Dexterous hand fruit and vegetable grabbing control method based on fuzzy variable impedance control

Abstract

The invention discloses a dexterous hand fruit and vegetable grabbing control method based on fuzzy variable impedance control. The method comprises the following steps: firstly, establishing a multi-finger collaborative kinematics model based on a spinor theory, and determining fingertip poses, grabbing force of each finger and multi-finger resultant force distribution; on the basis of force sealing, an optimization model with the minimum contact force as a target is constructed, an NSGA-II algorithm is adopted for solving, and the minimum grabbing force needed by stable and lossless grabbing is obtained; establishing an impedance control model in a Cartesian space, enabling a grabbing system to be equivalent to a mass-spring-damping system, analyzing the influence of an impedance parameter on grabbing performance, and determining an inertial parameter and a damping coefficient as key adjustment objects; fuzzy control is further introduced, the grabbing force error and the change rate of the grabbing force error serve as input and output inertia and damping adjustment amount, a fuzzy self-adaptive variable impedance controller is constructed, and dynamic optimization of impedance parameters and accurate tracking of the grabbing force are achieved. The fruit and vegetable grabbing device can be effectively applied to rapid, stable and lossless grabbing of various fruits and vegetables in an unstructured agricultural environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robots, in particular to the non-damage grasping control of a dexterous hand to fruits and vegetables in an agricultural picking robot. BACKGROUND

[0002] With the continuous development of agricultural intelligent equipment, automated and intelligent fruit and vegetable picking robots have gradually become a research hotspot in the field of agricultural picking. As the key component of the direct interaction between the robot and the fruit and vegetable, the performance of the end effector directly affects the picking efficiency and fruit quality. At present, the special effector designed for specific fruits and vegetables has poor adaptability when dealing with fruits of different shapes, hardness and surface characteristics. In contrast, the multi-fingered dexterous hand, due to its flexible structure and diverse grasping modes, shows stronger versatility and adaptability, and is expected to become an ideal effector for fruit and vegetable picking robots.

[0003] However, the dexterous hand still faces many challenges in the application of fruit and vegetable picking. The multi-fingered and multi-jointed characteristics of the dexterous hand result in complex grasping planning, and factors such as grasping posture, contact point selection and grasping force distribution will affect the grasping effect. More importantly, there is a contradiction between "grasping stability" and "grasping damage" in the grasping process: too large grasping force will cause mechanical damage to the fruit and vegetable, and too small grasping force will easily cause slipping. In addition, the fruit and vegetable itself has characteristics such as easy damage, easy deformation and unknown surface characteristics, combined with interference factors such as vibration, obstruction and light changes in the agricultural unstructured environment, which further increases the difficulty of grasping control of the dexterous hand.

[0004] In terms of control strategy, although the traditional impedance control can achieve the adjustment of grasping force by simulating the behavior of mass-spring-damper system, its parameters are usually fixed, which makes it difficult to adapt to the physical characteristics of different fruits and vegetables and the changes of dynamic environment. Fixed parameters are easy to cause force response overshoot or lag, which in turn causes fruit damage or picking failure. Therefore, there is an urgent need for a control method that can dynamically adjust the impedance parameters and adapt to various grasping conditions to achieve fast, stable and non-damage grasping of the dexterous hand to fruits and vegetables. SUMMARY

[0005] In order to achieve the above purpose, the present application proposes a dexterous hand grasping control method based on fuzzy variable impedance control, which comprises the following technical steps for solving the technical problems: Step 1: Construct the kinematics model of the dexterous hand grasping fruits and vegetables, establish the multi-fingered coordinated kinematics model based on the screw theory, determine the pose of each finger tip in the base coordinate system, the grasping force of each finger and the distribution of multi-fingered force; Step 2: Based on the force closure as the judgment basis, construct a multi-objective grasping force optimization model with the minimum contact force and the minimum deviation of contact force from the normal direction, solve it by using NSGA-II algorithm, and obtain the minimum grasping force required for stable and non-damage grasping; Step 3: Establish an impedance control model in Cartesian space, and then simulate the established impedance control model to obtain candidate values ​​for impedance control parameters, including candidate inertia coefficients. Candidate damping coefficient Candidate stiffness coefficient ; Step 4: Introduce fuzzy control, using the grasping force tracking error and its rate of change as input, and outputting the inertia adjustment ΔM and damping adjustment ΔB, to construct a fuzzy adaptive variable impedance controller based on candidate inertia coefficients. and candidate damping coefficient Calculate the inertia coefficient of the fuzzy adaptive variable impedance controller and damping coefficient The stiffness coefficient K of the fuzzy adaptive variable impedance controller will be set as the candidate stiffness coefficient. Fuzzy adaptive variable impedance controller based on inertia coefficient and damping coefficient It uses the desired pose, velocity, and acceleration of the fingertip in Cartesian space coordinates, the actual pose, velocity, and acceleration of the fingertip in Cartesian space coordinates, and the set stiffness coefficient K and desired gripping force as inputs to achieve precise tracking of the gripping force.

