Autonomously damped drop-type fruit harvesting platform and control method

Abstract

The application discloses a kind of independently damped falling fruit harvesting platform and control method;The harvesting platform includes mobile chassis and 4RRS rigid-flexible coupling parallel mechanism;The 4RRS rigid-flexible coupling parallel mechanism includes net stretching mechanical arm assembly and flexible harvesting net;The net stretching mechanical arm assembly includes multiple groups of harvesting mechanical arms and drive units;The end effector of each group of harvesting mechanical arms is connected with flexible collection net;Each group of drive units includes brushless DC motor, drive board, rotary encoder and current / voltage sensor;When fruit falls from high altitude onto flexible harvesting net, its impact force will be locally rigidized through flexible harvesting net and transmitted to harvesting mechanical arm;The control method of the independently damped falling fruit harvesting platform of the application utilizes 4RRS rigid-flexible coupling parallel mechanism, cooperates with fruit impact force identification and autonomous damping control algorithm, so that falling fruit can be stably, quickly and non-destructively harvested, the success rate and non-destructive rate of harvesting are improved, and labor intensity and cost are reduced.

Description

Technical Field

[0001] This invention belongs to the field of automated fruit harvesting, specifically relating to an autonomous vibration-damping drop-type fruit harvesting platform and control method. Background Technology

[0002] Drop harvesting is a method of harvesting fruits or medicinal plants by allowing them to fall naturally when ripe. This method is primarily used for varieties whose fruits easily detach from branches after ripening and can withstand a certain amount of impact, such as camellia oleifera, durian, coconut, and jackfruit. Taking durian drop harvesting as an example, in recent years, the durian industry has expanded rapidly due to market demand, e-commerce, and policy support, further increasing the demand for automated durian harvesting.

[0003] Durian, a unique tropical fruit, is renowned for its distinctive appearance and rich aroma. It is an irregularly shaped sphere, typically 10-20cm in length, with a dark green or yellowish-brown rind covered in numerous raised spikes of varying size and density. These spikes contribute to the difficulty of harvesting. Currently, there are no dedicated durian harvesting machines on the market, and existing harvesting methods generally fall into two categories: natural fall and manual picking. Both methods have the following drawbacks:

[0004] Natural fall method: This method utilizes the characteristic of ripe durians falling naturally. However, because ripe durians are quite heavy, direct impact can damage the flesh. Therefore, a large area of ​​netting or blankets needs to be laid under the plant as a passive vibration damping method (similar methods are used for harvesting camellia oleifera fruit) to catch durians that fall randomly from the tree and prevent damage. However, laying large areas of netting is costly and requires specific site conditions, making it difficult to popularize among small and medium-sized growers.

[0005] Manual harvesting: This method requires two people. One person climbs up the durian tree and cuts off the ripe durians. The other person stands below the tree and catches the falling durians with burlap. This method requires a high level of skill and extensive experience, and it is easy for the durians to be broken or damaged. Furthermore, the sharp thorns of the durians pose a significant safety hazard to the harvesters below the tree.

[0006] Therefore, in the process of harvesting durians, whether it is the passive vibration reduction of the natural falling method or the active vibration reduction of the manual picking method, the harvested fruit needs to be vibration-damped. Therefore, how to achieve efficient and safe vibration-damped harvesting under the plant is the key issue that limits the yield, transportation and preservation of fallen fruit and the safety of personnel. Summary of the Invention

[0007] To overcome the shortcomings of existing technologies, this invention provides a self-damping drop-type fruit harvesting platform. The drop-type fruit harvesting platform can identify the magnitude and direction of the impact force when the fruit falls in real time and achieve autonomous active vibration reduction, thereby achieving damage-free fruit harvesting.

[0008] The second objective of this invention is to provide a control method for an autonomous vibration-damping drop-type fruit harvesting platform.

[0009] The technical solution of the present invention to solve the above-mentioned technical problems is:

[0010] A self-vibration-damping drop-type fruit harvesting platform includes a mobile chassis and a 4RRS rigid-flexible coupling parallel mechanism mounted on the mobile chassis. The 4RRS rigid-flexible coupling parallel mechanism includes a net-stretching robotic arm assembly and a flexible harvesting net. The net-stretching robotic arm assembly includes multiple sets of harvesting robotic arms and multiple sets of drive units. The end effector of each set of harvesting robotic arms is connected to the flexible harvesting net. Each drive unit includes a brushless DC motor and a drive board for controlling the brushless DC motor. The drive board is equipped with a rotary encoder and a current / voltage sensor. The current / voltage sensor is connected to the armature winding of the brushless DC motor, and calculates the motor torque of the brushless DC motor in real time by measuring the current / voltage and utilizing the functional relationship between current / voltage and torque. The rotary encoder is coaxially connected to the output shaft of the brushless DC motor, and is used to detect the rotation angle, speed, and acceleration of the brushless DC motor in real time.

