A flexible end effector for harvesting economic forest fruits using a force-position coordinated control method

Abstract

This invention belongs to the field of agricultural harvesting robot technology, specifically relating to a flexible end effector for harvesting economic forest fruits using a force-position collaborative control method, aiming to enable the end effector to achieve high-quality harvesting. It includes a base, a gripping structure, and a drive mechanism. The gripping structure includes an active finger, first and second passive fingers, and correspondingly connected bases. The gripping structure also includes a first gear adjustment plate, a second gear adjustment plate, and a support plate; the gripping structure is fixed to the base, the base of the first passive finger is connected to the first gear adjustment plate via a mortise and tenon structure, the base of the second passive finger is connected to the second gear adjustment plate via a mortise and tenon structure, the support plate is connected to the first and second gear adjustment plates, and also to the active finger base; the drive mechanism is located inside the base and includes a miniature servo cylinder and a push rod, the servo cylinder is connected to one end of the push rod, and the other end of the push rod is connected to the active finger.

Description

Technical Field

[0001] This invention relates to the field of agricultural harvesting robot technology, and in particular to a flexible end effector for harvesting economic forest fruits using a force-position collaborative control method. Background Technology

[0002] Against the backdrop of the development of smart agriculture automation, the large-scale harvesting of economic forest fruits (such as citrus, apples, tomatoes, etc.) has placed higher demands on the adaptability and control precision of end effectors.

[0003] Existing end effectors lack structural adaptability, often resulting in fruit damage during harvesting. Furthermore, current control strategies for end effectors struggle to balance position positioning speed with flexible contact force control, failing to meet the demands of high-precision harvesting. Summary of the Invention

[0004] The purpose of this invention is to provide a flexible end effector for harvesting economic forest fruits using a force-position coordinated control method, which aims to enable the end effector to achieve high-quality harvesting.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a flexible end effector for harvesting economic forest fruits using a force-position coordinated control method. The flexible end effector for harvesting economic forest fruits includes: a base, a gripping structure, and a driving mechanism; the gripping structure includes an active finger, a first passive finger, and a second passive finger, as well as correspondingly connected active finger bases, first passive finger bases, and second passive finger bases; the gripping structure also includes a first gear adjustment plate, a second gear adjustment plate, and a support plate; the gripping structure is fixed on the base, the first passive finger base is connected to the first gear adjustment plate via a mortise and tenon structure, the second passive finger base is connected to the second gear adjustment plate via a mortise and tenon structure, the support plate is connected to the first gear adjustment plate and the second gear adjustment plate, and is also connected to the active finger base; the driving mechanism is disposed inside the base and includes a miniature servo cylinder and a push rod, the servo cylinder is connected to one end of the push rod, and the other end of the push rod is connected to the active finger. The force-position coordinated control method includes: S1: performing active finger kinematic analysis based on a simplified mechanical structure, deriving forward and inverse kinematic equations and establishing a force calculation model; S2: constructing a hybrid control architecture that combines active disturbance rejection control and model predictive control, introducing a custom dynamic feedforward module and a custom linear dynamic compensator; S3: building an experimental platform and verifying the effectiveness of the actuator and control method through a multi-index evaluation system.

[0007] The first and second passive finger bases of the flexible economic forest fruit harvesting end effector provided in this application embodiment are connected to the gear adjustment plate through a mortise and tenon structure, enabling rapid switching between multiple gears without additional tools. This greatly enhances flexibility and efficiency, and solves the problem of poor adaptability of existing fixed claw structures. The active finger and the two passive fingers form a triangular gripping structure, which, together with the rigid connection of the support plate, allows the opening angle of the three claws to be adjusted through the gear adjustment plate during gripping, ensuring that the contact force is evenly distributed on the fruit surface, thus helping to prevent damage to the fruit during harvesting.

[0008] In some embodiments, the flexible economic forest fruit harvesting end effector further includes a thin-film pressure sensor located on the side of the active finger. The thin-film pressure sensor is configured to connect to the controller, collect grasping force data of the grasping structure, and output the data to the controller.