[0006] Furthermore, in step 1, a single-finger kinematic model is established using spinor theory, including: establishing the transformation relationship between the base coordinate system of each finger and the global coordinate system; using the exponential product formula to represent the fingertip pose; calculating the spatial Jacobian matrix of each finger; and establishing the mapping relationship between joint torque and contact force.

[0007] Furthermore, in step 2, the gripping force optimization model includes the following constraints: force balance constraint, friction cone constraint, and joint torque constraint.

[0008] Furthermore, in step 3, the impedance control model is: In the formula, , , These are the desired inertia coefficient matrix, the desired damping coefficient matrix, and the desired stiffness coefficient moment, respectively. This represents the actual pose of the robotic arm's end effector. The desired pose of the robotic arm's end effector. For actual grasping power, For the desired grasping force.

[0009] Furthermore, the candidate inertia coefficient and candidate damping coefficient The desired stiffness coefficient matrix K represents the key adjustment parameters determined through simulation analysis. , .

[0010] Further, in step 4, fuzzy control is introduced, taking the tracking error of the gripping force and its rate of change as input, and outputting inertia adjustment amount ΔM and damping adjustment amount ΔB, including: The gripping force error And the error rate of change Is fuzzified into seven linguistic variables by a triangular membership function: {NB, NM, NS, ZO, PS, PM, PB}; According to the preset fuzzy rule table, fuzzy reasoning is carried out to obtain the fuzzy output of the inertia change amount ΔM and the damping change amount ΔB; The fuzzy rule is: when the gripping force error e f Belongs to fuzzy set , and the error rate of change Belongs to fuzzy set B h , then the output inertia coefficient change amount ΔM belongs to fuzzy set , and the damping coefficient change amount ΔB belongs to fuzzy set ; The membership degree of the horizontal row corresponding to the subset in the fuzzy set and the vertical column corresponding to the subset in the fuzzy set B h is the subset in the fuzzy set ; The membership degree of the horizontal row corresponding to the subset in the fuzzy set and the vertical column corresponding to the subset in the fuzzy set B h is the subset in the fuzzy set .

[0011] Further, the method involves a dexterous hand control system, including: A kinematic modeling module for constructing a dexterous hand gripping model based on screw theory; An optimization calculation module for solving the minimum gripping force using the NSGA-II algorithm; An impedance control module for implementing preliminary tracking of the gripping force; A fuzzy reasoning module for dynamically adjusting the impedance parameters M and B according to the gripping force error and rate of change.

[0012] The present application has the following beneficial effects relative to the prior art: 1. The balance between gripping stability and fruit damage is achieved: by constructing an optimization model based on force closure theory and solving the minimum necessary gripping force using the NSGA-II algorithm, the gripping force can be minimized while ensuring gripping stability and preventing fruit and vegetable slipping, thereby effectively avoiding mechanical damage to fruits and vegetables.

[0013] 2. Significantly improved the adaptability of the system in unstructured environments: By introducing fuzzy control logic, a fuzzy adaptive variable impedance controller is constructed, which can dynamically adjust impedance parameters in real time according to the grasping force error and its rate of change, overcoming the shortcomings of fixed impedance control parameters, inability to cope with differences in physical properties of fruits and vegetables, and environmental disturbances.

[0014] 3. Optimized the dynamic response performance of the grasping process: Through simulation analysis of impedance parameters and design of fuzzy rules, the system can intelligently adjust the response characteristics according to the actual situation. For example, fast response when the error is large, and suppression of overshoot and oscillation when approaching the target, thereby achieving faster response speed, smaller overshoot, and higher steady-state accuracy.

