A detection method and detection device for the grasping stability of a vibrating manipulator of an oil tea tree

Abstract

The present invention provides a method and a detection device for detecting the grasping stability of a vibrating mechanical hand for oil tea trees, which relates to the technical field of detecting the stability of mechanical hands, and includes: obtaining multiple sets of working parameters, effective friction forces and the proportion of branch damage during a complete grasping operation of the mechanical hand, establishing a neural network model based on these data, using the working parameters as inputs, and using the effective friction forces and the proportion of branch damage as labels for training to obtain a grasping stability detection model; inputting the working parameters of the current grasping operation into the model, outputting the predicted values of the effective friction force and the proportion of branch damage, and determining the stability reference value; determining a grasping stability correction factor by using the fruit mass and environmental parameters, and correcting the stability reference value to obtain a comprehensive stability index. Comparing the comprehensive stability index with a preset stability threshold to judge the stability level of the mechanical hand, and the present invention improves the efficiency and accuracy of detecting the grasping stability of the mechanical hand.

Description

Technical Field

[0001] The present invention relates to the technical field of manipulator stability detection, and particularly to a detection method and device for the grasping stability of a vibrating manipulator for oil tea trees. Background Art

[0002] With the advancement of agricultural modernization, the application of vibrating manipulators in fruit tree picking has gradually attracted attention, especially in the efficient picking process of oil tea tree fruits. The vibrating manipulator makes the branches resonate by grasping and vibrating the branches, thereby facilitating the shedding of fruits. This technology shows obvious advantages in improving picking efficiency and reducing labor costs. However, current vibrating manipulators still face many challenges in actual applications, especially in terms of grasping stability. The grasping force and vibration frequency during the vibration process are crucial for the picking effect of fruits, but due to the lack of real-time monitoring and feedback mechanisms, the manipulator often fails to grasp firmly during operation, resulting in fruit shedding or damage.

[0003] In addition, the existing technology for evaluating grasping stability mainly relies on manual experience and static tests, lacking adaptability to dynamic working environments and real-time analysis capabilities. Environmental factors, such as temperature, humidity, wind speed, and fruit quality, have complex effects on the grasping process, and the existing technology often fails to comprehensively consider these factors, resulting in poor operation effects of the manipulator under different environmental conditions. The lack of an effective real-time monitoring and adjustment mechanism not only increases the risk of fruit damage but also reduces the overall operation efficiency of the vibrating manipulator. Therefore, there is an urgent need for a new detection method that can real-time monitor and evaluate the stability of various parameters of the manipulator during the grasping process to cope with complex and changeable operating environments and achieve more efficient fruit picking.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The purpose of the present invention is to provide a detection method and device for the grasping stability of a vibrating manipulator for oil tea trees to solve the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A detection method for the grasping stability of a vibrating manipulator for oil tea trees, the specific steps including:

[0008] Step 1: Obtain the working parameters, effective friction force, and the proportion of branch damage caused by the vibrating manipulator during a complete grasping operation. Map the working parameters, effective friction force, and the proportion of branch damage for each grasping operation one by one to generate a sample data set. The working parameters include the grasping force, end acceleration, vibration frequency, and clamping width;

[0009] Step 2: Based on the data in the sample data set, establish a neural network model. Use the working parameters in the sample data set as the input of the model, and use the effective friction force and the proportion of branch damage in the sample data set as labels to train the neural network model to obtain a grasping stability detection model;

[0010] Step 3: Obtain the working parameters of the vibrating manipulator during the current grasping operation, input them into the trained grasping stability detection model. The model outputs the predicted value of the effective friction force and the predicted value of the proportion of branch damage. Determine the stability reference value of this grasping operation based on the predicted value of the effective friction force and the predicted value of the proportion of branch damage;

[0011] Step 4: Obtain the fruit mass and environmental parameters of the oil tea tree to be picked. Determine the grasping stability correction factor based on the fruit mass and environmental parameters, and use the grasping stability correction factor to correct the stability reference value to obtain a comprehensive stability index. The fruit mass is the average mass of the fruit of the oil tea tree to be picked, and the environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located;

[0012] Step 5: Compare the corrected comprehensive stability index with the preset stability threshold. According to the comparison result, judge the stability level of the vibrating manipulator to complete the detection of the grasping stability of the oil tea tree vibrating manipulator.

[0013] Furthermore, the generation method of the sample data set is: Map the working parameters of each grasping operation with the effective friction force and the proportion of branch damage caused one by one to form a data chain, and record the formed data chain as the sample data set;

[0014] Based on the data in the sample data set, establish a neural network model. Among them, select the long short-term memory network LSTM model as the basic model to establish, select the activation function and optimization algorithm. Among them, select the Tanh function as the activation function and select Adam as the optimization algorithm of the LSTM model;

[0015] The expression of the Tanh function is as follows:

[0016] ;

[0017] In the formula, represents the Tanh function, and the independent variable Represents the input weighted sum of neurons, that is, the result after weighted summation of the inputs received by the neurons from the previous layer; at the same time, set the hyperparameters of the LSTM model, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer; among them, the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32;

[0018] Randomly divide the sample data set into a training set and a test set. The working parameters include the grasping force, the end acceleration, the vibration frequency, and the clamping width; construct a grasping stability detection model based on the deep learning algorithm, use the working parameters in the training set as inputs, and use the effective friction force and the corresponding proportion of branch damage as labels to train the model to obtain a trained grasping stability detection model. Substitute the working parameters in the test set into the trained model to obtain the corresponding prediction results; calculate the error between the prediction results and the actual values in the test set; determine whether the error meets the preset error threshold; if it meets, output the trained model, that is, the grasping stability detection model; if it does not meet, return to continue training; the error is the mean absolute error, the root mean square error, and the coefficient of determination between the prediction results and the actual values in the test set;

[0019] The process of randomly dividing into a training set and a test set, the specific logic is: perform a random sorting process on the sample data set, and use 80% of the sorted sample data set as the training set, and the remaining 20% as the test set;

[0020] The input of the trained grasping stability detection model is the working parameter, and the output is the effective friction force and the proportion of branch damage.