[0011] Preferably, the harvesting robotic arms are in four groups, and the end effectors of the four groups of harvesting robotic arms are respectively connected to the four diagonals of the flexible harvesting net.

[0012] Preferably, each set of harvesting robotic arms is a two-degree-of-freedom rotary robotic arm; the two-degree-of-freedom rotary robotic arm includes a base, a lower swing arm, an upper swing arm, and an end effector disposed on the upper swing arm, wherein a lower joint is disposed between the base and the lower swing arm; an upper joint is disposed between the lower swing arm and the upper swing arm; correspondingly, each set of harvesting robotic arms has two sets of drive units, the two sets of drive units are respectively located at the upper joint and the lower joint of the two-degree-of-freedom rotary robotic arm, and are used to drive the swing of the upper swing arm and the lower swing arm in the two-degree-of-freedom rotary robotic arm.

[0013] Preferably, the flexible harvesting net is a square flexible coarse linen net, and the edges of the flexible harvesting net are sealed; each of the four opposite corners of the flexible harvesting net is provided with a connecting part that connects to the end effector of the harvesting robotic arm.

[0014] Preferably, the mobile chassis is a tracked mobile chassis.

[0015] A control method for a drop-type fruit harvesting platform with autonomous vibration reduction includes the following steps:

[0016] Step S1: The moving chassis transports the 4RRS rigid-flexible coupling parallel mechanism to the plant where the fruit to be harvested is to be placed; then, the driving units control the movement of each group of harvesting robotic arms to adjust the harvesting robotic arms to the initial state, thereby adjusting the posture of the flexible harvesting net to the pre-harvest state.

[0017] Step S2: When the fruit falls into the flexible harvesting net, the flexible harvesting net transmits the impact force of the falling fruit to the harvesting robot arm, causing the joints of the harvesting robot arm to swing, thereby causing the armature winding of the brushless DC motor in the drive unit to generate current / voltage changes, and the current / voltage change values ​​are collected by the current / voltage sensor.

[0018] Step S3: Construct a mathematical model for fruit impact force identification. Input the current / voltage change values ​​collected by the current / voltage sensor into the constructed mathematical model for fruit impact force identification. Obtain the magnitude and direction of the fruit impact force through the mathematical model for fruit impact force identification.

[0019] Step S4: The control unit formulates a corresponding vibration reduction harvesting strategy based on the magnitude and direction of the fruit impact force. Based on the formulated vibration reduction harvesting strategy, it generates corresponding control commands and uses these commands to control the movement trajectory of each group of harvesting robotic arms through the drive unit to drive the flexible harvesting net to move.

[0020] Step S5: After the vibration reduction harvesting of the fruit is completed, each group of drive units controls the coordinated movement of each group of harvesting robotic arms to make the flexible harvesting net pour the harvested fruit into the harvesting frame.

[0021] Step S6: Each group of drive units controls the movement of each group of harvesting robotic arms to return to the initial state. Then, steps S2-S5 are repeated to start the next round of fruit harvesting.

[0022] Preferably, in step S1, the harvesting robotic arms are in four groups, and the end effectors of the four groups of harvesting robotic arms are connected to the four diagonals of the flexible harvesting net; each group of harvesting robotic arms is a two-degree-of-freedom rotary robotic arm; the two-degree-of-freedom rotary robotic arm includes a base, a lower swing arm, an upper swing arm, and an end effector disposed on the upper swing arm, wherein a lower joint is disposed between the base and the lower swing arm; and an upper joint is disposed between the lower swing arm and the upper swing arm.

[0023] Preferably, in step S3, the step of identifying the location of the fruit impact force using the fruit impact force identification mathematical model is as follows:

[0024] Step S301: Using the center of the flexible harvesting net as the origin of the coordinate system, divide it into the first quadrant region, the second quadrant region, the third quadrant region, and the fourth quadrant region; then number the four sets of harvesting robotic arms.

[0025] Step S302: Compare the changes in current / voltage of the brushless DC motor in the harvesting robot arm collected by the current / voltage sensors in each group of drive units, determine the harvesting robot arm with the largest change in current / voltage, and determine the quadrant area where the fruit fell in the flexible harvesting net according to the number of the harvesting robot arm.

