A clamping force global sensing device and calibration method

Abstract

The present invention discloses a clamping force full-range sensing device and calibration method, which belongs to the field of clamping force sensing technology. It includes a clamping force sensing module, a clamping force clamping claw and a clamping body. The clamping force sensing module is located between the clamping body and the clamping claw. The clamping force sensing module is provided with a parallelogram flexible hinge mechanism. The parallelogram flexible hinge mechanism is provided with a parallelogram flexible hinge moving end and a parallelogram flexible hinge fixed end. The parallelogram flexible hinge fixed end is fixed to the clamping body through a sensing module clamping adapter plate, and the parallelogram flexible hinge moving end is connected to the clamping claw through a sensing module clamping claw adapter plate. The present invention adopts the above-mentioned clamping force full-range sensing device and calibration method to convert the clamping force into a linear micro-displacement in a single direction through the translational characteristics of the flexible hinge structure, and then uses a piezoelectric / piezoresistive sensor to detect the magnitude of the clamping force.

Description

Technical Field

[0001] The present invention relates to the field of clamping force sensing technology, and in particular to a clamping force global sensing device and a calibration method. Background Art

[0002] Gripping force sensing is a core technology in robotics, automation equipment, and precision manipulation. Gripping force sensing methods are primarily categorized as indirect and direct. Indirect sensing measures and analyzes other parameters related to the gripping force and then estimates the gripping force using algorithms or models. The accuracy of indirect gripping force sensing methods depends on the accuracy of the established models and algorithms, and is significantly affected by factors such as system nonlinearity and uncertainty.

[0003] For some complex working conditions and changing environments, the model may need to be continuously adjusted and optimized. Force sensors for direct gripping force sensing are typically installed at the gripper's fingertips. When an object is gripped, the sensor directly senses the force and converts information such as the force magnitude into measurable signals such as electrical and optical signals, thereby obtaining the gripping force value in real time. However, this method has measurement blind spots and is sensitive to the gripping point location, making it difficult to achieve accurate sensing in different gripping modes. Summary of the Invention

[0004] The purpose of the present invention is to provide a full-range sensing device and calibration method for clamping force, which converts the clamping force into a linear micro-displacement in a single direction through the translational characteristics of the flexible hinge structure, and then uses a piezoelectric / piezoresistive sensor to detect the magnitude of the clamping force.

[0005] The present invention provides a clamping force full-range sensing device and calibration method, including a clamping force sensing module, a clamping force jaw and a clamper body. The clamping force sensing module is located between the clamper body and the clamper jaw. The clamping force sensing module is provided with a parallelogram flexible hinge mechanism. The parallelogram flexible hinge mechanism is provided with a parallelogram flexible hinge moving end and a parallelogram flexible hinge fixed end. The parallelogram flexible hinge fixed end is fixed to the clamper body through a sensing module clamper adapter plate. The parallelogram flexible hinge moving end is connected to the clamper jaw through a sensing module jaw adapter plate. The sensor modules in the clamping force sensing module are installed on both sides of the parallelogram flexible hinge moving end.

[0006] Preferably, a clamper mechanical interface is provided on the upper portion of the clamper body, and a clamper electrical interface is provided on one side of the clamper body.

[0007] Preferably, the clamping force sensing module includes a sensor module, a sensing module clamper adapter plate, a parallelogram flexible hinge mechanism, a sensor module pre-tightening support, a sensor module pre-tightening screw, a sensing module clamp adapter plate, and a sensing module side plate; the clamper clamp is connected to the sensing module clamp adapter plate, the sensing module clamp adapter plate is connected to the clamping force sensing module side plate and is fixed to the moving end of the parallelogram flexible hinge mechanism by screws; the sensor module is installed on both sides of the parallelogram flexible hinge moving end through the sensor module pre-tightening support.