[0009] In some embodiments, the first gear adjustment plate and the second gear adjustment plate include multiple adjustment gears, and the distance between the different adjustment gears and the support plate is different; the first passive finger base and the second passive finger base are configured to be fixed on any one adjustment gear by means of a tenon and mortise structure.

[0010] In some embodiments, the active finger, the first passive finger, and the second passive finger are made of a soft material with a fin effect; the base material is made of carbon fiber 3D printed material.

[0011] In some embodiments, active finger kinematics analysis is performed based on a simplified mechanical structure, deriving forward and inverse kinematic equations and establishing a force calculation model, including: kinematic analysis constructs an equation system through a closed vector loop and the law of cosines to determine the kinematic relationship between the push rod and the gripping position.

[0012] In some embodiments, in the hybrid control architecture, a custom dynamics feedforward module provides input for model predictive control based on the relationship between force and displacement, and a custom linear dynamic compensator corrects prediction deviations through velocity errors.

[0013] In some embodiments, the hybrid control architecture uses an extended state observer to estimate and compensate for system disturbances in real time.

[0014] In some embodiments, the evaluation metrics of the experimental platform include steady-state arrival time, force tracking response time, and steady-state force tracking mean square error.

[0015] In some embodiments, the multi-index evaluation system performs dynamic performance quantification analysis based on the deviation between force sensor data and control commands. Attached Figure Description

[0016] Figure 1This is a schematic diagram of the structure of a flexible end effector for harvesting economic forest fruits from a specific angle, provided in an embodiment of this application.

[0017] Figure 2 This is a schematic diagram of a flexible economic forest fruit harvesting end effector from another angle, as provided in the embodiments of this application;

[0018] Figure 3 This is a flowchart of a force-position coordinated control method for a flexible economic forest fruit harvesting end effector provided in an embodiment of this application;

[0019] Figure 4 This is a schematic diagram of experimental data for a hybrid architecture under different gripping forces, provided in an embodiment of this application.

[0020] Figure 5 This is a graph of ablation experiment data provided in an embodiment of this application;

[0021] Figure 6 This is another ablation experiment data result diagram provided in the embodiment of this application. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of the invention, it should be understood that the terms "upper," "lower," "left," "right," "front," "rear," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or relative positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Unless otherwise specified, the above-mentioned orientational descriptions can be flexibly set in practical applications, provided that the relative positional relationships shown in the accompanying drawings are satisfied.

[0024] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0025] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "communication" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection. They can refer to a direct connection or an indirect connection through an intermediate medium, or a communication between the internal components of two elements. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0026] In embodiments of the invention, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, article, or apparatus that includes that element.

[0027] In embodiments of the present invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0028] Against the backdrop of the development of smart agriculture automation, the large-scale harvesting of economic forest fruits (such as citrus, apples, tomatoes, etc.) has placed higher demands on the adaptability and control precision of end effectors.

[0029] The end effector is the part of the robot that directly contacts the workpiece and performs the task. In agricultural harvesting robots, the end effector is the core component for completing the harvesting task, responsible for grasping, cutting, and conveying the fruit. It needs to directly deal with complex natural environments and fruits of different shapes and sizes; therefore, its design directly determines the harvesting efficiency, accuracy, and the degree of protection for the fruit.

[0030] Existing end effectors suffer from structural incompatibility, resulting in inadequate fruit protection during operation and frequent fruit damage. Furthermore, current control methods struggle to simultaneously achieve rapid and precise positioning of the end effector and flexible control of contact force, failing to meet the stringent requirements of high-precision harvesting operations.

[0031] In view of this, embodiments of this application provide a flexible end effector for harvesting economic forest fruits, exemplarily, such as... Figure 1 and Figure 2As shown. The end effector 100 includes: a base 1, a gripping structure 2, and a drive mechanism 3. The gripping structure 2 includes an active finger 21, a first passive finger 22, and a second passive finger 23, as well as correspondingly connected active finger bases 24, first passive finger bases 25, and second passive finger bases 26.