[0026] Preferably, in step S3, the process by which the mathematical model for identifying fruit impact force identifies the magnitude of the fruit impact force is as follows:

[0027] Step 311: During the fruit impact flexible harvesting net process, record the current / voltage data of the brushless DC motors in the upper and lower joints of the four sets of harvesting robotic arms in real time.

[0028] Step 312: In the recorded current / voltage data, find the moment when the current / voltage of the brushless DC motor in the upper and lower joints of each harvesting robot first reaches its extreme value, obtain the current / voltage value at that moment, and convert it into the corresponding motor torque according to the functional relationship between current / voltage and torque.

[0029] Step 313: Obtain the rotation angle of the brushless DC motors in the upper and lower joints of the harvesting robot arm through a rotary encoder, thereby obtaining the posture of the harvesting robot arm. Combined with the motor torque obtained in step S312, calculate the force borne by the end effector of each group of harvesting robots arm. By vector summing the forces borne by the end effectors of each group of harvesting robots arm, obtain the magnitude of the fruit impact force.

[0030] Preferably, in step S4, the specified steps of the vibration reduction harvesting strategy are as follows:

[0031] Step S401: Record current or voltage data in real time that are the same as the joint torque changes of the harvesting robotic arm;

[0032] Step S402: Based on the recorded current or voltage data, generate corresponding waveforms to detect the frequency and amplitude of the current or voltage in real time, thereby obtaining the frequency and amplitude of the vibration experienced by the four sets of harvesting robotic arms;

[0033] Step S403: The vibrations experienced by the four sets of harvesting robotic arms will affect the motor torque of the brushless DC motors in the upper and lower joints of each set of harvesting robotic arms. Taking the resulting motor torque as a known condition, the position, angular velocity, and angular acceleration of the upper and lower joints of each set of harvesting robotic arms are calculated in real time using inverse kinematics. Combined with the inverse dynamics equation of the second kind of Lagrange equation, the optimal counteracting torque that the upper and lower joints of each set of harvesting robotic arms need to generate to counteract the motor torque is calculated in real time. At the same time, a control command corresponding to the optimal counteracting torque is generated and sent to each drive unit of each set of harvesting robotic arms to cause the brushless DC motors at the upper and lower joints of each set of harvesting robotic arms to output the calculated optimal counteracting torque, thereby driving the flexible harvesting net to move and reduce or counteract the vibration energy of the fruit falling onto the flexible harvesting net.

[0034] Compared with existing technologies, the self-vibration-damping drop-type fruit harvesting platform of the present invention has the following advantages:

[0035] 1. Less damage to the fruit: The fruit enters the flexible harvesting net by falling. The flexible harvesting net has passive flexibility, while the 4RRS rigid-flexible coupling parallel mechanism can identify the magnitude and direction of the impact force of the fruit. Then, the control algorithm realizes active vibration reduction. The whole harvesting method is a combination of active and passive vibration reduction. This can quickly absorb and disperse the impact force of the fruit in a short time, thereby ensuring that the impact force of the falling fruit is minimized and thus avoiding damage to the fruit.

[0036] 2. Enhanced safety performance: Unlike traditional mechanical harvesting devices and manual harvesting methods, the harvesting method adopted in this invention requires no human intervention, fundamentally eliminating the risk of fruit falling and injuring people, thus ensuring the safety of the harvesting process.

[0037] 3. Higher degree of automation: The mobile chassis enables movement and navigation, and is combined with a rotary encoder, current / voltage sensor and brushless DC motor to form an electromechanical control system, thereby realizing real-time negative feedback closed-loop regulation. This design enables the autonomous vibration reduction falling fruit harvesting platform of this invention to automatically complete the harvesting of fruits. The harvesting process is fully automated, thereby greatly improving the harvesting efficiency.

[0038] 4. Enhanced adaptability: The self-vibration-damping drop-type fruit harvesting platform of this invention can adapt to fruits of different sizes, weights, and growth heights, as well as fruits with complex shapes and outlines, demonstrating strong adaptability to harvesting different types of fruits. Furthermore, the equipped mobile chassis allows for flexible movement in various terrains, providing all-terrain capability. This gives the self-vibration-damping drop-type fruit harvesting platform of this invention excellent terrain adaptability and fruit harvesting adaptability. Attached Figure Description

[0039] Figure 1 This is an overall structural diagram of the autonomous vibration-damping drop-type fruit harvesting platform of the present invention.