[0015] 4. Solving the key contradiction in agricultural picking: This method directly addresses the core contradiction between "grasping firmly" and "not damaging", providing an effective technical path for the reliable grasping of dexterous hands in complex and variable agricultural unstructured environments (such as vibration, occlusion, and fruit diversity) through a comprehensive solution that integrates kinematics, optimization algorithms, and intelligent control.

[0016] 5. Improved the level of intelligence and automation of control: This method does not rely on accurate fruit and vegetable models and environmental parameters, and can automatically complete online optimization of parameters and accurate tracking of grasping force through sensor feedback and fuzzy reasoning, reducing the dependence on prior knowledge and enhancing the robustness and practicality of the entire picking system. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, a brief introduction to the drawings needed in the embodiment or prior art description will be given below. Obviously, the drawings described below are some embodiments of the present application, and those skilled in the art can obtain other drawings based on these drawings without creative labor.

[0018] Figure 1 Base coordinate of dexterous hand; Figure 2 Thumb joint axis and coordinate diagram of dexterous hand; Figure 3 Transformation of contact coordinate system to fruit coordinate system; Figure 4 Impedance parameter simulation test; Figure 5 System block diagram of fuzzy variable impedance controller; Figure 6 Fuzzy language system; Figure 7 Point contact diagram; Figure 8 for a non-destructive contact model; Figure 9 for a friction factor determination test bench; Figure 10 for a non-destructive gripping force determination test bench; Figure 11 for a non-destructive operating force determination test bench; Figure 12 for a cone portion of a contact model; Figure 13 for a three-dimensional response map. DETAILED DESCRIPTION

[0019] In order to make the objectives, technical solutions and advantages of the present application clearer, and to facilitate the understanding and implementation of the present application by those skilled in the art, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0020] A dexterous hand grasping control method based on fuzzy variable impedance control, comprising the following steps: Step 1: build a dexterous hand grasping fruit and vegetable model, build a kinematics model of the dexterous hand based on the screw theory through the base coordinate relationship of the fingers, calculate the flexion / extension angle and adduction / abduction angle of each finger joint according to the motion information of each finger joint, convert the flexion / extension angle and adduction / abduction angle of each finger joint into the actual fingertip Cartesian position through the kinematics model module, obtain the single finger fingertip pose of the dexterous hand when grasping fruit and vegetables, and obtain the grasping force size of multiple fingers when grasping fruit and vegetables through the actual fingertip grasping force measured by the thin film force sensor of the fingertip of the dexterous hand.

[0021] As Figure 1 A global coordinate system is established at the base joint of the middle finger, and the base joint coordinates of the thumb, index finger, middle finger, ring finger and little finger are represented as , .

[0022] The pose of the base joint coordinates of each finger relative to the base joint of the middle finger is: (1) In the formula, is the transformation matrix of the base coordinates of the i-th finger relative to the base joint of the middle finger, is the base joint coordinate of the middle finger.

[0023] Each finger of the dexterous hand is equivalent to a mechanical arm, so the dexterous hand can be decomposed into multiple groups of mechanical arms to obtain the whole hand kinematics model. The coordinate system of a single finger is established, and the fingertip position and Jacobian matrix of each finger are solved.

[0024] definition Let be the coordinate transformation matrix from the base coordinate system to the end-effector coordinate system when the robotic arm is in the zero position. Here are the direction vectors of each joint axis. These are the position vectors of each joint axis. Let j be the rotational motion of the j-th joint. , It is the angle of the j-th joint. It is the motion spinor matrix in the form of the "hat operator" of the j-th joint.

[0025] Based on the POE kinematic model of an n-DOF manipulator, the fingertip pose of the j-th joint of the i-th finger in its base joint coordinate system is generated. (2) in, , —— The corresponding special orthogonal group Lie algebra, .

[0026] We'll analyze the thumb; the other fingers are analyzed using the same method. For example... Figure 2 As shown, establish a thumb-based coordinate system. With fingertip coordinate system .