[0032] The gripping structure 2 also includes a first gear adjustment plate 27, a second gear adjustment plate 28, and a support plate 29. The gripping structure 2 is fixed on the base 1. The first passive finger base 25 is connected to the first gear adjustment plate 27 via a mortise and tenon structure, and the second passive finger base 25 is connected to the second gear adjustment plate 28 via a mortise and tenon structure. The support plate 29 is connected to the first gear adjustment plate 27, the second gear adjustment plate 28, and also to the active finger base 24. The drive mechanism 3 is located inside the base 1 and includes a miniature servo cylinder 31 and a push rod 32. One end of the servo cylinder 31 is connected to the push rod 32, and the other end of the push rod 32 is connected to the active finger 21.

[0033] The active finger 21 is driven by a micro servo cylinder 31 via a drive mechanism 3, which drives a push rod 32 to control its opening and closing. It works in conjunction with the first passive finger 22 and the second passive finger 23 to pick and discard the fruit. The first passive finger 22 and the second passive finger 23 are connected to the gear adjustment plate by a tenon and mortise structure, which allows for quick plug-and-play gear switching along the slots of the gear adjustment plate.

[0034] In some embodiments, the first gear adjustment plate 27 and the second gear adjustment plate 28 include multiple adjustment gears, and the distance between the different adjustment gears and the support plate 29 is different; the first passive finger base 25 and the second passive finger base 26 are configured to be fixed on any one adjustment gear by means of a tenon and mortise structure.

[0035] For example, the first gear adjustment plate 27 and the second gear adjustment plate 28 include four adjustment levels, with the distance between them and the active finger 21 decreasing in each level, which can flexibly adapt to fruits of different sizes.

[0036] The different distances between the adjustment positions and the support plate 29 mean that the distances between the first passive finger 22 and the second passive finger 23 and the active finger 21 are also different, and can be adjusted in a timely manner according to actual conditions, such as the size of the fruit. The first adjustment plate 27 and the second adjustment plate 28 are fixed to the first passive finger base 25 and the second passive finger base 26 by fixing parts 20, forming a "convex-concave" tenon and mortise groove structure. The tenons of the first passive finger 22 and the second passive finger 23 are inserted into the mortises of the adjustment plates to achieve switching between different positions.

[0037] As one possible implementation, the flexible economic forest fruit harvesting end effector also includes a thin-film pressure sensor located on the side of the active finger. The thin-film pressure sensor is configured to connect to the controller, collect the grasping force data of the gripping structure, and output it to the controller.

[0038] For example, the miniature servo electric cylinder 31 (hereinafter referred to as the electric cylinder) is model LASF16-024D, with a stroke of 16mm, a peak thrust of 105N, a repeatability of ±0.03mm, and supports 50Hz control signals. The diaphragm pressure sensor has a range of 20g - 6kg.

[0039] A thin-film pressure sensor is positioned close to the gripping point to collect real-time contact force data between the gripping structure and the fruit, and feeds it back to the controller, forming a closed-loop system of "force sensing-control adjustment". This design allows the controller to dynamically adjust the output of the drive mechanism according to the actual gripping force, avoiding fruit damage or unstable gripping caused by improper gripping force in traditional open-loop control, and achieving synergistic optimization of flexible contact and stable gripping.

[0040] In some embodiments, the active finger, the first passive finger, and the second passive finger are made of a soft material with a fin effect, providing stable gripping force for the fruit within a diameter range of 10-100 mm. The base material includes carbon fiber 3D printing materials, such as Raise3D Industrial PA12 CF+, a nylon 6,12 (PA6,12; Nylon6,12) based carbon fiber reinforced composite material. The soft material forms a biomimetic contact surface through the fin effect, which can adaptively conform to the curvature of the fruit surface during gripping, uniformly distributing the contact force to the fruit skin, avoiding concentrated stress damage of traditional rigid materials, ensuring gripping stability, and buffering contact impact through material deformation. The carbon fiber 3D printing material combines the high strength of carbon fiber with the complex structural forming capabilities of 3D printing, enabling the base to maintain rigid support while significantly reducing weight.

[0041] This application also provides a force-position coordinated control method for a flexible economic forest fruit harvesting end effector, exemplarily referring to... Figure 3 The method includes:

[0042] S1: Based on a simplified mechanical structure, perform active finger kinematics analysis, derive forward and inverse kinematic equations, and establish a force calculation model.