[0040] Figure 2 This is a schematic diagram of the harvesting robotic arm and its drive unit.

[0041] Figure 3 This is a schematic diagram of a mobile chassis.

[0042] Figure 4 This is a flowchart illustrating the control method of the autonomous vibration reduction drop-type fruit harvesting platform of the present invention.

[0043] Figure 5 This is a flowchart of the harvesting method of the autonomous vibration-damping drop-type fruit harvesting platform of the present invention.

[0044] Figure 6 This is an equivalent diagram of the localized stiffening effect of a flexible harvesting net.

[0045] In the diagram: 1-Mobile chassis; 2-Drive unit; 3-Harvesting robotic arm; 4-Flexible harvesting net; 5-Controller; A, B, C, and D are the numbers of the harvesting robotic arm. Detailed Implementation

[0046] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0047] See Figures 1-6 The self-vibration-damping falling fruit harvesting platform of the present invention includes a mobile chassis and a 4RRS rigid-flexible coupling parallel mechanism disposed on the mobile chassis.

[0048] See Figures 1-6 The mobile chassis is connected to the 4RRS rigid-flexible coupling parallel mechanism and is used to drive the movement and navigation of the 4RRS rigid-flexible coupling parallel mechanism. The mobile chassis is a tracked chassis and can be implemented with reference to existing tracked mobile chassis on the market. In this embodiment, the mobile chassis includes a chassis shell, an all-terrain rubber track assembly, and a DC drive motor. The all-terrain rubber track assembly is installed outside the chassis shell, and the DC drive motor is installed inside the chassis shell. The all-terrain rubber track assembly moves under the drive of the DC drive motor.

[0049] Therefore, in this embodiment, the mobile chassis is divided into inner and outer parts by the chassis shell; the chassis shell of the mobile chassis is made of steel, aluminum alloy and carbon fiber to ensure that the structural strength and weight of the self-vibration-damping falling fruit harvesting platform of the present invention are moderate.

[0050] The mobile chassis is externally connected to two rubber tracks that can rotate 360 ​​degrees horizontally. These rubber tracks are more space-saving than high-manganese steel tracks. In addition, the rubber material is elastic, which can play a certain role in shock absorption. It also has good grip, is not easy to slip, is more wear-resistant, and has good comfort performance, which can further assist in the completion of autonomous shock absorption.

[0051] The mobile chassis houses a DC drive motor with a stainless steel cover; the all-terrain rubber track assembly is connected to the DC drive motor to realize the electric drive movement function of the autonomous vibration-damping drop-type fruit harvesting platform of the present invention.

[0052] See Figures 1-6 The 4RRS rigid-flexible coupling parallel mechanism includes a net-stretching robotic arm assembly and a flexible harvesting net. The net-stretching robotic arm assembly is used to realize the precise movement and autonomous vibration reduction function of the harvesting robotic arm. The flexible harvesting net is connected to the end of the net-stretching robotic arm assembly and can absorb and disperse the impact force of falling fruit during the harvesting process.

[0053] See Figures 1-6 The web-shearing robotic arm assembly includes a harvesting robotic arm and a drive unit.

[0054] See Figures 1-6 The harvesting robotic arms consist of four sets, and the end effectors of the four sets of harvesting robotic arms are respectively connected to the four diagonals of the flexible harvesting net. In this embodiment, each set of harvesting robotic arms is a two-degree-of-freedom rotary robotic arm. Correspondingly, each set of harvesting robotic arms has two sets of drive units, which are located on two joints of the two-degree-of-freedom rotary robotic arm and are used to drive the swing of the two arms in the two-degree-of-freedom rotary robotic arm.

[0055] See Figures 1-6 The flexible harvesting net is a square flexible coarse linen net. The flexible harvesting net is edge-sealed and has connection parts at each of its four opposite corners that connect to the end effector of the harvesting robotic arm.

[0056] In this embodiment, the flexible harvesting net is made of high-strength, high-wear-resistant polyester fiber and nylon. It can withstand long-term use and frequent washing, ensuring its durability and stability. The softness of the material can be appropriately adjusted according to the size and shape of the fruit, thereby protecting the quality and value of the fruit. It is made of environmentally friendly materials, which meets the requirements of modern society for environmental protection. The lightweight material makes it easy to carry and move.

[0057] See Figures 1-6The drive unit includes a brushless DC motor, a rotary encoder, and a current / voltage sensor. The rotary encoder is coaxially connected to the output shaft of the brushless DC motor and is used to measure and feedback the position, speed, and acceleration of the brushless DC motor in real time. The current / voltage sensor is connected to the armature winding of the brushless DC motor and is used to monitor the motor torque of the brushless DC motor in real time.