[0008] Preferably, the sensor module includes a piezoelectric / piezoresistive sensor, a sensor module pre-tightening adjustment block and a sensor module push block. The sensor module pre-tightening adjustment block is provided on one side of the piezoelectric / piezoresistive sensor, and the sensor module push block is provided on the other side of the piezoelectric / piezoresistive sensor.

[0009] Preferably, the parallelogram flexible hinge mechanism includes a parallelogram flexible hinge, a parallelogram flexible hinge moving end, a parallelogram flexible hinge fixed end and a clamping force sensing module side panel fixing screw hole.

[0010] Preferably, the piezoelectric / piezoresistive sensor is a thin film sensor, and the output differential voltage signal is:

[0011] ;

[0012] 、 is the voltage value in the initial preload state; 、 is the voltage value after clamping force;

[0013] Differential voltage signal The Gaussian process regression model is used to map the clamping force value, and the model satisfies:

[0014] ;

[0015] in, is the clamping force mapping function; is the force-voltage conversion function that needs to be fitted; is the mean function; is the covariance function; is a Gaussian process.

[0016] Preferably, the parallelogram flexible hinge mechanism is a symmetrical double parallelogram structure, and the moving end of the parallelogram flexible hinge and the fixed end of the parallelogram flexible hinge are connected by a flexible hinge.

[0017] Preferably, the method comprises the following steps:

[0018] Step S1: Screw in the pre-tightening screw to push the pre-tightening adjustment block to move along the pre-tightening direction, apply the initial pre-tightening force to the piezoelectric / piezoresistive sensors on both sides, and record the initial voltage value. and ;

[0019] Step S2: Apply a known clamping force F at different positions of the clamping jaws, and collect the corresponding differential voltage signals as follows:

[0020] ;

[0021] 、 is the real-time voltage under stress; 、 It is the reference voltage of preload state;

[0022] Data alignment:

[0023] Generate a cubic spline interpolation function for the acquired differential voltage signal, and then resample it according to the acquisition frequency of the known clamping force;

[0024] Normalize the data:

[0025] ;

[0026] is the normalized differential voltage; is the mean value of the differential voltage; is the standard deviation of the differential voltage;

[0027] Denoising and smoothing:

[0028] ;

[0029] is the voltage signal after filtering; is the standardized voltage sequence; n is the filter window half-parameter; k is the time offset; t is the current time point;

[0030] Divide the processed data into training data set and validation data set;

[0031] Step S3, training the Gaussian process model;

[0032] ;

[0033] in, is the clamping force mapping function; is the force-voltage conversion function that needs to be fitted; is the mean function; is the covariance function; is a Gaussian process;

[0034] Using the covariance function to establish the correlation between the differential input signal and the clamping force;

[0035] ;

[0036] Establish the correlation between the differential input signal and the clamping force, is the maximum variance of the sensor output, is the length scale;

[0037] Use the training dataset to train the Gaussian process model and optimize the hyperparameters of the covariance function. , its log-marginal likelihood function;

[0038] ;

[0039] is the known clamping force matrix; The covariance matrix is ​​obtained after training; is the observation noise; is the total covariance matrix; is the input matrix for training; Indicates the number of samples in the training sample set;

[0040] By maximizing the marginal likelihood function Determine optimal hyperparameters;

[0041] Step S4: using the trained Gaussian process regression model, inputting a new sensor signal, and outputting a predicted value of the clamping force;

[0042] ;

[0043] is the predicted clamping force; is the new input value; is the input matrix for training; The covariance matrix is ​​obtained after training; is the observation noise; is the total covariance matrix; is the known clamping force matrix;

[0044] Calculate the prediction variance to assess reliability:

[0045] ;

[0046] is the covariance of the new input value with respect to itself; is the covariance of the new input value with respect to all training values; is the observation noise; is the total covariance matrix; is the covariance of all training values ​​with respect to the new input value.

[0047] Preferably, the parallelogram flexible hinge mechanism can be replaced by a hinge structure with different stiffness parameters.

