. The morphological values ​​of the citrus samples measured in this embodiment are shown in Table 1.

[0042] Table 1. Morphological values ​​of citrus samples

[0043] Based on the morphological values ​​of the sample citrus fruits in Table 1, a functional relationship between the transverse diameter, longitudinal diameter, and peel thickness was established using multiple linear regression analysis and the least squares method, which served as a regression model for peel thickness.

[0044] The expression for the peel thickness regression model in this embodiment is: z=f(x,y)=0.013092x+0.029703y-0.5125.

[0045] Therefore, after obtaining the longitudinal and transverse diameters of the citrus fruit through step S10, the peel thickness of the citrus fruit can be obtained by substituting them into the peel thickness regression model.

[0046] Step S30: Determine the maximum and minimum clamping forces of the flexible grippers based on the peel thickness and geometric parameters of the citrus fruit.

[0047] Specifically, the steps for determining the minimum clamping force of the flexible gripper include: Step a10: Obtain pulp density and peel density; In this embodiment, the density of all sample citrus fruits and peels can be measured, and the average value can be calculated to obtain pulp density and peel density.

[0048] Step a20: Determine the volume of the pulp and the volume of the peel based on the peel thickness, longitudinal diameter, and transverse diameter.

[0049] The pulp volume in this embodiment The calculation expression is: ; The volume of the fruit peel in this embodiment The calculation expression is: .

[0050] Step a30: Determine the pulp quality based on the pulp volume and density, and determine the peel quality based on the peel volume and density.

[0051] The formula for calculating the pulp weight in this embodiment is: ; in, For the quality of the fruit pulp, This refers to the density of the fruit pulp.

[0052] The formula for calculating the peel quality in this embodiment is: ; in, For the quality of the fruit peel, This refers to the density of the fruit peel.

[0053] Step a40: Determine the total weight of the citrus fruit based on the weight of the pulp and peel, and determine the minimum clamping force of the flexible gripper based on the total weight of the citrus fruit.

[0054] In this embodiment, when the frictional force between the flexible gripper and the citrus fruit is equal to the weight of the citrus fruit, the normal force on the single-sided contact surface of the gripper is... F M That is, the minimum clamping force.

[0055] The total mass of citrus fruits in this embodiment for: .

[0056] Minimum clamping force in this embodiment F M for: ; In the formula, g is the acceleration due to gravity. It is the static friction coefficient.

[0057] Specifically, the steps for determining the maximum clamping force of the flexible gripper include: Step b10: Input the fruit image into the pre-built maturity prediction model to obtain the maturity index of the citrus, which includes: firmness and sugar content; wherein, the maturity prediction model is constructed using the random forest algorithm.

[0058] The construction steps of the maturity prediction model in this embodiment are as follows: 1. Collect Ehime 38 mandarin orange samples at different stages of maturity (300-500 samples are recommended, covering three levels: immature, mature, and overripe).

[0059] 2. For each sample: ① Acquire RGB-D images. You can directly use the processing method in step S10 to obtain preprocessed RGB-D images, and then extract color features, texture features, and morphological and three-dimensional features from the preprocessed RGB-D images. Among them, color features can be calculated based on RGB space, including the average R value, G value, B value, and R / G ratio of the fruit region (the higher the maturity, the larger the R / G; when Ehime 38 is mature, R / G ≈ 1.3-1.6). Among them, texture features are extracted by converting RGB images to grayscale images and then using the Gray-Level Co-occurrence Matrix (GLCM) to extract quantization indicators; Among them, the morphological and three-dimensional features include fruit shape index and surface roughness. The fruit shape index is the ratio of the transverse diameter to the longitudinal diameter of the citrus sample, which serves as an auxiliary indicator for determining the maturity level. The surface roughness is based on the RGB-D depth map, and the height standard deviation of the point cloud on the fruit surface is calculated (mature citrus ≤ 0.5 mm, immature / overripe ≥ 0.8 mm), reflecting the smoothness of the peel.

[0060] ② The measured maturity indicators (sweetness was measured using a handheld saccharimeter and hardness using a digital hardness tester) were used as the "label values"; 3. Dataset partitioning: 80% as the training set and 20% as the test set.

[0061] 4. Train the random forest regression model using the training set, and test the trained random forest regression model using the test set to obtain the final trained model. Use the final trained model as the maturity prediction model.

[0062] Step b20: Input the peel thickness, longitudinal diameter, transverse diameter, hardness and sugar content into the pre-constructed critical equivalent stress prediction model to obtain the critical equivalent stress of the citrus. The critical equivalent stress of the citrus is used as the maximum clamping force of the flexible gripper. The critical equivalent stress prediction model is constructed using the finite element analysis method.