[0058] The rotary encoder uses the angle of the harvesting robot arm as the data to drive and change the working magnetic field of the brushless DC motor, thereby driving the brushless DC motor to produce a nonlinear damping effect. The harvesting robot arm acts as the execution device, and the signals generated by both are transmitted through a negative feedback system. Based on the changes in the actions of the rotary encoder and the current / voltage sensor, the energy brought by the impact force of the fruit is finally converted into the heat energy generated by the brushless DC motor under nonlinear damping and the mechanical energy of the flexible harvesting net, thereby achieving autonomous vibration reduction and damage-free fruit harvesting.

[0059] See Figures 1-6 The brushless DC motor is equipped with a torque sensing and vibration reduction control algorithm based on FOC control. The FOC control algorithm precisely controls the magnitude and direction of the magnetic field, making the motor's motion torque stable, with low noise, high efficiency, and high-speed dynamic response.

[0060] See Figures 1-6 The rotary encoder is a non-contact magnetic encoder, which has the advantages of high sensing position accuracy, vibration resistance, shock resistance, high protection level and long service life, and has a wider magnetic field strength sensing range.

[0061] See Figures 1-6 The current / voltage sensor samples the data through a precision resistor and converts it into voltage via I / V. The voltage signal is then processed by the drive unit via the A / DC module and sent to the controller as input parameters for the fruit impact force identification mathematical model, used to calculate the magnitude and direction of the fruit impact force.

[0062] See Figures 1-6The impact force of the falling fruit is transmitted to the output shaft of the brushless DC motor through the flexible harvesting net and the harvesting robotic arm, causing a change in the magnetic field of the armature winding of the brushless DC motor, which in turn causes a change in the current / voltage. The magnitude and orientation of the fruit impact force and the orientation of the flexible harvesting net are different, and the changes in the current / voltage of the armature winding of the brushless DC motor at the joint of each set of harvesting robotic arms are also different. Therefore, by using the current output torque equation of the armature winding of the brushless DC motor, combined with the mechanical transmission characteristics of the flexible harvesting net, and combined with the 4RRS rigid-flexible coupling parallel mechanism, a mathematical model for fruit impact force identification can be established. This model can identify the magnitude and orientation of the fruit impact force without the addition of additional sensors, which can be used to achieve precise autonomous vibration reduction control in the future.

[0063] See Figures 1-6 The control method for the autonomous vibration reduction falling fruit harvesting platform of the present invention includes the following steps:

[0064] Step S1: The mobile chassis transports the 4RRS rigid-flexible coupling parallel mechanism to the plant where the fruit to be harvested is to be placed; the DC drive motor of the mobile chassis is self-locked to achieve overall fixation; the harvesting robotic arm on the 4RRS rigid-flexible coupling parallel mechanism is driven by a brushless DC motor to adjust the harvesting robotic arm to the initial state, thereby adjusting the posture of the flexible harvesting net to the preparatory form before harvesting, and preparing for vibration reduction in the fruit drop harvesting.

[0065] Step S2: When the fruit falls into the flexible harvesting net, the flexible harvesting net transmits the impact force of the falling fruit to the harvesting robotic arm, causing the joints of the harvesting robotic arm to swing. At this time, the joint torque of the four harvesting robotic arms changes differently, causing the current / voltage generated by the armature winding of the brushless DC motor to change, and the change value of current / voltage is collected by the current / voltage sensor.

[0066] Step S3: Construct a mathematical model for fruit impact force identification. Input the current / voltage change values ​​collected by the current / voltage sensor into the constructed mathematical model for fruit impact force identification. The mathematical model for fruit impact force identification analyzes these current / voltage change values ​​and can identify the magnitude and direction of the fruit impact force without adding additional sensors.

[0067] Step S4: The controller formulates a corresponding vibration reduction harvesting strategy based on the magnitude and direction of the fruit impact force. According to the formulated vibration reduction harvesting strategy, it generates control commands and uses them to control the movement trajectory of each group of harvesting robotic arms through the drive unit, thereby driving the flexible harvesting net to move, so as to reduce or offset the vibration impact of the fruit falling on the flexible harvesting net, thereby achieving fruit drop-type non-destructive harvesting.

[0068] Step S5: After the vibration-damping harvesting of the fruit is completed, the drive unit controls the coordinated movement of each group of harvesting robotic arms to cause the flexible harvesting net to pour the harvested fruit into the harvesting frame.