[0048] Therefore, the present invention adopts the above-mentioned clamping force full-range sensing device and calibration method to convert the clamping force into a linear micro-displacement in a single direction through the translational characteristics of the flexible hinge structure, and then uses a piezoelectric / piezoresistive sensor to detect the magnitude of the clamping force.

[0049] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a schematic diagram of the overall structure of a clamping force global sensing device and calibration method according to the present invention;

[0051] Figure 2 This is an exploded schematic diagram of the structure of a clamping force sensing module of a clamping force global sensing device and calibration method of the present invention;

[0052] Figure 3 This is an exploded schematic diagram of a sensor module of a clamping force global sensing device and calibration method according to the present invention;

[0053] Figure 4 A schematic diagram of the structure of a parallelogram flexible hinge mechanism of a clamping force global sensing device and calibration method according to the present invention;

[0054] Figure 5 This is a schematic diagram of another parallelogram flexible hinge mechanism structure of a clamping force global sensing device and calibration method of the present invention;

[0055] Figure 6 The figure is a flow chart of a clamping force global sensing device and calibration method according to the present invention.

[0056] Reference numerals

[0057] 1. Gripping force sensing module; 2. Gripper jaws; 3. Gripper body; 4. Gripper electrical interface; 5. Gripper mechanical interface; 11. Sensor module; 12. Gripper adapter plate for sensing module; 13. Parallelogram flexible hinge mechanism; 14. Sensor module pre-tightening support; 15. Sensor module pre-tightening screw; 16. Gripper adapter plate for sensing module; 17. Gripping force sensing module side panel; 111. Piezoelectric / piezoresistive sensor; 112. Sensor Module pre-tightening adjustment block; 113, sensor module push block; 131, parallelogram flexible hinge; 132, parallelogram flexible hinge motion end; 133, parallelogram flexible hinge fixed end; 134, clamping force sensing module side panel fixing screw hole; 6, another parallelogram flexible hinge mechanism; 61, flexible hinge; 62, parallelogram flexible hinge motion output end; 63, parallelogram flexible hinge jaw connection platform; 64, pre-tightening screw hole. DETAILED DESCRIPTION

[0058] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0059] Unless otherwise defined, technical or scientific terms used in the present invention shall have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs.

[0060] The words "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. Words such as "include" or "comprise" mean that the elements or objects preceding the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0061] Example 1

[0062] like Figure 1-6 As shown in the figure, a global gripping force sensing device and calibration method of the present invention includes a gripping force sensing module 1, a gripping force jaw 2, and a gripper body 3. A gripper mechanical interface 5 is provided above the gripper body 3, and a gripper electrical interface 4 is provided on one side of the gripper body 3. The gripper electrical interface 4 completes the electrical connection between the gripper drive and the force sensing sensor, and the gripper mechanical interface 5 completes the connection between the gripper and the robot.

[0063] The gripping force sensing module 1 includes a sensor module 11, a sensing module gripper adapter plate 12, a parallelogram flexible hinge mechanism 13, a sensor module preload support 14, a sensor module preload screw 15, a sensing module jaw adapter plate 16, and a sensing module side plate 17. The gripper jaw 2 is connected to the sensing module jaw adapter plate 16, which is then connected to the gripping force sensing module side plate 17 and secured to the parallelogram flexible hinge motion end 132 via screws.

[0064] The sensor module 11 is mounted on both sides of the parallelogram flexible hinge mechanism's moving end 132 via sensor module preload supports 14. The sensor module 11 includes a piezoelectric / piezoresistive sensor 111, a sensor module preload adjustment block 112, and a sensor module push block 113. The sensor module preload adjustment block 112 is located on one side of the piezoelectric / piezoresistive sensor 111, and the sensor module push block 113 is located on the other side of the piezoelectric / piezoresistive sensor 111. To preload the sensor module 11, the piezoelectric / piezoresistive sensor 111 is subjected to a preload force, enabling it to detect an initial pressure value.