[0021] Furthermore, the working parameters include the grasping force, the end acceleration, the vibration frequency, and the clamping width, where:

[0022] The grasping force is the normal pressure generated at the contact surface between the manipulator gripper and the branch, and the average value of the grasping force in the complete picking cycle is used as the grasping force for this grasping operation;

[0023] The end acceleration is the vibration acceleration of the end effector of the manipulator. Collect the acceleration time series data of the manipulator during the vibration picking stage, and use the average value of the acceleration in the stable vibration period as the end acceleration of the vibrating manipulator; among them, the stable vibration period refers to the vibration stage where the fluctuation range of the vibration excitation frequency applied by the manipulator does not exceed And the duration ≥ 3 seconds;

[0024] The vibration frequency is the vibration excitation frequency actively applied by the manipulator. The average value of the vibration frequency during the stable vibration period during the grasping operation is taken as the vibration frequency of this grasping operation;

[0025] The clamping width is the net contact distance between the inner sides of the two jaws, that is, the distance between the inner sides of the two jaws during the stable vibration period after applying the rated grasping force;

[0026] The effective friction force refers to the friction force component that actually plays an anti-slip role between the manipulator jaw and the tree branch contact surface. During the complete grasping operation cycle, the average value of the effective friction force during the stable vibration period is taken as the output value of the grasping stability detection model;

[0027] The specific logic for calculating the effective friction force is as follows:

[0028] Measure the normal pressure component of the manipulator jaw on the tree branch through a six-dimensional force sensor , and deduct the influence of the manipulator's own weight. The formula is as follows:

[0029] ;

[0030] In the formula, is the original output force vector of the sensor, represents the modulus of the force vector, is the contact surface inclination angle between the jaw and the tree branch, is the jaw mass, is the acceleration due to gravity;

[0031] Determine the effective friction force according to the following formula:

[0032] ;

[0033] In the formula, is the effective friction force, is the friction coefficient, calibrated through experiments, is the normal pressure component of the manipulator jaw on the tree branch;

[0034] The proportion of tree branch damage refers to the ratio of the area of skin breakage or structural damage of the tree branch caused by the manipulator grasping to the total contact area. Its calculation formula is as follows:

[0035] ;

[0036] In the formula, is the proportion of tree branch damage, is the area of skin breakage or structural damage of the tree branch caused by the manipulator grasping, is the actual contact area between the jaw and the tree branch.

[0037] Further, obtain the working parameters of the vibrating manipulator during the current grasping operation, and input them into the trained grasping stability detection model. The model outputs the predicted value of the effective friction force and the predicted value of the proportion of branch damage , where the working parameters include the grasping force, the end acceleration, the vibration frequency, and the clamping width;

[0038] Determine the stability reference value of this grasping operation based on the predicted value of the effective friction force and the predicted value of the proportion of branch damage. The formula is as follows:

[0039] ;

[0040] In the formula, is the stability reference value, is the predicted value of the effective friction force, is the predicted value of the proportion of branch damage, and are preset weights, , and satisfy , is the optimal effective friction force value under the current working condition, is the effective friction force allowable deviation range coefficient, The value range of is

[0041] Among them, the function for determining is as follows:

[0042] ;

[0043] In the formula, is the friction coefficient, calibrated through experiments, is the mass of the jaw, is the acceleration due to gravity.

[0044] Further, obtain the fruit mass and environmental parameters of the oil tea tree to be picked, and determine the grasping stability correction factor based on the fruit mass and environmental parameters. The method for determining the fruit mass is as follows: Randomly collect multiple fruits and calculate their mass average, and take the mass average as the average mass of the fruits of the oil tea tree to be picked. The environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located; Collect the temperature, wind speed, and humidity data multiple times at the same moment, and take the average value as the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located;

[0045] Among them, the formula for determining the grasping stability correction factor is as follows:

[0046] ;

[0047] In the formula, is the grasping stability correction factor, is the fruit mass, is the rated load weight of the manipulator, determined according to the mechanical design parameters, is the temperature, is the reference temperature, is the humidity, is the reference humidity, is the wind speed, is the reference wind speed; is the mass influence index, calibrated through load experiments, reflecting the non-linear influence of mass change on stability; is the wind speed attenuation coefficient, calibrated through wind tunnel experiments, reflecting the suppression intensity of wind speed on stability.

[0048] Furthermore, the grasping stability correction factor is used to correct the stability reference value to obtain the comprehensive stability index. The formula is as follows:

[0049] ;

[0050] In the formula, is the comprehensive stability index, is the stability reference value, is the grasping stability correction factor, is the preset weight, and .

[0051] Furthermore, the corrected comprehensive stability index is compared with the preset stability threshold. According to the comparison result, the stability level of the vibrating manipulator is judged;

[0052] Among them, the logic for comparing the corrected comprehensive stability index with the preset stability threshold is as follows:

[0053] When , it means that the comprehensive stability index is less than the threshold at this time, indicating that the current stability of the vibrating manipulator is poor, and the stability level of the manipulator is judged to be: unstable;

[0054] When , it means that the comprehensive stability index is equal to the threshold at this time, indicating that the current stability of the vibrating manipulator is in a boundary state, and the stability level of the manipulator is judged to be: acceptable;

[0055] When , it means that the comprehensive stability index is greater than the threshold at this time, indicating that the current stability of the vibrating manipulator is good, and the stability level of the manipulator is judged to be: stable;

[0056] In the formula, is the comprehensive stability index, is the stability threshold.

[0057] Further, the method for determining the stability threshold is as follows: collect the historical stability data of the manipulator under different working conditions, calculate the mean and standard deviation of the comprehensive stability index through statistical analysis, and determine the stability threshold according to the following formula:

[0058] ;

[0059] In the formula, is the stability threshold, is the mean of the comprehensive stability index, is the standard deviation of the comprehensive stability index, is the safety factor, and its usual value range is from 1 to 2 to ensure a certain safety margin in practical applications;

[0060] Conduct experiments under the set operating conditions to observe the stability performance of the manipulator, verify the effectiveness of the set threshold through experimental data, ensure that the threshold can accurately distinguish the three stability levels of unstable, acceptable, and stable, evaluate and adjust the set stability threshold, and continuously monitor its comprehensive stability index during the actual operation of the manipulator, and regularly adjust the stability threshold according to the real-time feedback and performance to ensure its adaptability and effectiveness.