[0027] The kinematic parameters of the thumb POE model are shown in Table 1. In the zero position state, the fingertip coordinate system... Relative thumb base coordinate system The transition matrix is (3) Table 1. Kinematic parameters of the exponential product of thumb. Therefore, substituting the parameters from Table 1 and Equation 3 into Equation 2, we obtain the thumb tip posture of the dexterous hand as follows: (4) In the formula, , , . express , express , express , express r and e are taken as 1 and 2, respectively. — Thumb The positional variables of the joints, r and e, are both less than the maximum value of the number of finger joints j; For the spatial Jacobian of a serial robot, spinor theory has... (5) where , Ad is the adjoint transformation.

[0028] Substituting the model parameters in Table 1 into equation 5, the spatial Jacobian of the thumb is: (6) In the contact coordinate system, the joint torque and the mapping relationship of the contact force are: (7) The spatial instantaneous velocity of the thumb fingertip can be expressed by the kinematic screw as: (8) where is the configuration space of the thumb fingertip in the thumb base coordinate system.

[0029] According to the grasping fruit and vegetable model, the contact coordinate system, the fruit and vegetable centroid coordinate system, and the world coordinate system are established, and the fruit and vegetable centroid coordinate system is selected as the world coordinate system, as shown in Figure 3 .

[0030] The pose of the ith contact coordinate system relative to the fruit and vegetable coordinate system is , and the contact force in the fruit and vegetable coordinate system is expressed as: (9) where . In the equation , G i is the grasping matrix of the ith contact point, is the force screw basis of the frictional point contact, is a 3 × 3 unit matrix; When the contact force of the k fingers acts on the fruit, the mapping relationship between the contact force screw and the external force screw is (10) where , In the equation, is the grasping matrix; is the contact force matrix; is the external force screw balanced by the contact resultant force.

[0031] Step 2: Determine the stable picking judgment rule based on force closure, constrain the picking force balance, non-destructive contact linearization, and joint torque, and construct the grasping fruit and vegetable grasping force optimization model. Based on the NSGA-II algorithm, the minimum contact force is optimized to obtain the minimum grasping force required when grasping fruit and vegetables, i.e. to obtain the expected grasping force. The expected stable grasping is achieved with the minimum grasping force and operating force, ensuring that the fruit and vegetables are not damaged.

[0032] Taking the minimum grasping force of each contact point as the index, the minimum grasping force index function , The smaller the value, the higher the fruit and vegetable loss rate. (11) wherein, In the formula, is the grasping force matrix; is the grasping force weight matrix. Since the dexterous hand finger grasps is a finger surface contact, the finger surface is divided into k point contacts, so is the force of each point in the z-axis direction, as shown in Figure 7 .

[0033] The friction cone angle at the contact point The angle between the contact force and the normal at the i-th contact point wherein , , is the force of each point in the x, y, z-axis direction. Under the premise of ensuring the minimum grasping force of each finger tip, the finger tip force is closest to the normal direction, and the minimum operating force index function is established , (12).

[0034] It is worth noting that the friction cone angle is unknown, and the friction coefficient between the finger and the fruit needs to be solved first. This solving process needs the help of instruments, and the specific method is as follows: The dexterous hand mainly relies on operating force to pick, so it is assumed that there is friction point contact between the finger and the fruit and vegetable skin, and the rolling or sliding of the finger on the fruit and vegetable skin is not considered. As shown in Figure 8 , the contact force when picking fruit and vegetables can be expressed as , , , are the components of the contact force in the x, y, z directions, and the contact force is located inside the friction cone, that is, the operating force is less than or equal to the product of the grasping force and the friction factor . The contact force component perpendicular to the contact surface is the main cause of fruit and vegetable damage, and the tangential operating force to the contact surface will cause fruit skin scratches. At the same time, the contact force close to the edge of the friction cone will increase the possibility of sliding. Therefore, the contact force of the fruit and vegetable should be kept within a certain threshold to avoid uncertain damage and improve contact stability, and the maximum contact force in the lossless state is decomposed into lossless grasping force and lossless operating force . Therefore, the constraint of the frictionless lossless point contact model is: (13) As shown in Figure 9As shown, water honey peach is more susceptible to damage than other fruits and vegetables, so water honey peach "Xia Hui 5" is used as the test object to measure its static friction factor. The measuring equipment is MXD-02 type friction factor instrument (Jinan Lan Guang Electromechanical Technology Co., Ltd.), the contact material is polyurethane rubber, and the Shore hardness is 40A. During the test, the polyurethane rubber (size 150mmx80mmx2mm) is fixed on the test sliding table through the clamp, the fine rope is connected with the tensile sensor and the water honey peach, the water honey peach is stably placed on the contact material, the weight of the friction factor instrument is set equal to the mass of the water honey peach, and the moving speed of the test sliding table is 100mm / min. In order to reduce the test error, 3 water honey peach samples are tested for 3 times respectively and the average value is taken, and the test results are shown in Table 2.