[0065] The clamping force sensing module 1 is located between the clamp body 3 and the clamp jaw 2. It is equipped with a parallelogram flexible hinge mechanism 13. The parallelogram flexible hinge mechanism 13 includes a parallelogram flexible hinge 131, a parallelogram flexible hinge moving end 132, a parallelogram flexible hinge fixed end 133, and a clamping force sensing module side panel fixing screw hole 134.

[0066] The parallelogram flexible hinge mechanism 13 is equipped with a parallelogram flexible hinge moving end 132 and a parallelogram flexible hinge fixed end 133. The parallelogram flexible hinge fixed end 133 is fixed to the gripper body 3 via the sensing module gripper adapter plate 12. The parallelogram flexible hinge moving end 132 is connected to the gripper jaw 2 via the sensing module jaw adapter plate 16. The sensor modules 11 in the clamping force sensing module 1 are mounted on both sides of the parallelogram flexible hinge moving end 132. The parallelogram flexible hinge mechanism 13 is a symmetrical double parallelogram structure, with the parallelogram flexible hinge moving end 132 and the parallelogram flexible hinge fixed end 133 connected by a flexible hinge.

[0067] The parallelogram flexible hinge mechanism 13, when the parallelogram flexible hinge moving end 132 is subjected to force, the parallelogram flexible hinge 131 undergoes a slight deformation, and one side of the parallelogram flexible hinge moving end 132 will increase the squeezing of the sensor module push block 113, while the other side will reduce the preload force to squeeze the piezoelectric / piezoresistive sensor 111, thereby forming a differential pressure signal.

[0068] The piezoelectric / piezoresistive sensor 111 is a thin film sensor, and its output differential voltage signal is:

[0069] ;

[0070] 、 is the voltage value in the initial preload state; 、 is the voltage value after clamping force;

[0071] Differential voltage signal The Gaussian process regression model is used to map the clamping force value, and the model satisfies:

[0072] ;

[0073] in, is the mean function; is the covariance function; is the clamping force mapping function; is the force-voltage conversion function that needs to be fitted; is a Gaussian process.

[0074] A calibration method for a clamping force global sensing device comprises the following steps:

[0075] Step S1: Screw in the sensor module pre-tightening screw 15 to push the sensor module pre-tightening adjustment block 112 to move along the inclined surface. Apply initial pre-tightening force to the piezoelectric / piezoresistive sensors 111 on both sides and record the initial voltage value. and ;

[0076] Step S2: Apply a known clamping force at different positions of the clamping jaw 2, and collect the corresponding differential voltage signal as follows:

[0077] ;

[0078] 、 is the real-time voltage under stress; 、 It is the reference voltage of preload state;

[0079] Data alignment:

[0080] Generate a cubic spline interpolation function for the acquired differential voltage signal, and then resample it according to the acquisition frequency of the known clamping force;

[0081] Normalize the data:

[0082] ;

[0083] is the normalized differential voltage; is the mean value of the differential voltage; is the standard deviation of the differential voltage;

[0084] Denoising and smoothing:

[0085] ;

[0086] is the voltage signal after filtering; is the standardized voltage sequence; n is the filter window half-parameter; k is the time offset; t is the current time point;

[0087] Divide the processed data into training data set and validation data set;

[0088] Step S3, training the Gaussian process model;

[0089] ;

[0090] in, is the clamping force mapping function; is the force-voltage conversion function that needs to be fitted; is the mean function; is the covariance function; is a Gaussian process;

[0091] Using the covariance function to establish the correlation between the differential input signal and the clamping force;

[0092] ;

[0093] Establish the correlation between the differential input signal and the clamping force, is the maximum variance of the sensor output, is the length scale.