[0061] The present invention also provides a grasping stability detection device for an oil tea tree vibrating manipulator. The grasping stability detection device for an oil tea tree vibrating manipulator is used to execute the above-mentioned grasping stability detection method for an oil tea tree vibrating manipulator, and includes:

[0062] A data acquisition module, which is used to obtain the working parameters, effective friction force, and the proportion of branch damage caused by the vibrating manipulator in a complete grasping operation, map the working parameters, effective friction force, and the proportion of branch damage in each grasping operation one by one to generate a sample data set. The working parameters include grasping force, end acceleration, vibration frequency, and clamping width;

[0063] A grasping stability detection model construction module, which based on the data in the sample data set, establishes a neural network model, uses the working parameters in the sample data set as the input of the model, and uses the effective friction force and the proportion of branch damage in the sample data set as labels to train the neural network model to obtain a grasping stability detection model;

[0064] A stability prediction module is used to obtain the working parameters of the vibrating manipulator during the current grasping operation, input them into the trained grasping stability detection model, and the model outputs the predicted value of the effective friction force and the predicted value of the proportion of branch damage. Based on the predicted value of the effective friction force and the predicted value of the proportion of branch damage, the stability reference value of this grasping operation is determined. The working parameters include the grasping force, the end acceleration, the vibration frequency, and the clamping width;

[0065] An environmental dynamic correction module is used to obtain the fruit quality and environmental parameters of the oil tea tree to be picked, determine the grasping stability correction factor based on the fruit quality and environmental parameters, and use the grasping stability correction factor to correct the stability reference value to obtain the comprehensive stability index. The fruit quality is the average quality of the fruits of the oil tea tree to be picked, and the environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located;

[0066] A stability determination module is used to compare the corrected comprehensive stability index with a preset stability threshold, and based on the comparison result, judge the stability level of the vibrating manipulator to complete the detection of the grasping stability of the oil tea tree vibrating manipulator.

[0067] Compared with the prior art, the beneficial effects of the present invention are:

[0068] The method for detecting the grasping stability of the oil tea tree vibrating manipulator provided by the present invention can monitor and evaluate the grasping stability of the manipulator in real time, and solve the problem of insufficient stability existing in the traditional picking method. By establishing a neural network model, it can accurately predict the effective friction force and the proportion of branch damage to ensure the grasping effect under different operating conditions. This method not only improves the success rate of fruit picking, reduces the damage caused by unstable grasping, but also provides a data basis for the dynamic adjustment of the manipulator, thereby improving the operation efficiency and stability of the manipulator. In addition, the correction factor combining the environmental parameters and the fruit quality makes the stability evaluation more comprehensive and accurate, further promoting the transformation of agricultural mechanization to intelligence and providing strong support for the efficient picking of oil tea trees. Description of the Drawings

[0069] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0070] Figure 2 It is a schematic diagram of the overall device module of the present invention;

[0071] Figure 3 It is a fitting curve of the predicted value of the effective friction force and the stability reference value;

[0072] Figure 4 It is a fitting curve of the predicted value of the proportion of branch damage and the stability reference value. Detailed Embodiments

[0073] To make the objectives, technical solutions, and advantages of the present invention more clear and understandable, the present invention will be further described in detail below in conjunction with specific embodiments.

[0074] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meaning understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second", and similar terms used in the present invention do not indicate any order, quantity, or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0075] Embodiment:

[0076] Please refer to Figure 1 , the present invention provides a technical solution:

[0077] A method for detecting the grasping stability of a vibrating manipulator for oil tea trees, the specific steps include:

[0078] Step 1: Obtain the working parameters, effective friction force, and the proportion of branch damage caused in a single complete grasping operation of the vibrating manipulator, map the working parameters, effective friction force, and the proportion of branch damage in each grasping operation one by one to generate a sample data set, and the working parameters include the grasping force, end acceleration, vibration frequency, and clamping width;

[0079] Step 2: Based on the data in the sample data set, establish a neural network model, use the working parameters in the sample data set as the input of the model, and use the effective friction force and the proportion of branch damage in the sample data set as labels to train the neural network model to obtain a grasping stability detection model;

[0080] In this embodiment, the method for generating the sample data set is: map the working parameters of each grasping operation to the effective friction force and the proportion of branch damage caused one by one to form a data chain, and record the formed data chain as the sample data set;

[0081] Based on the data in the sample data set, establish a neural network model. Among them, select the long short-term memory network LSTM model as the basic model to establish, select the activation function and optimization algorithm, and select the Tanh function as the activation function and Adam as the optimization algorithm for the LSTM model;

[0082] The expression of the Tanh function is as follows:

[0083] ;

[0084] In the formula, represents the Tanh function, and the independent variable represents the weighted sum of the inputs of the neuron, that is, the result after the inputs received by the neuron from the previous layer are weighted and summed; at the same time, the hyperparameters of the LSTM model are set. The hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer; among them, the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32;

[0085] The sample data set is randomly divided into a training set and a test set. The working parameters include the grasping force, the end acceleration, the vibration frequency, and the clamping width; a grasping stability detection model is constructed based on the deep learning algorithm. The working parameters in the training set are used as inputs, and the effective friction force and the corresponding proportion of branch damage are used as labels to train the model to obtain a trained grasping stability detection model. The working parameters in the test set are substituted into the trained model to obtain the corresponding prediction results; calculate the error between the prediction results and the actual values in the test set; determine whether the error meets the preset error threshold; if it meets, output the trained model, that is, the grasping stability detection model; if it does not meet, return to continue training; the error is the mean absolute error, the root mean square error, and the coefficient of determination between the prediction results and the actual values in the test set;

[0086] The process of randomly dividing into a training set and a test set, the specific logic is: the sample data set is randomly sorted, and 80% of the sorted sample data set is used as the training set, and the remaining 20% is used as the test set;

[0087] The input of the trained grasping stability detection model is the working parameter, and the output is the effective friction force and the proportion of branch damage.

[0088] The advantages of Step 1 and Step 2 are that by systematically obtaining multiple sets of working parameters of the vibrating manipulator in the grasping operation and the corresponding effective friction force and the proportion of branch damage, a sample data set is formed, providing a solid data basis for the subsequent training of the model. This data-driven method significantly improves the accuracy and generalization ability of the model. Compared with the traditional experience-based evaluation method, it can more scientifically identify the key factors affecting grasping stability, thus effectively reducing the occurrence of false predictions and unstable operations.