[0069] Step S6: The drive unit controls the movement of each group of harvesting robotic arms to return to the initial state, and then repeats steps S2-S5 to start the next round of fruit harvesting.

[0070] See Figures 1-6 The principle and construction method of the mathematical model for fruit impact force identification include the following steps:

[0071] When the fruit falls, the flexible harvesting net deforms upon impact. This deformation causes localized stiffening of the net, effectively creating four sets of robotic arms at the four corners, each holding the falling fruit with ropes. Simultaneously, the deformation of the net transmits the impact force to the robotic arms and their drive units, causing variations in the torque of the upper and lower joints. The load on each robotic arm differs depending on where the fruit falls within the net, resulting in different torque variations in the upper and lower joints. The variations in joint torque across the four robotic arms at different fruit positions within the net are shown in the table below.

[0072]

[0073] As shown in the table, when the fruit falls in the exact center of the flexible harvesting net, the magnitude and trend of the joint torque of the four harvesting robotic arms are basically the same. When the fruit falls in the first, second, third, and fourth quadrants of the flexible harvesting net, the robotic arm closest to the fruit's fall point has the largest joint torque and the fastest response. In turn, the two robotic arms closest to the fruit's fall point have basically the same joint torque magnitude and trend. However, the two robotic arms furthest from the fruit's fall point have the slowest response to joint torque changes.

[0074] When the fruit falls onto the x-axis or y-axis of the flexible harvesting net, the joint torque magnitude and trend of the two harvesting robotic arms closest to the fruit's fall point are basically the same; while the joint torque magnitude and trend of the two harvesting robotic arms farther from the fruit's fall point are also basically the same; however, the joint torque of the two harvesting robotic arms closest to the fruit's fall point changes faster and experiences greater torque variation.

[0075] When the joint torque of the harvesting robot changes, the brushless DC motor at the joint of the harvesting robot will generate current or voltage accordingly. The changes in these currents or voltages are the same as the changes in the joint torque of the harvesting robot. Therefore, the position of the fruit falling into the flexible harvesting net can be deduced from the changes in current / voltage generated by the upper and lower joints of the harvesting robot, thus forming a mathematical model for fruit impact force identification.

[0076] Therefore, the process by which the mathematical model for identifying fruit impact force identifies the location of the fruit impact force is as follows:

[0077] Step S301: Using the center of the flexible harvesting net as the origin of the coordinate system, divide it into the first quadrant region, the second quadrant region, the third quadrant region, and the fourth quadrant region; then number the four sets of harvesting robotic arms.

[0078] Step S302: Compare the changes in current / voltage in each group of harvesting robotic arms collected by the current / voltage sensors in the drive unit, determine the harvesting robotic arm with the largest change in current / voltage, and determine the quadrant area where the fruit fell in the flexible harvesting net according to the number of the harvesting robotic arm.

[0079] Furthermore, the process by which the mathematical model for identifying fruit impact force identifies the magnitude of the impact force is as follows:

[0080] During the contact between the falling fruit and the flexible harvesting net, the net uses its elasticity to cushion the impact. When the fruit's velocity reaches zero, the impact force is essentially converted into the elastic force of the net. This elastic deformation force is then transmitted to the end effectors of the four harvesting robotic arms through the net's deformation. Furthermore, this elastic deformation force continues to be transmitted to the brushless DC motors in the upper and lower joints of the robotic arms, resulting in changes in the torque of the upper and lower joints. Based on this chain reaction, conversely, the above... The process yields a mathematical model for identifying fruit impact force. Since there is a functional relationship between the current / voltage and the torque of the brushless DC motor, the current / voltage of the brushless DC motor can be measured first to obtain the torque experienced by the brushless DC motor. This torque can then be converted back to the force experienced by the end effector of the harvesting robot arm. This allows us to obtain the magnitude of the force experienced by the end effectors of the four sets of harvesting robot arms. Combined with the force spectrum model of the flexible harvesting net, the magnitude of the fruit impact force can be obtained, and the location of its action can also be determined based on the distribution of the fruit impact force.