[0094] Use the training dataset to train the Gaussian process model and optimize the hyperparameters of the covariance function. , its log-marginal likelihood function;

[0095] ;

[0096] is the known clamping force matrix; The covariance matrix is ​​obtained after training; is the observation noise; is the total covariance matrix; is the input matrix for training; Indicates the number of samples in the training sample set;

[0097] By maximizing the marginal likelihood function Determine optimal hyperparameters.

[0098] Use the validation dataset for testing and calculate the error (RMSE / MAE);

[0099] RMSE (Root Mean Square Error) root mean square error;

[0100] ;

[0101] MAE (Mean Absolute Error) mean absolute error;

[0102] ;

[0103] Step S4: using the trained Gaussian process regression model, inputting a new sensor signal, and outputting a predicted value of the clamping force;

[0104] ;

[0105] is the predicted clamping force; is the new input value; is the input matrix for training; The covariance matrix is ​​obtained after training; is the observation noise; is the total covariance matrix; is the known clamping force matrix;

[0106] Calculate the prediction variance to assess reliability:

[0107] .

[0108] is the covariance of the new input value with respect to itself; is the covariance of the new input value with respect to all training values; is the observation noise; is the total covariance matrix; is the covariance of all training values ​​with respect to the new input value.

[0109] The parallelogram flexible hinge mechanism 13 can be replaced by a hinge structure with different stiffness parameters.

[0110] The parallelogram flexible hinge mechanism 13 can be replaced by a hinge mechanism with different stiffness parameters. Figure 5The parallelogram flexible hinge mechanism 6 is a replaceable hinge mechanism. The parallelogram flexible hinge mechanism 6 includes a flexible hinge 61, a parallelogram flexible hinge motion output end 62, a parallelogram flexible hinge clamping claw connection platform 63, and a pre-tightening screw hole 64. The parallelogram flexible hinge clamping claw connection platform 63 is located at the center of the parallelogram flexible hinge mechanism 6, and the flexible hinge 61 is located on the outside of the parallelogram flexible hinge clamping claw connection platform 63. The parallelogram flexible hinge clamping claw connection platform 63 and the pre-tightening screw hole 64 are provided on the left side of the parallelogram flexible hinge clamping claw connection platform 63. Through the above replacement, while maintaining the original installation structure, higher precision, stronger anti-interference performance, and longer life of clamping force detection can be achieved.

[0111] Therefore, the present invention utilizes the aforementioned global clamping force sensing device and calibration method, converting the clamping force into a single-direction linear micro-displacement through the translational characteristics of the flexible hinge structure. This is then used to detect the clamping force using a piezoelectric / piezoresistive sensor. This addresses the blind spots in clamping force measurement and the sensitivity to the clamping point position that exist with force sensors installed at the gripper fingertips.

[0112] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can still be modified or replaced by equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A clamping force global sensing device, characterized in that: The clamping force sensing module comprises a clamping force clamping claw and a clamping body. The clamping force sensing module is located between the clamping body and the clamping claw. The clamping force sensing module is provided with a parallelogram flexible hinge mechanism. The parallelogram flexible hinge mechanism is provided with a parallelogram flexible hinge moving end and a parallelogram flexible hinge fixed end. The parallelogram flexible hinge fixed end is fixed to the clamping body through a clamping adapter plate of the sensing module. The parallelogram flexible hinge moving end is connected to the clamping claw through a clamping claw adapter plate of the sensing module. The sensor modules in the clamping force sensing module are installed on both sides of the parallelogram flexible hinge moving end. The clamping force sensing module includes a sensor module, a sensing module clamp adapter plate, a parallelogram flexible hinge mechanism, a sensor module pre-tightening support, a sensor module pre-tightening screw, a sensing module clamp adapter plate, and a sensing module side plate; the clamping jaws of the clamp are connected to the sensing module clamp adapter plate, the sensing module clamp adapter plate is connected to the clamping force sensing module side plate and is fixed to the parallelogram flexible hinge motion end by screws; the sensor module is installed on both sides of the parallelogram flexible hinge motion end through the sensor module pre-tightening support; The parallelogram flexible hinge mechanism is a symmetrical double parallelogram structure, and the parallelogram flexible hinge moving end and the parallelogram flexible hinge fixed end are connected by a flexible hinge; The sensor module includes a piezoelectric / piezoresistive sensor, a sensor module pre-tightening adjustment block and a sensor module pushing block. The sensor module pre-tightening adjustment block is provided on one side of the piezoelectric / piezoresistive sensor, and the sensor module pushing block is provided on the other side of the piezoelectric / piezoresistive sensor.