[0089] Compared with the prior art, the beneficial effects of this solution are reflected in that the grasping stability detection model based on neural network has the ability of real-time prediction, enabling the vibrating manipulator to adjust the grasping strategy in a dynamic working environment in a timely manner. Through deep learning training on the sample data set, the model can not only adapt to different environmental conditions and fruit quality variations, but also provide accurate feedback for the operation of the manipulator, thereby improving the success rate and efficiency of fruit picking, reducing fruit damage, and optimizing the overall operation effect. In the overall solution, the implementation of steps 1 and 2 promotes the systematic collection and intelligent analysis of data, making the entire grasping stability detection process more scientific and intelligent. This not only provides an accurate basis for real-time monitoring and evaluation, but also lays a foundation for subsequent stability correction and grade judgment, helping to achieve efficient, precise, and safe operation of the vibrating manipulator, and further promoting the progress of the oil tea tree picking technology.

[0090] Step 3: Obtain the working parameters of the vibrating manipulator under the current grasping operation, input them into the trained grasping stability detection model, and the model outputs the predicted value of the effective friction force and the predicted value of the proportion of branch damage. Determine the stability reference value of this grasping operation based on the predicted value of the effective friction force and the predicted value of the proportion of branch damage;

[0091] In this embodiment, the working parameters include the grasping force, the end acceleration, the vibration frequency, and the clamping width, where:

[0092] The grasping force is the normal pressure generated at the contact surface between the manipulator jaw and the branch. Take the average value of the grasping force in the complete picking cycle as the grasping force of this grasping operation;

[0093] The end acceleration is the vibration acceleration of the end effector of the manipulator. Collect the acceleration time series data of the manipulator during the vibration picking stage, and take the average value of the acceleration in the stable vibration period as the end acceleration of the vibrating manipulator; where the stable vibration period refers to the vibration stage where the fluctuation range of the vibration excitation frequency applied by the manipulator does not exceed and the duration is ≥ 3 seconds;

[0094] The vibration frequency is the vibration excitation frequency actively applied by the manipulator. Take the average value of the vibration frequency in the stable vibration period during the grasping operation as the vibration frequency of this grasping operation;

[0095] The clamping width is the net contact distance between the inner sides of the two jaws, that is, the distance between the inner sides of the two jaws during the stable vibration period after applying the rated grasping force;

[0096] The effective friction force refers to the friction force component that actually plays an anti-slip role between the manipulator jaw and the branch contact surface. In the complete grasping operation cycle, take the average value of the effective friction force in the stable vibration period as the output value of the grasping stability detection model;

[0097] The specific logic for calculating the effective friction force is as follows:

[0098] Measure the normal pressure component of the manipulator gripper on the branch through a six-axis force sensor , and deduct the influence of the manipulator's own weight. The formula is as follows:

[0099] ;

[0100] In the formula, is the original output force vector of the sensor, represents the modulus of the force vector, is the contact surface inclination angle between the gripper and the branch, is the mass of the gripper, is the acceleration due to gravity;

[0101] Determine the effective friction force according to the following formula:

[0102] ;

[0103] In the formula, is the effective friction force, is the friction coefficient, calibrated through experiments, is the normal pressure component of the manipulator gripper on the branch;

[0104] The proportion of branch damage refers to the ratio of the area of skin breakage or structural damage of the branch caused by the manipulator's grasping to the total area of the contact area. Its calculation formula is as follows:

[0105] ;

[0106] In the formula, is the proportion of branch damage, is the area of skin breakage and structural damage of the branch caused by the manipulator's grasping, is the actual contact area between the gripper and the branch.

[0107] is used to effectively characterize the safety and stability of the manipulator's grasping. A higher value means that there may be a greater risk of instability in the manipulator's grasping operation, resulting in branch damage or a decrease in fruit quality. Therefore, from a technical effect perspective, reducing the proportion of branch damage is the key to improving the performance of the manipulator, which can ensure the quality of fruit picking, reduce economic losses, and improve the application reliability of the manipulator in agricultural operations.

[0108] Obtain the working parameters of the vibrating manipulator under the current grasping operation, and input them into the trained grasping stability detection model. The model outputs the predicted value of the effective friction force and the predicted value of the proportion of branch damage ; where the current grasping operation is a complete grasping operation;

[0109] Based on the predicted value of effective friction and the predicted value of the proportion of branch damage, determine the stability reference value of this grasping operation. The formula is as follows:

[0110] ;

[0111] In the formula, is the stability reference value, is the predicted value of effective friction, is the predicted value of the proportion of branch damage, and are preset weights, , ; is the optimal effective friction value under the current working condition, is the coefficient of the allowable deviation range of effective friction, The value of is defines the allowable deviation range of the effective friction relative to the optimal value . When , the contribution of the friction term drops to 0. The smaller the value of , the more sensitive is to the friction deviation; the larger the value of

[0112] In this formula, the dependent variable Specifically reflects the grasping stability degree of the manipulator under specific grasping operation conditions. Its meaning is to combine the predicted value of effective friction and the predicted value of the proportion of branch damage, and through the comprehensive analysis of these two indicators, evaluate whether the manipulator has sufficient stability during the grasping process to effectively prevent damage and ensure the success of grasping. The calculation of this value can provide a quantitative stability standard, which has an important technical effect on the operation safety of the manipulator, helping the operator to timely judge the risks during the grasping process and adjust the operation strategy accordingly, so as to improve the success rate of fruit picking and reduce the potential damage risk. Within a certain range, a higher effective friction means that the manipulator can more effectively resist sliding when grasping the branch, reducing the risk of sliding, thus improving the grasping stability. However, when the effective friction is too high, it may cause an increase in branch damage due to excessive grasping force, thus affecting the stability. A higher proportion of branch damage means that the risk of branch damage increases during the grasping process, and the grasping effect of the manipulator is unstable; that is to say, and show a negative correlation.

[0113] The formula uses the natural logarithm function to effectively transform the input non-linear relationship into a linear relationship, making the result more stable and controllable. The natural logarithm function can change smoothly when the value is small or close to zero, avoiding drastic fluctuations in the model output, which is very important in practical applications.

[0114] Among them, the function based on which is determined is as follows:

[0115] ;

[0116] In the formula, is the friction coefficient, calibrated through experiments, is the mass of the jaw, is the acceleration due to gravity.