[0081] See Figures 1-6 The specified steps of the vibration reduction and recovery strategy are as follows:

[0082] Step S401: Record current or voltage data in real time that are the same as the joint torque changes of the harvesting robotic arm;

[0083] Step S402: Based on the recorded current or voltage data, generate corresponding waveforms to detect the frequency and amplitude of the current or voltage in real time, thereby obtaining the frequency and amplitude of the vibration experienced by the four sets of harvesting robotic arms;

[0084] Step S403: The vibrations experienced by the four sets of harvesting robotic arms will affect the DC brushless motors in the upper and lower joints of each set of harvesting robotic arms, thus using the resulting motor torque effect as a known condition, the position θ and angular velocity of the upper and lower joints in each set of harvesting robotic arms are calculated in real time using inverse kinematics. and angular acceleration By combining the inverse dynamics equations of the second kind of Lagrange equations, the optimal counteracting torque required by the upper and lower joints of each harvesting robot arm to counteract the influence of the motor torque is calculated in real time. At the same time, control commands corresponding to the optimal reverse torque are generated and sent to the drive units of each group of harvesting robotic arms to cause the brushless DC motors at the upper and lower joints of each group of harvesting robotic arms to output the calculated optimal reverse torque, thereby driving the flexible harvesting net to move, thereby reducing or offsetting the vibration energy of the fruit falling on the flexible harvesting net.

[0085] See Figures 1-6 The control method of the autonomous vibration reduction falling fruit harvesting platform of the present invention utilizes a 4RRS rigid-flexible coupling parallel mechanism, combined with fruit impact force identification and autonomous vibration reduction control algorithm, so as to stably, quickly and without damage harvesting falling fruit, improve the harvesting success rate and damage-free rate, reduce labor intensity and cost, and can be used for the non-destructive automated harvesting of fruits with high growth and large weight, such as camellia fruit, coconut, durian, and jackfruit.

[0086] The above are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above content. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A control method for a self-vibration-damping drop-type fruit harvesting platform, characterized in that, The autonomous vibration-damping drop-type fruit harvesting platform includes a mobile chassis and a 4RRS rigid-flexible coupling parallel mechanism mounted on the mobile chassis. The 4RRS rigid-flexible coupling parallel mechanism includes a net-stretching robotic arm assembly and a flexible harvesting net. The net-stretching robotic arm assembly includes multiple sets of harvesting robotic arms and multiple sets of drive units. The end effector of each set of harvesting robotic arms is connected to the flexible harvesting net. Each drive unit includes a brushless DC motor and a drive board for controlling the brushless DC motor. The drive board is equipped with a rotary encoder and a current / voltage sensor. The current / voltage sensor is connected to the armature winding of the brushless DC motor, and calculates the motor torque in real time by measuring the current / voltage and utilizing the functional relationship between current / voltage and torque. The rotary encoder is coaxially connected to the output shaft of the brushless DC motor, and is used to detect the rotation angle, speed, and acceleration of the brushless DC motor in real time. The specific method includes the following steps: Step S1: The moving chassis transports the 4RRS rigid-flexible coupling parallel mechanism to the plant where the fruit to be harvested is to be placed; then, the driving units control the movement of each group of harvesting robotic arms to adjust each group of harvesting robotic arms to the initial posture, thereby adjusting the posture of the flexible harvesting net to the pre-harvest state. Step S2: When the fruit falls into the flexible harvesting net, the flexible harvesting net transmits the impact force of the falling fruit to the harvesting robot arm, causing the joints of the harvesting robot arm to swing, thereby causing the armature winding of the brushless DC motor in the drive unit to generate current / voltage changes, and the current / voltage change values ​​are collected by the current / voltage sensor. Step S3: Construct a mathematical model for fruit impact force identification. Input the current / voltage change values ​​collected by the current / voltage sensor into the constructed mathematical model for fruit impact force identification. Obtain the magnitude and direction of the fruit impact force through the mathematical model for fruit impact force identification. Step S4: The control unit formulates a corresponding vibration reduction harvesting strategy based on the magnitude and direction of the fruit impact force. Based on the formulated vibration reduction harvesting strategy, it generates corresponding control commands and uses these commands to control the movement trajectory of each group of harvesting robotic arms through the drive unit to drive the flexible harvesting net to move. Step S5: After the vibration reduction harvesting of the fruit is completed, each group of drive units controls the coordinated movement of each group of harvesting robotic arms to make the flexible harvesting net pour the harvested fruit into the harvesting frame. Step S6: Each group of drive units controls the movement of each group of harvesting robotic arms to return to the initial state. Then, steps S2-S5 are repeated to start the next round of fruit harvesting.

2. The control method for the autonomous vibration-damping drop-type fruit harvesting platform according to claim 1, characterized in that, The harvesting robotic arms consist of four sets, and the end effectors of the four sets of harvesting robotic arms are respectively connected to the four diagonals of the flexible harvesting net.