2. A clamping force global sensing device according to claim 1, characterized in that: A clamper mechanical interface is provided on the upper portion of the clamper body, and a clamper electrical interface is provided on one side of the clamper body.

3. The clamping force global sensing device according to claim 1, characterized in that: The parallelogram flexible hinge mechanism comprises a parallelogram flexible hinge, a parallelogram flexible hinge moving end, a parallelogram flexible hinge fixed end and a clamping force sensing module side plate fixing screw hole.

4. The clamping force global sensing device according to claim 1, characterized in that: The piezoelectric / piezoresistive sensor is a thin film sensor that outputs a differential voltage signal: ； 、 is the real-time voltage under stress; 、 It is the reference voltage of preload state; Differential voltage signal The Gaussian process regression model is used to map the clamping force value, and the model satisfies: ； in, is the mean function; is the covariance function; is the clamping force mapping function; is the force-voltage conversion function that needs to be fitted; is a Gaussian process.

5. A calibration method for a clamping force global sensing device according to any one of claims 1 to 4, characterized in that: The following steps are involved: Step S1: Screw in the pre-tightening screw to push the pre-tightening adjustment block to move along the pre-tightening direction, apply the initial pre-tightening force to the piezoelectric / piezoresistive sensors on both sides, and record the initial voltage value. and ; Step S2: Apply a known clamping force F at different positions of the clamping jaws, and collect the corresponding differential voltage signals as follows: ； 、 is the real-time voltage under stress; 、 It is the reference voltage of preload state; Data alignment: Generate a cubic spline interpolation function for the acquired differential voltage signal, and then resample it according to the acquisition frequency of the known clamping force; Normalize the data: ； is the normalized differential voltage; is the mean value of the differential voltage; is the standard deviation of the differential voltage; Denoising and smoothing: ； is the voltage signal after filtering; is the standardized voltage sequence; n is the filter window half-parameter; k is the time offset; t is the current time point; Divide the processed data into training data set and validation data set; Step S3, training the Gaussian process model; ； in, is the clamping force mapping function; is the force-voltage conversion function that needs to be fitted; is the mean function; is the covariance function; is a Gaussian process; Using the covariance function to establish the correlation between the differential input signal and the clamping force; ； is the maximum variance of the sensor output, is the length scale; Use the training dataset to train the Gaussian process model and optimize the hyperparameters of the covariance function. , its log-marginal likelihood function; is the known clamping force matrix; The covariance matrix is ​​obtained after training; is the observation noise; is the total covariance matrix; is the input matrix for training; Indicates the number of samples in the training sample set; By maximizing the marginal likelihood function Determine optimal hyperparameters; Step S4: using the trained Gaussian process regression model, inputting a new sensor signal, and outputting a predicted value of the clamping force; ； is the predicted clamping force; is the new input value; is the input matrix for training; The covariance matrix is ​​obtained after training; is the observation noise; is the total covariance matrix; is the known clamping force matrix; Calculate the prediction variance to assess reliability: ； is the covariance of the new input value with respect to itself; is the covariance of the new input value with respect to all training values; is the observation noise; is the total covariance matrix; is the covariance of all training values ​​with respect to the new input value.

6. The clamping force global sensing device according to claim 1, characterized in that: The parallelogram flexible hinge mechanism can be replaced by hinge structures with different stiffness parameters.

Images