[0117] Table 1: Statistical table of stability reference values

[0118] ;

[0119] Please refer to Figures 3 - 4, in this data analysis, relevant data on the predicted value of effective friction force, the predicted proportion of branch damage, and the optimal effective friction force under specific environmental conditions are presented, as well as the stability benchmark value calculated based on these data; the predicted value of effective friction force reflects the friction ability of the surface under specific conditions. Within an appropriate range, the higher the value, the better the surface can resist sliding and provide higher stability. The stability benchmark value reaches its maximum when the effective friction force is around 5.5 and then gradually decreases, which conforms to the optimal effect of effective friction force. This phenomenon indicates that there is an optimal friction force value for the grasping effect. If this value is exceeded, the increase in branch damage may be caused by excessive grasping force, thus affecting stability. The predicted proportion of branch damage refers to the expected proportion of damage that may be suffered during use. The higher the value, the more serious the damage, which may lead to a decrease in effective friction force and thus affect overall safety. The minimum anti-slip friction force is the lowest friction force threshold required to ensure safe operation. If the actual effective friction force is lower than this value, there is a risk of sliding. As can be seen from the table, when the value of effective friction force is less than 5.5, as the effective friction force increases, the stability benchmark value also increases accordingly, indicating that the increase in friction force helps to improve overall stability; at the same time, the decrease in the proportion of branch damage further promotes the increase in the stability benchmark value, emphasizing the importance of maintaining the structure to prevent damage. In summary, this table provides an important reference for decision-makers to evaluate the safety of materials under different environmental conditions and take corresponding measures to improve overall stability and safety.

[0120] The advantage of step 3 is that by inputting the working parameters of the current grasping operation of the vibrating manipulator into the trained stability detection model, it can predict the effective friction force and the proportion of branch damage in real time. This process realizes the dynamic evaluation of grasping stability, enabling the manipulator to adjust the grasping strategy in a timely manner according to the real-time data feedback to ensure stability under different operating conditions. This dynamic feedback mechanism significantly improves the operation safety and accuracy.

[0121] Compared with the prior art, the beneficial effect of step 3 in this solution is reflected in its ability to quickly and effectively evaluate the stability of the manipulator in actual operation. This model-based real-time detection method replaces the traditional static evaluation means, does not rely on manual experience judgment, reduces the risk of fruit damage caused by improper operation, and improves the automation and intelligence level of the entire picking process. In the overall solution, the adoption of step 3 promotes the intelligent operation of the manipulator. Through real-time monitoring and prediction, the manipulator can not only adapt to different environments and fruit states, but also optimize the grasping parameters according to the feedback data, thus achieving more efficient fruit picking. This process closely combines the grasping stability detection with the actual operation, providing strong technical support for the efficient, accurate, and safe operation of the manipulator, and promoting the continuous progress of the oil tea tree picking technology.

[0122] Step 4: Obtain the fruit quality and environmental parameters of the oil tea tree to be picked. Determine the grasping stability correction factor based on the fruit quality and environmental parameters, and use the grasping stability correction factor to correct the stability reference value to obtain the comprehensive stability index. The fruit quality is the average quality of the fruits of the oil tea tree to be picked, and the environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located;

[0123] In this embodiment, obtain the fruit quality and environmental parameters of the oil tea tree to be picked. Determine the grasping stability correction factor based on the fruit quality and environmental parameters. The fruit quality is the average quality of the fruits of the oil tea tree to be picked, and the environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located; Collect the temperature, wind speed, and humidity data multiple times at the same moment, and use the average value as the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located;

[0124] Among them, the formula for determining the grasping stability correction factor is as follows:

[0125] ;

[0126] In the formula, is the grasping stability correction factor, is the fruit quality, is the rated load weight of the manipulator, determined according to the mechanical design parameters, is the temperature, is the reference temperature, ; is the humidity, is the reference humidity, ; is the wind speed, is the reference wind speed, ; is the quality influence index, , calibrated through a load experiment, reflecting the non-linear influence of mass change on stability; is the wind speed attenuation coefficient, , calibrated through a wind tunnel experiment, reflecting the suppression intensity of wind speed on stability.

[0127] The grasping stability correction factor defined by this formula is used as the dependent variable, comprehensively quantifying the stability compensation requirements of the manipulator in a specific operating environment. The larger its value, the more favorable the environmental conditions are for stable grasping. The technical effect is to dynamically adjust the stability criterion to adapt to the actual working conditions. Among the independent variables: the fruit mass ratio reflects the load effect. As the fruit mass increases, it significantly reduces through the power function ​The value reflects the negative impact of increased load on stability; the deviations of temperature, humidity from the reference values affect through the Lorentz function. When the environmental parameters deviate from the reference values, the denominator term increases, resulting in attenuation, which conforms to the law of the characteristics of biological materials deteriorating with the environment; the wind speed acts on through the exponential attenuation term . The greater the wind speed, the more severe the attenuation and this effect is strengthened, reflecting the linear destructive effect of wind disturbance on grasping. All independent variables in the formula are negatively correlated with η: an increase in fruit mass, a deviation of environmental parameters from the reference values, or an increase in wind speed will all cause the value to decrease.

[0128] The design of this formula has clear physical significance and mathematical rationality: the mass term adopts the power function form, which can accurately reflect the non-linear effect of the load rate on stability. The exponent enhances the sensitivity of mass change; the temperature term and humidity term adopt the Lorentz function form, with a gentle change near the reference value and a rapid attenuation when deviating from the reference value, which conforms to the law of the change of biological material characteristics; the wind speed term describes the inhibitory effect of wind speed on stability through the exponential attenuation function. Its dimensionless treatment ensures the dimensional consistency of the formula. The attenuation coefficient is calibrated through wind tunnel experiments to accurately reflect the dynamic response of oil tea tree branches under wind speed interference. Each sub-item is coupled in a multiplicative form, which not only reflects the synergistic influence of multiple factors, but also ensures the final correction factor through normalization design. The larger the value, the more favorable the environmental conditions.

[0129] The grasping stability correction factor is used to correct the stability reference value to obtain the comprehensive stability index. The formula is as follows:

[0130] ;

[0131] In the formula, is the comprehensive stability index, is the stability reference value, is the grasping stability correction factor, is the preset weight, and .

[0132] This formula dynamically adjusts the stability reference value by introducing the grasping stability correction factor to obtain the comprehensive stability index . The formula adopts a linear weighted form, where The setting has clear engineering rationality: the correction factor reflects the environmental adaptation requirements. When approaches 1, it can be increased by up to . This not only avoids misjudgment caused by excessive correction but also reserves a reasonable margin for stability. When approaches 0, it degenerates to to ensure that the benchmark stability judgment is not disturbed. The value of the weight is calibrated through 200 sets of field experiments. Under the dual constraints of a fruit drop rate < 5% and a branch damage rate < 8%, it is proven that can optimally balance environmental adaptability and control stability. This linear form is computationally efficient and complements the non-linear benchmark value to jointly form a complete stability evaluation system.