[0063] The steps for constructing the critical equivalent stress prediction model in this embodiment are as follows: Step 1: Construct a basic database of "size-maturity-mechanical parameters"; where "size" refers to the longitudinal diameter, transverse diameter, and peel thickness of the citrus; "maturity" refers to the sugar content and firmness of the citrus; and "mechanical parameters" refers to the elastic modulus (E) and yield strength (σ) of the peel at the corresponding maturity level (determined through material tensile testing).

[0064] Step 2: Simulate critical equivalent stress based on finite element analysis (AnsysWorkbench): 1. Geometric modeling: Based on different sizes of transverse diameter, longitudinal diameter, and peel thickness, establish a layered elliptical model (peel + pulp) to ensure that the model is geometrically consistent with the actual fruit. 2. Assigning material properties: Based on the correlation model in step 1, assign corresponding elastic modulus (E) and yield strength (σ) to citrus fruits of different maturity levels. 3. Solution and Critical Value Determination: After running the simulation, monitor the distribution of equivalent stress in the peel. When the equivalent stress first reaches the yield strength (σ) of the corresponding maturity, it is the critical equivalent stress of the "size-maturity" combination. Step 3: Establish a multivariate prediction model: The finite element simulation results from step 2 are fitted with the "size-maturity" input parameters to form the final multivariate prediction model, as follows: Input variables: transverse diameter (x), longitudinal diameter (y), skin thickness (z), sugar content (T), fruit firmness (H); Output variable: Critical equivalent stress (σeᵩ); The equation was fitted using a combination of multiple linear regression and sequential quadratic programming. σeᵩ=a·x+b·y+c·z+d·T+e·H+f; In the formula, a, b, c, d, e, and f are regression coefficients, which are obtained through training with finite element simulation data, and the final multivariate prediction model is used as the critical equivalent stress prediction model.

[0065] Step S40: Determine the maximum and minimum stroke of the electric actuator based on the maximum and minimum clamping forces of the flexible gripper.

[0066] In this embodiment, the steps for determining the maximum and minimum stroke of the electric actuator based on the maximum and minimum clamping forces of the flexible gripper are as follows: using the maximum and minimum clamping forces as input parameters, the input parameters are input into a pre-built stroke-clamping force regression model to obtain the stroke output, which includes: the maximum stroke of the electric actuator corresponding to the maximum clamping force and the minimum stroke of the electric actuator corresponding to the minimum clamping force.

[0067] The steps for constructing the stroke-clamping force regression model are as follows: Step c10: Obtain several sets of clamping force samples of citrus fruits, including minimum clamping force samples and maximum clamping force samples.

[0068] In this embodiment, the clamping force samples can be divided into two groups: one group is the minimum clamping force sample, and the other group is the maximum clamping force sample. The two groups of samples are used to establish the stroke-clamping force regression model corresponding to the minimum clamping force and the stroke-clamping force regression model corresponding to the maximum clamping force.

[0069] Step c20: Simulate the stroke and clamping force samples of the electric actuator based on the finite element method to obtain several stroke-clamping force sequences. Each stroke-clamping force sequence includes a set of clamping force samples and the stroke of the electric actuator corresponding to the clamping force sample. Several stroke-clamping force sequences in this embodiment are shown in Table 2.

[0070] Table 2. Stroke-Clamping Force Sequence Table

[0071] in, This is the stroke of the electric actuator corresponding to the minimum clamping force.

[0072] Step c30: Iteratively fit several stroke-clamping force sequences based on Newton's iteration method to obtain the initial regression model.

[0073] The equation obtained after fitting in this embodiment is: A* +B; Solve for the coefficients A and bias B using Newton's iteration method to obtain the optimal coefficients A and the optimal bias B.

[0074] The steps of Newton's iterative method are as follows: using the 10 groups in Table 2 and The objective function is to minimize the sum of squared residuals. By iteratively correcting the coefficients A and bias B, the optimal coefficients A and bias B that minimize the residuals can be quickly approximated.

[0075] Step c40: Based on the sequential quadratic programming algorithm, the coefficients of the initial regression model are constrained, verified, and corrected to obtain the corrected regression model. The corrected regression model is then used as the stroke-clamping force regression model.

[0076] Step S50: Determine the stroke range based on the maximum and minimum strokes of the electric actuator, and determine the safety margin value of the stroke range according to the double-sided inward contraction rule.