[0035] Table 2: Static friction factor measurement results Among them, the non-destructive contact force can be decomposed into non-destructive gripping force and non-destructive operating force . The non-destructive gripping force test platform mainly includes dexterous hand and FSR402 type film pressure sensor, as shown in Figure 10 . The film pressure sensor is arranged under the fingertip material, the dexterous hand fingertip is applied to the centroid above the water honey peach for 10s, and the value of the film pressure sensor is recorded, that is, the gripping force. From 2 to 8N, every interval of 0.5N is a group of gripping force test, a total of 13 groups. After placing in a room temperature environment of 25℃ for 48h, the browning of the skin and flesh of the water honey peach is observed, and the maximum gripping force in the group without obvious browning is recorded as the non-destructive gripping force .

[0036] The non-destructive operating force test platform mainly includes an iron stand, a dexterous hand, a FSR402 type film pressure sensor, and a digital tensile meter (0~50.00N, accuracy: 0.01N), as shown in Figure 11 . The water honey peach is placed on a tray, the tray is connected to the iron stand in the vertical direction through a fine rope, and the water honey peach is connected to the digital tensile meter in the horizontal direction. The dexterous hand fingertip applies a stable pressure to the water honey peach above, and keeps the film pressure sensor at a specified value . By gradually increasing the tension, the reading of the digital tensile meter is changed, and it is kept for 10s, which is defined as the operating force. The operating force is from 2 to 8N, every interval of 0.5N is a test, a total of 13 groups. After placing in a room temperature environment of 25℃ for 48h, the browning of the skin and flesh of the water honey peach is observed, and the maximum operating force in the group without obvious browning is recorded as the non-destructive operating force.

[0037] The mechanical property parameters of water honey peach measured by the test are: the average value of the static friction factor is 0.51, the average value of the non-destructive gripping force is 7.5N, and the average value of the non-destructive operating force is 3.5N.

[0038] The upper half-cone part of the lossless grasp model can be decomposed as Figure 12 , the normal line is the line perpendicular to the plane (the normal line is the z-axis as shown in Figure 12 ), and a is the included angle between the cone and the normal line.

[0039] The value range of a is ( When , the contact force is located on the conical surface of the friction cone, which is at the critical point of relative sliding; when , the contact force coincides with the normal line of the cone, and the grasp is the most stable.

[0040] To minimize the action force, the optimization objective function is constructed The multi-objective contact force optimization model of multi-finger lossless picking fruits and vegetables can be summarized as In the formula, represents the decision variable, ; is the picking force; is the external force screw balanced by the contact force; , are the maximum and minimum joint torque, respectively; is the constraint of the frictionless point contact model.

[0041] Based on the NSGA-II algorithm, the picking contact force is optimized to obtain the expected grasping force.

[0042] Step 3: Based on the dexterous hand dynamics and kinematics, the relationship matrix from the task space to the joint space of the dexterous finger tip is determined, the dexterous hand grasping fruits and vegetables is regarded as a virtual mass-spring-damper model of the second-order system, and the impedance control model in the Cartesian coordinate system is established.

[0043] The dynamics equation of the dexterous hand finger in the joint space is as follows: (15) In the formula, is the inertia matrix, is the Coriolis matrix, is the gravity moment vector, and are the joint torque and external torque vector, respectively.

[0044] From the dexterous hand kinematics model in step 1, the actual pose, velocity and acceleration of the finger tip in the Cartesian coordinate system are obtained as follows: (16) (17) (18) External torque and generalized external force vector , joint torque and fingertip torque are respectively: (19) (20) The dynamics model of the dexterous hand fingertip in Cartesian space can be obtained by integrating equations (15) to (20) as follows: (21) where the matrix from task space to joint space has the following relationship: (22) The impedance control model in Cartesian coordinates is: (23) where, , , are the desired inertia coefficient matrix, desired damping coefficient matrix and desired stiffness coefficient matrix of the impedance model respectively, is the actual pose of the fingertip, is the desired pose of the fingertip, is the actual grasping force, is the desired grasping force, is the grasping force error, .