[0133] In this formula, the dependent variable reflects the actual stability level of the manipulator under the current working conditions. The larger its value, the better the system stability. The technical effect is to adapt to different operating environments by dynamically correcting the benchmark value. The independent variables include the stability benchmark value and the correction factor . Among them, characterizes the inherent stability of the manipulator under standard working conditions, quantifies the influence degree of environmental factors: mass, temperature, humidity, and wind speed on stability; the weight is used to adjust the correction intensity, and this value is calibrated through experiments to ensure that it is neither overly corrected nor overly conservative. In the formula, is positively correlated with both and : when increases, the comprehensive stability index increases accordingly; conversely, when decreases, approaches the benchmark value . This design not only ensures environmental adaptability but also maintains the robustness of the system.

[0134] The advantage of Step 4 is that by obtaining the fruit quality and environmental parameters of the oil tea tree to be picked, it can provide a more comprehensive and accurate correction factor for the grasping stability. This process ensures that the stability detection is more in line with the actual operating conditions, can effectively consider the influence of environmental changes such as temperature, wind speed, and humidity on the grasping performance, and thus realizes dynamic and targeted adjustments. This correction mechanism effectively improves the adaptability of the manipulator in complex environments and reduces the risk of unstable grasping caused by environmental factors.

[0135] Compared with the prior art, the beneficial effect of step 4 in this solution is reflected in that by comprehensively considering the fruit quality and environmental parameters, it can more accurately evaluate the stability during the actual grasping process. Traditional methods often ignore the changes in environmental factors, resulting in inaccurate predictions and increasing the risk of fruit damage. The introduction of this step enables the model to flexibly respond to different operating conditions, improving the success rate and stability of grasping, and ensuring the efficiency and safety of the fruit picking process. In the overall solution, the adoption of step 4 promotes the intelligence and data-driven of the grasping stability detection method. Through the correction of environmental parameters in the real scenario, the entire detection process is not limited to static analysis but moves towards dynamic adjustment. This innovative adjustment mechanism not only improves the operation efficiency of the manipulator but also provides higher safety protection for fruit picking, laying a solid foundation for the further development of the vibration manipulator technology for oil tea trees.

[0136] Step 5: Compare the corrected comprehensive stability index with the preset stability threshold, and based on the comparison result, judge the stability level of the vibration manipulator to complete the detection of the grasping stability of the vibration manipulator for oil tea trees;

[0137] In this embodiment, the corrected comprehensive stability index is compared with the preset stability threshold, and based on the comparison result, the stability level of the vibration manipulator is judged;

[0138] Among them, the specific logic for comparing the corrected comprehensive stability index with the preset stability threshold is as follows:

[0139] When , it means that at this time the comprehensive stability index is less than the threshold, indicating that the current stability of the vibration manipulator is poor, and the stability level of the manipulator is determined as: unstable;

[0140] When , it means that at this time the comprehensive stability index is equal to the threshold, indicating that the current stability of the vibration manipulator is in a boundary state, and the stability level of the manipulator is determined as: acceptable;

[0141] When , it means that at this time the comprehensive stability index is greater than the threshold, indicating that the current stability of the vibration manipulator is good, and the stability level of the manipulator is determined as: stable;

[0142] In the formula, is the comprehensive stability index, is the stability threshold.

[0143] The method for determining the stability threshold is: collect the historical stability data of the manipulator under different working conditions, calculate the mean and standard deviation of the comprehensive stability index through statistical analysis, and determine the stability threshold according to the following formula:

[0144] ;

[0145] In the formula, is the stability threshold, is the mean value of the comprehensive stability index, is the standard deviation of the comprehensive stability index, is the safety factor, , to ensure a certain safety margin in practical applications;

[0146] Conduct experiments under the set operating conditions to observe the stability performance of the manipulator, verify the effectiveness of the set threshold through experimental data, ensure that the threshold can accurately distinguish the three stability levels of unstable, acceptable, and stable, evaluate and adjust the set stability threshold, and continuously monitor its comprehensive stability index during the actual operation of the manipulator, and regularly adjust the stability threshold according to real-time feedback and performance to ensure its adaptability and effectiveness.

[0147] The advantage of step 5 is that by comparing the corrected comprehensive stability index with the preset stability threshold, it can effectively judge the stability level of the vibrating manipulator. This process not only provides a clear stability evaluation criterion but also provides a basis for subsequent operation decisions, ensuring the efficiency and safety of the manipulator during operation. This dynamic evaluation mechanism can provide real-time feedback on the working state of the manipulator, enabling timely corresponding measures to be taken to reduce the risk of fruit damage.

[0148] Compared with the prior art, the beneficial effect of step 5 in this solution lies in its systematic and quantitative stability detection method, which improves the accuracy of the overall grasping stability evaluation. Traditional methods often rely on manual experience or simple qualitative analysis, lacking scientificity and accuracy. Through the quantitative analysis of the comprehensive stability index, the working state of the manipulator can be more accurately identified, avoiding losses caused by human judgment errors and improving the success rate of fruit picking. In the overall solution, the adoption of step 5 promotes the intelligent and automated level of grasping stability detection. Through real-time comparison and judgment, the manipulator can quickly adjust the operation strategy according to the stability level and optimize the picking process. This innovative mechanism not only improves the working efficiency of the manipulator but also ensures the quality of fruit picking, promotes the progress of the oil tea tree picking technology, and helps to achieve a more efficient agricultural production mode.

[0149] Please refer to Figure 2 , an oil tea tree vibrating manipulator grasping stability detection device, comprising:

[0150] The data acquisition module is used to obtain the working parameters, effective friction force, and the proportion of branch damage caused by a multi-group of vibrating manipulators during a complete grasping operation. It maps the working parameters, effective friction force, and the proportion of branch damage for each grasping operation one by one to generate a sample data set. The working parameters include the grasping force, end acceleration, vibration frequency, and clamping width.

[0151] The grasping stability detection model construction module, based on the data in the sample data set, establishes a neural network model. Using the working parameters in the sample data set as the input of the model and the effective friction force and the proportion of branch damage in the sample data set as the labels, it trains the neural network model to obtain the grasping stability detection model.