[0043] In some embodiments, active finger kinematics analysis is performed based on a simplified mechanical structure, deriving forward and inverse kinematic equations and establishing a force calculation model, including: kinematic analysis constructs an equation system through a closed vector loop and the law of cosines to determine the kinematic relationship between the push rod and the gripping position.

[0044] For example, based on a simplified mechanical structure, forward and inverse kinematic equations are established using the properties of closed vector loops and the law of cosines to clarify the relationship between the grasping position, velocity, and acceleration of the push rod and the end effector. The forces acting on the active finger are analyzed, and force calculation formulas are derived based on the principle of torque balance. A geometric model of the end effector is constructed in Matlab to simulate the motion of the push rod and the fingertip. Under specific initial conditions, the position, velocity, and acceleration information of relevant points are recorded and compared with the kinematic equation calculation results to verify the accuracy of the forward and inverse kinematic calculations.

[0045] S2: Construct a hybrid control architecture that combines active disturbance rejection control and model predictive control, and introduce a custom dynamic feedforward module and a custom linear dynamic compensator.

[0046] A staged force-position coordinated control strategy is adopted. Based on the geometric dimensions of the target fruit, the optimal gripping stroke of the end effector's active finger is calculated using an inverse kinematics model to deduce the expected motion trajectory and required dynamic load of the electric cylinder push rod. A hybrid control architecture based on the coordination of Active Disturbance Rejection Control (ADRC) and Model Predictive Control (MPC) is constructed, introducing a custom dynamic feedforward module (CPV), an extended state observer (ESO), and a custom linear dynamic compensator (LSEF). The specific mathematical models and parameter settings of the ESO, CPV, LSEF, and MPC modules are determined to achieve precise coordinated control of staged force and position.

[0047] In the hybrid control architecture, a custom dynamics feedforward module provides input to model predictive control based on the force-displacement relationship, while a custom linear dynamic compensator corrects prediction deviations through velocity errors. The hybrid control architecture uses an extended state observer to perform real-time estimation and compensation for system disturbances.

[0048] Specifically, the ESO, as the core unit for disturbance observation and dynamic decoupling, expands the unmodeled dynamics of the system (frictional nonlinearity, inertial coupling) into additional state variables, thereby achieving real-time estimation and compensation of the total disturbance. For the end-effector drive system, a third-order ESO is constructed, with the following dynamic equations:

[0049]

[0050] in, These represent the actual position and actual velocity of the observation system, respectively. This is a state extension quantity, representing the total disturbance of the system, including friction, external disturbances, and model errors. For MPC output commands, These represent the nonlinear observation gains.

[0051]

[0052] In traditional Active Disturbance Rejection Control (ADRC) architectures, the tracking differentiator module is typically used to generate a smooth reference trajectory and extract the differential signal. However, in precision transmission scenarios, its dynamic performance is easily degraded due to phase delay and noise sensitivity. Therefore, this application employs a custom dynamic feedforward module (CPV) that calculates the displacement x and velocity v under the action of force F using corresponding formulas, and uses these as inputs to Model Predictive Control (MPC). The introduction of this module effectively improves the optimization efficiency and dynamic response capability of MPC while retaining the anti-disturbance advantages of ADRC.

[0053]

[0054] in, , Let be the initial position and initial velocity, t be the time, and a be the acceleration.

[0055] As one possible implementation, a custom linear dynamic compensator acts as a coordinating hub between MPC and ESO, undertaking the dual functions of dynamic error compensation and disturbance feedforward suppression. Its mathematical model can be described as follows:

[0056]

[0057] in, The reference speed for MPC prediction, denoted as ESO, k is the velocity observed, F is the actual force measured by the sensor, and m is the equivalent mass of the electric cylinder.