[0045] The actual grasping force is measured by the membrane force sensor of the finger position, and the actual pose of the fingertip is obtained by the position sensor, and the control rate is: (24) The fruit and vegetable model is simplified as a spring system: (25) where, is the local normal stiffness of the fruit and vegetable, is the position of the fruit and vegetable skin when there is no contact.

[0046] The desired pose of the fingertip is obtained according to the desired grasping force (26) Step 4: Set the impedance control parameters in the impedance control model obtained in step 3, and then simulate the established impedance control model, including: setting one of the impedance control parameters, the expected inertia coefficient matrix, the expected damping coefficient matrix and the expected stiffness coefficient matrix as n different values, and the other two as constant values, obtaining n groups of impedance control parameters corresponding to the impedance control model under three simulation conditions, taking the actual fingertip pose, the expected pose of the fingertip and the expected grasping force as the input of the impedance control model, and obtaining the actual grasping force output by the impedance control model.

[0047] The impedance control model corresponding to each of the n groups of impedance control parameters under the three simulation conditions includes: First: simulate the stiffness coefficient K, set K from 500 N / m to 1000 N / m, and set M = 1 N·s 2 / m, B = 20 N·s / m, the local normal stiffness of fruits and vegetables is 1000 N / m, and the expected grasping force is 5 N.

[0048] Second: simulate the inertia coefficient M, set M from 1 N·s 2 / m to 5 N·s 2 / m, and set B = 20 N·s / m, K = 1000 N / m, the stiffness of fruits and vegetables is 1000 N / m, and the expected grasping force is 5 N.

[0049] Third: simulate the damping coefficient B, set B from 8 N·s / m to 50 N·s / m, and set M = 0.1 N·s 2 / m, K = 1000 N / m, the stiffness of fruits and vegetables is 1000 N / m, and the expected grasping force is 5 N.

[0050] It is worth noting that the stiffness coefficient K of fruits and vegetables is not equal to the local normal stiffness of fruits and vegetables , but they should be set to the same size.

[0051] The simulation results are as Figure 4 .

[0052] The stiffness coefficient is negatively related to the steady-state error of the grasping force. When the stiffness coefficient is larger, the actual pose of the finger tip is closer to the expected pose. Increasing the stiffness coefficient improves the ability of the finger to approach the expected pose and the expected grasping force, making the steady-state deviation smaller.

[0053] When the inertia coefficient takes a larger value, the grasping force rises slowly and the oscillation period is long. A smaller inertia coefficient helps the grasping force to follow the step command more quickly, shortens the transition time and makes the system more stable.

[0054] Too large damping coefficient sacrifices the response speed of grasping force, and too small damping coefficient leads to unstable oscillation of grasping force.

[0055] The stiffness coefficient K mainly affects the steady-state error of the system, and has little effect on the dynamic performance. A larger K helps to reduce the pressing depth and the grasping force, thereby reducing the risk of fruit and vegetable damage. Therefore, K is set to a larger fixed value (not dynamically adjusted). The inertia coefficient M and the damping coefficient B significantly affect the dynamic response performance of the system, and therefore the impedance control parameters that affect the stability of the grasping force are mainly the inertia coefficient and the damping coefficient . According to the impedance parameter simulation, candidate values of the impedance control parameters are obtained, including a candidate inertia coefficient = 1 N·s 2 / m, a candidate damping coefficient = 20 N·s / m, and a candidate stiffness coefficient = 1000 N / m.

[0056] Step 5: Compare the actual fingertip grasping force with the grasping force threshold to determine the contact state between the dexterous hand and the fruit and vegetable. If the actual grasping force is less than the grasping force threshold, it indicates that the dexterous hand is in free space, and the output grasping force error is set to 0; otherwise, the difference between the expected fingertip grasping force and the actual fingertip grasping force is compared by a comparator to obtain the grasping force error , and the rate of change of the grasping force error ; The fuzzy controller outputs the inertia coefficient change amount , the damping coefficient change amount according to the grasping force error and the error change rate ; the impedance control module adjusts the actual grasping force according to the outputs of the fuzzy controller and the expected pose, speed, and acceleration of the fingertip in the Cartesian space coordinates, the actual pose, speed, and acceleration of the fingertip in the Cartesian space coordinates, and the set stiffness coefficient K and expected grasping force. The system block diagram of the fuzzy variable impedance controller is as Figure 5 .