[0077] In this embodiment, the stroke range is the difference between the maximum stroke and the minimum stroke. Due to numerical and calculation errors during the experiment, the stroke range is the theoretical value under ideal conditions. In order to ensure the control safety of the electric push rod in the actual process (i.e., the fruit will not be crushed), this embodiment uses the double-sided inward contraction rule to determine the safety margin value of the stroke range.

[0078] The bilateral contraction rule is as follows: the travel range is divided into three equal parts, and one part is used as the safety margin value of the travel range.

[0079] Step S60: Compensate for the maximum and minimum strokes based on the safety margin value to obtain the safe control range of the electric actuator's stroke, and generate control commands for the electric actuator based on the safe control range of the electric actuator's stroke.

[0080] The travel safety control range in this embodiment is: [(D Smax -AM), (D Smin -AM)], where D Smin For the minimum travel distance, D Smax The maximum stroke is denoted by AM, which is the safety margin value. In this embodiment, the safety margin value is 1.5 mm.

[0081] As a further optimization of this embodiment, the method further includes: constructing a maturity level of citrus based on the maturity index of citrus, wherein the maturity level includes: immature, mature and overripe; and dynamically correcting the safety margin value based on the maturity level of citrus.

[0082] In this embodiment, for immature citrus (high critical equivalent stress): AM=1.2mm (strong mechanical stability, the margin can be appropriately reduced); for mature citrus (low critical equivalent stress): AM=1.5mm (the default value in this embodiment to avoid overload); for overripe citrus (extremely low critical equivalent stress): AM=1.8mm (further expand the safety range to prevent damage).

[0083] Therefore, the control method in this embodiment acquires the geometric morphological parameters of the citrus fruit using an RGB-D camera and combines them with a multivariate prediction model. This allows for precise adaptation to Ehime citrus fruits of different sizes and ripeness levels, dynamically adjusting the clamping force and the stroke of the electric push rod. This ensures both clamping stability (preventing fruit slippage) and avoids plastic deformation caused by excessive clamping. Furthermore, it eliminates the need for expensive tactile and force sensors, significantly reducing research and manufacturing costs.

[0084] Example 3 This embodiment also provides an electronic device, including a memory, a controller, and a computer program stored in the memory and executable on the controller. When the controller executes the computer program, it implements the control method of the integrated citrus clamping and cutting end effector in Embodiment 2.

[0085] This embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a controller, implements the control method for the integrated citrus-clipping end effector described in Embodiment 2. Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0086] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a controller of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the controller of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0087] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A control method for an integrated end effector for citrus fruit clamping and shearing, characterized in that, The actuator includes a flexible gripper for holding citrus fruits and an electric push rod for driving the flexible gripper to open and close. The method includes: Acquire images of citrus fruits, identify the fruit images, and obtain the geometric morphological parameters of the citrus fruits; The geometric morphological parameters of citrus fruits are input into a pre-constructed peel thickness regression model to calculate the peel thickness of citrus fruits; Based on the peel thickness and geometric parameters of the citrus fruit, the maximum and minimum clamping forces of the flexible grippers are determined. The maximum and minimum strokes of the electric actuator are determined based on the maximum and minimum clamping forces of the flexible grippers. The stroke range is determined based on the maximum and minimum strokes of the electric linear actuator, and the safety margin of the stroke range is determined according to the double-sided inward contraction rule. The maximum and minimum strokes are compensated based on the safety margin value to obtain the safe control range of the electric linear actuator's stroke. Control commands for the electric linear actuator are then generated based on this safe control range.

2. The control method for the integrated clamping and shearing end effector for citrus fruits according to claim 1, characterized in that, The geometric parameters include: transverse diameter and longitudinal diameter. The steps for determining the minimum clamping force of the flexible gripper include: Obtain the density of the pulp and the density of the peel; The volume of the pulp and the volume of the peel are determined based on the thickness, longitudinal diameter, and transverse diameter of the peel. The pulp quality is determined based on the pulp volume and density, and the peel quality is determined based on the peel volume and density. The total weight of the citrus fruit is determined based on the weight of the pulp and peel, and the minimum clamping force of the flexible gripper is determined based on the total weight of the citrus fruit.

3. The control method for the integrated clamping and shearing end effector for citrus fruits according to claim 2, characterized in that, The steps to determine the maximum clamping force of a flexible gripper include: Fruit images are input into a pre-built maturity prediction model to obtain citrus maturity indicators, including firmness and sugar content; wherein, the maturity prediction model is constructed using a random forest algorithm. The peel thickness, longitudinal diameter, transverse diameter, hardness, and sugar content are input into a pre-constructed critical equivalent stress prediction model to obtain the critical equivalent stress of the citrus. The critical equivalent stress of the citrus is used as the maximum clamping force of the flexible gripper. The critical equivalent stress prediction model is constructed using the finite element analysis method.