[0152] The stability prediction module is used to obtain the working parameters of the vibrating manipulator during the current grasping operation, input them into the trained grasping stability detection model. The model outputs the predicted value of the effective friction force and the predicted value of the proportion of branch damage. Based on the predicted value of the effective friction force and the predicted value of the proportion of branch damage, it determines the stability reference value for this grasping operation. The working parameters include the grasping force, end acceleration, vibration frequency, and clamping width.

[0153] The environmental dynamic correction module is used to obtain the fruit quality and environmental parameters of the oil tea tree to be picked. Based on the fruit quality and environmental parameters, it determines the grasping stability correction factor, and uses the grasping stability correction factor to correct the stability reference value to obtain the comprehensive stability index. The fruit quality is the average quality of the fruit of the oil tea tree to be picked, and the environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located.

[0154] The stability determination module is used to compare the corrected comprehensive stability index with a preset stability threshold. According to the comparison result, it determines the stability level of the vibrating manipulator, and completes the detection of the grasping stability of the oil tea tree vibrating manipulator.

[0155] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulation to get a formula that is closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0156] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by the combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0157] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, and it may be located in one place or distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0158] As mentioned above, the above are only specific implementation manners of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application.

Claims

1. A method for detecting the grasping stability of a vibrating manipulator for oil tea trees, characterized in that, The specific steps are as follows: Step 1: Obtain the working parameters, effective frictional force, and the proportion of branch damage caused during a complete grasping operation of the vibrating manipulator. Map the working parameters, effective frictional force, and the proportion of branch damage for each grasping operation one by one to generate a sample data set. The working parameters include the grasping force, end acceleration, vibration frequency, and clamping width; Step 2: Based on the data in the sample data set, establish a neural network model. Use the working parameters in the sample data set as the input of the model, and use the effective frictional force and the proportion of branch damage in the sample data set as labels to train the neural network model to obtain a grasping stability detection model; Step 3: Obtain the working parameters of the vibrating manipulator during the current grasping operation, and input them into the trained grasping stability detection model. The model outputs the predicted value of the effective frictional force and the predicted value of the proportion of branch damage. Determine the stability reference value of this grasping operation based on the predicted value of the effective frictional force and the predicted value of the proportion of branch damage; Step 4: Obtain the fruit quality and environmental parameters of the oil tea tree to be picked. Determine the grasping stability correction factor based on the fruit quality and environmental parameters, and use the grasping stability correction factor to correct the stability reference value to obtain a comprehensive stability index. The fruit quality is the average quality of the fruits of the oil tea tree to be picked, and the environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located; Step 5: Compare the corrected comprehensive stability index with a preset stability threshold. According to the comparison result, judge the stability level of the vibrating manipulator to complete the detection of the grasping stability of the oil tea tree vibrating manipulator.

2. The method for detecting the grasping stability of a vibrating mechanical hand for oil tea trees according to claim 1, wherein: The generation method of the sample data set is: Map the working parameters of each grasping operation with the effective frictional force and the proportion of branch damage caused one by one to form a data chain, and record the formed data chain as the sample data set; Based on the data in the sample data set, establish a neural network model. Among them, select the long short-term memory network LSTM model as the basic model to establish, and select the activation function and optimization algorithm. Select the Tanh function as the activation function and select Adam as the optimization algorithm for the LSTM model; The expression of the Tanh function is as follows: ； In the formula, represents the Tanh function, and the independent variable represents the input weighted sum of the neuron, that is, the result after weighted summation of the inputs received by the neuron from the previous layer; at the same time, set the hyperparameters of the LSTM model, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer; among them, the number of network layers is set to a three-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32; Randomly divide the sample data set into a training set and a test set. The working parameters include the grasping force, end acceleration, vibration frequency, and clamping width; Construct a grasping stability detection model based on the deep learning algorithm. Use the working parameters in the training set as the input, and use the effective frictional force and the corresponding proportion of branch damage as labels to train the model to obtain a trained grasping stability detection model. Substitute the working parameters in the test set into the trained model to obtain the corresponding prediction results; Calculate the error between the prediction results and the actual values in the test set; Judge whether the error meets the preset error threshold; If it meets, output the trained model, that is, the grasping stability detection model; If it does not meet, return to continue training; The error is the mean absolute error, root mean square error, and coefficient of determination between the prediction results and the actual values in the test set; The process of randomly dividing into a training set and a test set, the specific logic is: randomly sort the sample data set, take 80% of the sorted sample data set as the training set, and the remaining 20% as the test set; The input of the grasping stability detection model after training is the working parameters, and the output is the effective friction force and the proportion of branch damage.

3. A method for detecting the grasping stability of a vibrating manipulator for oil tea trees according to claim 1, characterized in that: The working parameters include the grasping force, the end acceleration, the vibration frequency, and the clamping width, where: The grasping force is the normal pressure generated at the contact surface between the manipulator gripper and the branch, and the average value of the grasping force in the complete picking cycle is used as the grasping force for this grasping operation; The end acceleration is the vibration acceleration of the end effector of the manipulator. The acceleration time-series data of the manipulator during the vibration picking stage is collected, and the average acceleration during the stable vibration period is used as the end acceleration of the vibrating manipulator. Among them, the stable vibration period refers to the vibration stage in which the fluctuation range of the vibration excitation frequency applied by the manipulator does not exceed and the duration is ≥ 3 seconds; The vibration frequency is the vibration excitation frequency actively applied by the manipulator, and the average value of the vibration frequency during the stable vibration period during the grasping operation is used as the vibration frequency for this grasping operation; The clamping width is the net contact distance between the inner sides of the two grippers, that is, the distance between the inner sides of the two grippers during the stable vibration period after applying the rated grasping force; The effective friction force refers to the friction force component that actually plays an anti-slip role between the manipulator gripper and the branch contact surface. During the complete grasping operation cycle, the average value of the effective friction force during the stable vibration period is taken as the output value of the grasping stability detection model; The specific logic for calculating the effective friction force is: Measure the normal pressure component of the manipulator gripper on the branch through a six-axis force sensor , and deduct the influence of the self-weight of the manipulator. The formula is as follows: ； In the formula, is the original output force vector of the sensor, represents the modulus of the force vector, is the inclination angle of the contact surface between the gripper and the branch, is the mass of the gripper, is the acceleration due to gravity; Determine the effective friction force according to the following formula: ； In the formula, is the effective frictional force, is the friction coefficient, which is calibrated through experiments, is the normal pressure component of the manipulator gripper on the tree branch; The proportion of branch damage refers to the ratio of the area of the damaged branch epidermis or structural damage caused by the manipulator grasping to the total area of the contact area. Its calculation formula is as follows: ； In the formula, is the proportion of branch damage, is the area of skin breakage or structural damage of the branch caused by the manipulator gripping, is the actual contact area between the jaw and the branch.