[0057] The above fuzzy controller outputs the inertia coefficient change amount , the damping coefficient change amount according to the grasping force error and the error change rate , including: 51) Set the grasping force error range to , the rate of change of the grasping force error to , the inertia coefficient change amount to , and the damping coefficient change amount to .

[0058] The input observation quantities are and ) and output control quantity ( and Each set is divided into seven fuzzy subsets: {NB, NM, NS, ZO, PS, PM, PB}. (Grasping force error) (Error rate of change) (Change in inertia coefficient) and The (dampening coefficient change) is fuzzified into a fuzzy subset: {negative large, negative medium, negative, zero, positive small, positive medium, positive large}, with corresponding records of {NB, NM, NS, ZO, PS, PM, PB}. 52) Input fuzzification: Fuzzifying the input data using the grasping force error. and the rate of change of gripping force error (or ) Fuzzification to obtain the membership degree on different fuzzy subsets; calculate Membership degree on 7 fuzzy subsets (27) calculate Membership degree on 7 fuzzy subsets (28) 53) The membership degree of the input observation is used for fuzzy inference according to the set fuzzy rules to obtain the fuzzy subsets corresponding to the changes in inertia coefficient and damping coefficient; In this embodiment, based on the simulation results of the impedance control parameters obtained in step 4, fuzzy control rules for the impedance control parameters based on force control are formulated as shown in Tables 3 and 4.

[0059] (1) When and When it is large, Smaller size to improve response speed.

[0060] (2) When Larger but When smaller, Larger size to enhance stability.

[0061] (3) When and When approaching zero, .

[0062] (4) When or When it is large, Larger, to suppress oscillations.

[0063] (5) When and When smaller, Smaller size to improve response speed.

[0064] Table 3. Fuzzy Rule Table for Changes in Inertia Coefficient Table 4. Fuzzy Rule Table for Damping Coefficient Change Rule format: (29) That is: when the grasping force error e f Belongs to fuzzy sets And the rate of change of error Belongs to fuzzy set B h When the change in the output inertia coefficient ΔM belongs to the fuzzy set... Meanwhile, the change in damping coefficient ΔB belongs to a fuzzy set. ; 54) Based on the candidate inertia coefficient and candidate damping coefficient Calculate the coefficient of inertia and damping coefficient The stiffness coefficient K will be set as the candidate stiffness coefficient. ; The solutions for M and B are: (30) (31) Set the stiffness coefficient K = 1000 N / m. 55) Stiffness coefficient of fruits and vegetables Expected grasping power , Given that the desired pose can be obtained according to equation (26), After obtaining the desired pose value, the actual pose for controlling the dexterity hand movement adjustment can be solved by inverse kinematics of the dexterity hand. The impedance controller takes the inertia coefficient and damping coefficient output by the fuzzy controller, as well as the desired position, velocity, and acceleration in Cartesian space, and the fingertip position, velocity, and acceleration in Cartesian space, as inputs. Using dexterous hand kinematics, it adjusts the actual grasping force of the dexterous hand to approximate the desired grasping force. .

[0065] By traversing all combinations of input variables, inference and defuzzification are performed based on the fuzzy rule base, and the mapping relationship between input and output is plotted as a three-dimensional response surface, resulting in the three-dimensional response diagram of the fuzzy controller, as shown below. Figure 13 .

[0066] Triangle membership degree and linguistic variables, such as Figure 6 .

[0067] It should be noted that the embodiments described herein are merely exemplary of the spirit of the present application. A person skilled in the art of the present application can make various modifications or supplements to the described embodiments or use similar ways instead, without departing from the spirit of the present application or exceeding the scope defined by the appended claims.