4. The control method for the integrated clamping and shearing end effector for citrus fruits according to claim 3, characterized in that, The method further includes: Based on the maturity index of citrus, a maturity level of citrus is constructed, which includes: immature, mature and overripe; The safety margin value is dynamically adjusted based on the maturity level of the citrus fruit.

5. The control method for the integrated clamping and shearing end effector for citrus fruits according to claim 1, characterized in that, Based on the maximum and minimum clamping forces of the flexible grippers, the maximum and minimum strokes of the electric actuator (301) are determined, including: Using the maximum and minimum clamping forces as input parameters, the input parameters are fed into a pre-built stroke-clamping force regression model to obtain the stroke output, which includes: the maximum stroke of the electric actuator corresponding to the maximum clamping force and the minimum stroke of the electric actuator corresponding to the minimum clamping force.

6. The control method for the integrated clamping and shearing end effector for citrus fruits according to claim 5, characterized in that, The method further includes: constructing a stroke-clamping force regression model, including: Obtain several sets of clamping force samples of citrus fruits, including minimum clamping force samples and maximum clamping force samples; Based on the finite element method, the stroke and clamping force samples of the electric actuator are simulated to obtain several stroke-clamping force sequences. Each stroke-clamping force sequence includes a set of clamping force samples and the stroke of the electric actuator corresponding to the clamping force sample. The initial regression model is obtained by iteratively fitting several stroke-clamping force sequences based on Newton's iteration method. The coefficients of the initial regression model are constrained, verified, and corrected using a sequential quadratic programming algorithm to obtain a corrected regression model, which is then used as the stroke-clamping force regression model.

7. The control method for the integrated clamping and shearing end effector for citrus fruits according to claim 1, characterized in that, The method further includes: constructing a peel thickness regression model, including: Obtain morphological values ​​of several sets of citrus samples. Each set of morphological values ​​includes transverse diameter measurement, longitudinal diameter measurement, and peel thickness measurement. Based on the least squares method, linear fitting was performed on several groups of transverse diameter measurements, longitudinal diameter measurements, and peel thickness measurements to obtain a peel thickness regression model.

8. A citrus-grabbing end effector, characterized in that, The actuator includes: a frame (1), on which are mounted: Controller (2), for performing the control method of the citrus clamping and shearing end effector according to any one of claims 1-7; The clamping device (3) includes a flexible gripper and an electric push rod (301) for driving the flexible gripper to open and close. The electric push rod (301) is electrically connected to the controller (2). An image acquisition device, which is electrically connected to the controller (2), is used to acquire images of citrus fruits; The shearing device (4) is located above the clamping device (3) and is electrically connected to the controller (2) for cutting the stem of the citrus fruit on the flexible gripper.

9. The citrus clamping and shearing integrated end effector according to claim 8, characterized in that, The shearing device (4) includes: a mounting plate (401) mounted on a frame (1), the front end of the mounting plate (401) being provided with a fixed blade (405) and a movable blade (406), the movable blade (406) being rotatably connected to the mounting plate (401); a servo motor (402) is provided on the mounting plate (401), a connecting arm (403) is mounted on the drive end of the servo motor (402), a first connecting rod (404) is hinged on the connecting arm (403), and the end of the first connecting rod (404) away from the connecting arm (403) is hinged to the movable blade (406).

10. The citrus clamping and shearing integrated end effector according to claim 8, characterized in that, The clamping device (3) further includes: a push rod frame (308) deployed in front of the electric push rod (301), the telescopic end of the electric push rod (301) is connected to the middle of the push rod frame (308), and slide rails (302) are deployed on the frame (1) on both sides of the electric push rod (301). A slider (303) is slidably connected on the slide rail (302), and the slider (303) is fixedly connected to the push rod frame (308). The push rod frame (308) is hinged to a second connecting rod (304) at both ends. A finger clamping mounting bracket (305) is provided in front of the push rod frame (308). The finger clamping mounting bracket (305) is mounted on the frame (1). The flexible gripper includes two flexible gripping fingers (306), and a gripping finger fixing member (307) is provided on the end of each flexible gripping finger (306). One end of any gripping fixing member is hinged to the gripping finger mounting frame (305), and the other end of any gripping fixing member is hinged to the corresponding second connecting rod (304).

Images