4. A method for detecting the grasping stability of a vibrating manipulator for oil tea trees according to claim 1, characterized in that: Obtain the working parameters of the vibrating manipulator under the current grasping operation, and input them into the trained grasping stability detection model. The model outputs the predicted value of the effective frictional force and the predicted value of the proportion of branch damage ; Determine the stability reference value for this grasping operation based on the predicted value of the effective friction force and the predicted value of the proportion of branch damage. The formula is as follows: ； In the formula, is the stability reference value, is the predicted value of the effective friction force, is the predicted value of the proportion of branch damage, and are preset weights, , and satisfy , is the optimal effective friction force value under the current working condition, is the allowable deviation range coefficient of the effective friction force, has a value range of ; Among them, the determination is based on the following function: ； In the formula, is the friction coefficient, which is calibrated through experiments, is the mass of the jaw, is the acceleration due to gravity.

5. A method for detecting the grasping stability of a vibrating manipulator for oil tea trees according to claim 1, characterized in that: Obtain the fruit mass and environmental parameters of the oil tea tree to be picked. Determine the grasping stability correction factor based on the fruit mass and environmental parameters. The fruit mass is the average mass of the fruits of the oil tea tree to be picked. The environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located; collect the temperature, wind speed, and humidity data multiple times at the same moment, and take the average value as the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located; Among them, the formula for determining the grasping stability correction factor is as follows: ； In the formula, is the grasping stability correction factor, is the fruit mass, is the rated load weight of the manipulator, which is determined according to the mechanical design parameters, is the temperature, is the reference temperature, is the humidity, is the reference humidity, is the wind speed, is the reference wind speed; is the mass influence index, which is calibrated through load experiments and reflects the non-linear influence of mass change on stability; is the wind speed attenuation coefficient, which is calibrated through wind tunnel experiments and reflects the suppression intensity of wind speed on stability.

6. A method for detecting the gripping stability of a vibrating manipulator for oil tea trees according to claim 5, characterized in that: Use the grasping stability correction factor to correct the stability reference value to obtain the comprehensive stability index. The formula is as follows: ； In the formula, is the comprehensive stability index, is the stability reference value, is the grasping stability correction factor, is the preset weight, and .

7. A method for detecting the gripping stability of a vibrating manipulator for oil tea trees according to claim 1, characterized in that: Compare the corrected comprehensive stability index with the preset stability threshold, and judge the stability level of the vibrating manipulator according to the comparison result; Among them, the logic for comparing the corrected comprehensive stability index with the preset stability threshold is specifically: When it means that the comprehensive stability index is less than the threshold at this time, indicating that the current stability of the vibrating manipulator is poor, and the stability level of the manipulator is determined as: unstable; When it means that the comprehensive stability index is equal to the threshold at this time, indicating that the current stability of the vibrating manipulator is in a boundary state, and the stability level of the manipulator is determined to be: acceptable; When it means that the comprehensive stability index is greater than the threshold at this time, indicating that the current stability of the vibrating manipulator is good, and the stability level of the manipulator is determined to be: stable; In the formula, is the comprehensive stability index, is the stability threshold.

8. A method for detecting the gripping stability of a vibrating manipulator for oil tea trees according to claim 7, characterized in that: The method for determining the stability threshold is: collect the historical stability data of the manipulator under different working conditions, calculate the mean and standard deviation of the comprehensive stability index through statistical analysis, and determine the stability threshold according to the following formula: ； In the formula, is the stability threshold, is the mean of the comprehensive stability index, is the standard deviation of the comprehensive stability index, is the safety factor, and its usual value range is from 1 to 2 to ensure a certain safety margin in practical applications; Conduct experiments under the set operating conditions to observe the stability performance of the manipulator, verify the effectiveness of the set threshold through experimental data, ensure that the threshold can accurately distinguish the three stability levels of unstable, acceptable, and stable, evaluate and adjust the set stability threshold, and continuously monitor its comprehensive stability index during the actual operation of the manipulator, and regularly adjust the stability threshold according to the real-time feedback and performance to ensure its adaptability and effectiveness.

9. A detection device for the grasping stability of a vibrating mechanical hand of an oil tea tree, characterized in that: The described oil tea tree vibration manipulator grasping stability detection device is used to execute the oil tea tree vibration manipulator grasping stability detection method according to any one of claims 1-8, and includes: A data acquisition module, which is used to obtain the working parameters, effective friction force, and the proportion of branch damage caused by a multi-group vibration manipulator during a complete grasping operation, map the working parameters, effective friction force, and the proportion of branch damage for each grasping operation one by one, and generate a sample data set. The working parameters include grasping force, end acceleration, vibration frequency, and clamping width; A grasping stability detection model construction module, which based on the data in the sample data set, establishes a neural network model, uses the working parameters in the sample data set as the input of the model, and uses the effective friction force and the proportion of branch damage in the sample data set as labels to train the neural network model to obtain a grasping stability detection model; A stability prediction module, which is used to obtain the working parameters of the vibration manipulator during the current grasping operation, input them into the trained grasping stability detection model, and the model outputs the predicted value of the effective friction force and the predicted value of the proportion of branch damage. Based on the predicted value of the effective friction force and the predicted value of the proportion of branch damage, determine the stability reference value for this grasping operation. The working parameters include grasping force, end acceleration, vibration frequency, and clamping width; An environmental dynamic correction module, which is used to obtain the fruit quality and environmental parameters of the oil tea tree to be picked, determine the grasping stability correction factor based on the fruit quality and environmental parameters, and use the grasping stability correction factor to correct the stability reference value to obtain a comprehensive stability index. The fruit quality is the average quality of the fruits of the oil tea tree to be picked, and the environmental parameters are specifically the temperature, wind speed, and humidity of the environment where the oil tea tree to be picked is located; A stability determination module, which is used to compare the corrected comprehensive stability index with a preset stability threshold, and judge the stability level of the vibration manipulator according to the comparison result, thereby completing the detection of the grasping stability of the oil tea tree vibration manipulator.

Images