Claims

1. A dexterous hand fruit and vegetable grasping control method based on fuzzy variable impedance control, characterized in that, Includes the following steps: Step 1: Construct a kinematic model of dexterous hand grasping fruits and vegetables, establish a multi-finger cooperative kinematic model based on spinor theory, and determine the pose of each fingertip in the base coordinate system, the grasping force of each finger, and the distribution of the resultant force of multiple fingers; Step 2: Based on force closure, construct a multi-objective gripping force optimization model with minimum contact force and minimum deviation between contact force and normal direction as the criterion. Solve the model using the NSGA-II algorithm to obtain the minimum gripping force required for stable and non-destructive gripping. Step 3: Establish an impedance control model in Cartesian space, and then simulate the established impedance control model to obtain candidate values ​​for impedance control parameters, including candidate inertia coefficients. Candidate damping coefficient Candidate stiffness coefficients ; Step 4: Introduce fuzzy control, using the grasping force tracking error and its rate of change as input, and outputting the inertia adjustment ΔM and damping adjustment ΔB, to construct a fuzzy adaptive variable impedance controller based on candidate inertia coefficients. and candidate damping coefficient Calculate the inertia coefficient of the fuzzy adaptive variable impedance controller and damping coefficient The stiffness coefficient K of the fuzzy adaptive variable impedance controller will be set as the candidate stiffness coefficient. Fuzzy adaptive variable impedance controller based on inertia coefficient and damping coefficient It uses the desired pose, velocity, and acceleration of the fingertip in Cartesian space coordinates, the actual pose, velocity, and acceleration of the fingertip in Cartesian space coordinates, and the set stiffness coefficient K and desired gripping force as inputs to achieve precise tracking of the gripping force.

2. The method according to claim 1, characterized in that, In step 1, a single-finger kinematic model is established using spinor theory, including: establishing the transformation relationship between the base coordinate system of each finger and the global coordinate system; using the exponential product formula to represent the fingertip pose; calculating the spatial Jacobian matrix of each finger; and establishing the mapping relationship between joint torque and contact force.

3. The method according to claim 1, characterized in that, In step 2, the gripping force optimization model includes the following constraints: force balance constraint, friction cone constraint, and joint torque constraint.

4. The method according to claim 1, characterized in that, In step 3, the impedance control model is as follows: In the formula, , , These are the expected inertia coefficient matrix, expected damping coefficient matrix, and expected stiffness coefficient matrix of the impedance model, respectively. This represents the actual pose of the robotic arm's end effector. The desired pose of the robotic arm's end effector. For actual grasping power, For the desired grasping force.

5. The method according to claim 4, characterized in that, The candidate inertia coefficient and candidate damping coefficient The desired stiffness coefficient matrix K represents the key adjustment parameters determined through simulation analysis. , .

6. The method according to claim 1, characterized in that, In step 4, fuzzy control is introduced, taking the grasping force tracking error and its rate of change as input, and outputting the inertial adjustment amount ΔM and the damping adjustment amount ΔB, including: Grasping force error and error change rate The linguistic variables are fuzzed using the triangular membership function: {NB, NM, NS, ZO, PS, PM, PB}. Based on a preset fuzzy rule table, fuzzy inference is performed to obtain fuzzy outputs of the inertia change ΔM and the damping change ΔB; the fuzzy rule is: when the gripping force error e f Belongs to fuzzy sets And the rate of change of error Belongs to fuzzy set B h When the change in the output inertia coefficient ΔM belongs to the fuzzy set... Meanwhile, the change in damping coefficient ΔB belongs to a fuzzy set. ; Fuzzy set in the fuzzy rule table of inertia coefficient change The row corresponding to the middle subset and the fuzzy set B h The membership degree of a subset whose columns are orthogonal is a fuzzy set. A subset of; Fuzzy set in the fuzzy rule table of damping coefficient change The row corresponding to the middle subset and the fuzzy set B h The membership degree of a subset whose columns are orthogonal is a fuzzy set. A subset of; The fuzzy rule table for the change in inertia coefficient is as follows: The fuzzy rule table for the change in damping coefficient is as follows: 。 7. A dexterous hand control system for implementing the method according to any one of claims 1-6, characterized in that, include: The kinematic modeling module is used to construct dexterous hand grasping models based on spinor theory. An optimized calculation module is used to solve for the minimum gripping force using the NSGA-II algorithm; Impedance control module is used to achieve initial tracking of the gripping force; The fuzzy inference module is used to dynamically adjust the impedance parameters M and B based on the grasping force error and the rate of change.

8. A fruit and vegetable harvesting robot, characterized in that, It includes the dexterous hand control system as described in claim 7.

Images