[0058] The MPC module, as the core optimization unit of the ADRC-MPC hybrid architecture, achieves coordinated optimization of high-precision trajectory tracking and strong anti-disturbance capability through rolling time-domain optimization and a perturbation feedforward mechanism. This module uses the theoretically unperturbed position x and velocity v generated by the CPV module as dual reference inputs. The MPC constructs a prediction model based on Newton-Euler dynamics, and its state equation is expressed as:

[0059]

[0060] in, and These represent the position and velocity state at step k. To control the cycle, m is the equivalent mass of the electric cylinder. The control variable to be optimized is used. A quadratic cost function is then constructed within the prediction time domain N:

[0061]

[0062] Where Q and R are the weights for state tracking and input rate of change, respectively. The total disturbance estimate of the ESO output is obtained by... Achieve disturbance feedforward compensation.

[0063] S3: Build an experimental platform and verify the effectiveness of the actuator and control method through a multi-index evaluation system.

[0064] In some embodiments, the evaluation metrics of the experimental platform include steady-state arrival time, force tracking response time, and steady-state force tracking mean square error.

[0065] The steady-state arrival time is defined as the time interval from the start of force control to the point where the absolute difference between the control force and the steady-state force is less than 0.1N for three consecutive steps. The force tracking response time is determined by the time interval from the start of force control to the point where the absolute difference between the force sensor reading and the control command force is less than 0.1N for three consecutive steps. The steady-state force tracking mean square error is based on data within a 10-20 second time window, and the formula is as follows:

[0066]

[0067] In some embodiments, the multi-index evaluation system performs dynamic performance quantification analysis based on the deviation between force sensor data and control commands. By analyzing the deviation between the measured data collected in real time by the force sensor and the control commands, the grasping performance of the end effector is transformed into quantifiable dynamic indicators (such as steady-state arrival time, force tracking response time, etc.). This data-driven approach intuitively presents the response characteristics of the control strategy under different operating conditions. For example, the deviation curve can clearly identify the entire process of the system transitioning from dynamic to steady state after force control is initiated, providing precise targets for performance optimization.

[0068] For example, an experimental platform was built, and a stepped increasing thrust of 1N to 30N was applied under no-load conditions of the electric cylinder. Push rod position data was collected, and the equivalent mass *m* of the electric cylinder was calculated using quadratic polynomial fitting, differentiation, and Newton's second law. A standard apple with a diameter of 55.59mm was used as the test object, and a target clamping force was set. The end effector employed a two-stage control strategy for grasping experiments. Steady-state arrival time, force tracking response time, and steady-state force tracking mean square error (MSE) were quantitatively analyzed to evaluate the effectiveness of the control strategy.

[0069] Reference Figure 4 , Figure 4 For experimental data of hybrid architecture under different gripping forces, based on Figure 4 It can be seen that the average steady-state arrival time of the hybrid architecture is 3.08s, the average force tracking response time is 0.85s, and the average steady-state force tracking mean square error is 0.00397N. 2These quantitative indicators fully verify the three core characteristics of the control system: First, the model feedforward compensation mechanism of the CPV module endows the system with fast response capability, with response time controlled within 1 second in 90% of operating conditions; Second, the cooperative working mode of the ESO disturbance observer and the LSEF linear compensator ensures that the system achieves high-precision force tracking control, and the steady-state error can be stably maintained within ±0.1N; Third, the multi-step predictive optimization mechanism of MPC significantly improves force tracking efficiency and effectively shortens force tracking time.

[0070] This application also verifies the roles of MPC and LSEF in the hybrid control architecture through ablation experiments, exemplarily referring to... Figure 5 and Figure 6 The experiment compared the performance of the hybrid control architecture with CPV, ESO and LSEF modules intact, the hybrid control architecture with MPC removed and the hybrid control architecture with LSEF removed.

[0071] Reference Figure 5 Compared with the complete hybrid control architecture, the removal of the MPC module and the results of ten sets of experimental data under different gripping forces show that the average steady-state setup time of the system is extended by 60.06%, and the force tracking response time deteriorates by 323.5%. This demonstrates the important role of the multi-step predictive optimization mechanism of the MPC module in improving the dynamic performance of the system, especially its key role in accelerating the convergence process and improving the force tracking response time.

[0072] It should be understood that the steady-state force tracking mean square error did not change significantly compared to the complete architecture. This phenomenon reveals that the linear compensation mechanism of the LSEF module can still effectively maintain the steady-state control accuracy of the system in the absence of MPC predictive optimization, thus verifying the effectiveness of the LSEF module design.