3. The control method for the autonomous vibration-damping drop-type fruit harvesting platform according to claim 1, characterized in that, Each set of harvesting robotic arms is a two-degree-of-freedom rotary robotic arm; the two-degree-of-freedom rotary robotic arm includes a base, a lower swing arm, an upper swing arm, and an end effector disposed on the upper swing arm, wherein a lower joint is disposed between the base and the lower swing arm; an upper joint is disposed between the lower swing arm and the upper swing arm; correspondingly, each set of harvesting robotic arms has two sets of drive units, which are respectively located at the upper joint and the lower joint of the two-degree-of-freedom rotary robotic arm, and are used to drive the swing of the upper swing arm and the lower swing arm in the two-degree-of-freedom rotary robotic arm.

4. The control method for the autonomous vibration-damping drop-type fruit harvesting platform according to claim 1, characterized in that, The flexible harvesting net is made of square flexible coarse linen cloth and the edges of the flexible harvesting net are sealed; each of the four opposite corners of the flexible harvesting net is provided with a connecting part that connects to the end effector of the harvesting robotic arm.

5. The control method for the autonomous vibration-damping drop-type fruit harvesting platform according to claim 1, characterized in that, The mobile chassis is a tracked mobile chassis.

6. The control method for the autonomous vibration-damping drop-type fruit harvesting platform according to claim 1, characterized in that, In step S3, the step of identifying the location of the fruit impact force using the fruit impact force identification mathematical model is as follows: Step S301: Using the center of the flexible harvesting net as the origin of the coordinate system, divide it into the first quadrant region, the second quadrant region, the third quadrant region, and the fourth quadrant region; then number the four sets of harvesting robotic arms. Step S302: Compare the changes in current / voltage of the brushless DC motor in the harvesting robot arm collected by the current / voltage sensors in each group of drive units, determine the harvesting robot arm with the largest change in current / voltage, and determine the quadrant area where the fruit fell in the flexible harvesting net according to the number of the harvesting robot arm.

7. The control method for the autonomous vibration-damping drop-type fruit harvesting platform according to claim 6, characterized in that, In step S3, the process by which the mathematical model for identifying fruit impact force identifies the magnitude of the fruit impact force is as follows: Step S311: During the fruit impact flexible harvesting net process, record the current / voltage data of the brushless DC motors in the upper and lower joints of the four sets of harvesting robotic arms in real time. Step S312: In the recorded current / voltage data, find the moment when the current / voltage of the brushless DC motor in the upper and lower joints of each harvesting robot arm first reaches its extreme value, obtain the current / voltage value at that moment, and convert it into the corresponding motor torque according to the functional relationship between current / voltage and torque. Step S313: Obtain the rotation angle of the brushless DC motors in the upper and lower joints of the harvesting robot arm through a rotary encoder, thereby obtaining the posture of the harvesting robot arm. Combined with the motor torque obtained in step S312, calculate the force borne by the end effector of each group of harvesting robot arms. By vector summing the forces borne by the end effectors of each group of harvesting robot arms, obtain the magnitude of the fruit impact force.

8. The control method for the autonomous vibration-damping drop-type fruit harvesting platform according to claim 7, characterized in that, In step S4, the steps for formulating the vibration reduction and recovery strategy are as follows: Step S401: Record current or voltage data in real time that are the same as the joint torque changes of the harvesting robotic arm; Step S402: Generate corresponding waveforms based on the recorded current or voltage data to detect the frequency and amplitude of the current or voltage in real time, thereby obtaining the frequency and amplitude of the vibration experienced by the four sets of harvesting robotic arms. Step S403: The vibrations experienced by the four sets of harvesting robotic arms will affect the motor torque of the brushless DC motors in the upper and lower joints of each set of harvesting robotic arms. Taking the resulting motor torque as a known condition, the position, angular velocity, and angular acceleration of the upper and lower joints of each set of harvesting robotic arms are calculated in real time using inverse kinematics. Combined with the inverse dynamics equation of the second kind of Lagrange equation, the optimal counteracting torque that the upper and lower joints of each set of harvesting robotic arms need to generate to counteract the motor torque is calculated in real time. At the same time, a control command corresponding to the optimal counteracting torque is generated and sent to each drive unit of each set of harvesting robotic arms to cause the brushless DC motors at the upper and lower joints of each set of harvesting robotic arms to output the calculated optimal counteracting torque, thereby driving the flexible harvesting net to move and reduce or counteract the vibration energy of the fruit falling onto the flexible harvesting net.

Images