[0073] Reference Figure 6 Compared with the complete hybrid control architecture, removing the LSEF module and replacing it with a traditional nonlinear state error feedback (NLSEF) controller, according to ten sets of experimental data under different gripping forces, shows that the system steady-state settling time is shortened by 0.088%, the force tracking response time is extended by 0.024%, and the average error is also improved to some extent.

[0074] This result demonstrates both the potential advantages of nonlinear control in specific scenarios and exposes its inherent limitations. While the NLSEF module achieves more precise steady-state control through a nonlinear feedback mechanism, the number of parameters it requires adjustment (three or more) is far greater than that of the LSEF module (only one). This leads to a series of problems in practical engineering applications: First, the parameter tuning process is cumbersome and time-consuming. Second, the parameter's adaptability to different operating conditions is poor. Third, the system stability is more sensitive to parameter fluctuations. In contrast, the LSEF module, with its simple linear architecture and single adjustable parameter, significantly improves the system's engineering applicability and parameter robustness while ensuring control performance.

[0075] In the description of this specification, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0076] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A flexible end effector for harvesting economic forest fruits using a force-position coordinated control method, characterized in that, The flexible economic forest fruit harvesting end effector includes: a base, a gripping structure, and a drive mechanism; The gripping structure includes an active finger, a first passive finger, a second passive finger, and corresponding active finger bases, first passive finger bases, and second passive finger bases connected to each other. The gripping structure further includes a first gear adjustment plate, a second gear adjustment plate, and a support plate; the gripping structure is fixed on the base, the first passive finger base is connected to the first gear adjustment plate through a tenon and mortise structure, the second passive finger base is connected to the second gear adjustment plate through a tenon and mortise structure, and the support plate is connected to the first gear adjustment plate, the second gear adjustment plate, and also to the active finger base; The drive mechanism is located inside the base and includes a miniature servo cylinder and a push rod. The servo cylinder is connected to one end of the push rod, and the other end of the push rod is connected to the active finger. The force-position coordinated control method includes: S1: Based on a simplified mechanical structure, perform active finger kinematics analysis, derive forward and inverse kinematic equations, and establish a force calculation model; S2: Construct a hybrid control architecture that combines active disturbance rejection control and model predictive control, and introduce a custom dynamic feedforward module and a custom linear dynamic compensator; S3: Build an experimental platform and verify the effectiveness of the actuator and control method through a multi-index evaluation system.

2. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 1, characterized in that, The flexible economic forest fruit harvesting end effector also includes a thin-film pressure sensor, which is located on the side of the active finger. The thin-film pressure sensor is configured to connect to the controller, collect the grasping force data of the gripping structure, and output it to the controller.

3. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 1, characterized in that, The first and second gear adjustment plates include multiple adjustment levels, and the distance between the different adjustment levels and the support plate is different; the first and second passive finger bases are configured to be fixed to any one adjustment level by means of a tenon and mortise structure.

4. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 1, characterized in that, The active finger, the first passive finger, and the second passive finger are made of soft materials with a fin effect; The base material includes carbon fiber 3D printing material.

5. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 1, characterized in that, The method for performing active finger kinematics analysis based on a simplified mechanical structure, deriving forward and inverse kinematic equations, and establishing a force calculation model includes: The kinematic analysis constructs an equation system using a closed vector loop and the law of cosines to determine the kinematic relationship between the push rod and the grab position.

6. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 1, characterized in that, In the hybrid control architecture, the custom dynamics feedforward module provides input for model predictive control based on the relationship between force and displacement, and the custom linear dynamic compensator corrects the prediction deviation through velocity error.

7. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 6, characterized in that, The hybrid control architecture uses an extended state observer to estimate and compensate for system disturbances in real time.

8. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 1, characterized in that, The evaluation metrics for the experimental platform include steady-state arrival time, force tracking response time, and steady-state force tracking mean square error.

9. The flexible economic forest fruit harvesting end effector using a force-position coordinated control method according to claim 1, characterized in that, The multi-index evaluation system performs dynamic performance quantification analysis based on the deviation between force sensor data and control commands.

Images