A robot operation posture control method and system based on tactile sensor

Abstract

The present invention relates to the field of robot control technology, and in particular to a robot operation posture control method and system based on a tactile sensor, comprising the following steps: S1: collecting tactile data of the contact area between the robot end effector and the target object; S2: filtering the tactile data; S3: extracting relevant tactile features from the data processed by S2; S4: comparing the tactile features extracted by S3 with the preset standard tactile state, and calculating the posture error vector; S5: generating an adjustment instruction according to the posture error vector calculated by S4; S6: executing the adjustment instruction to adjust the posture of the robot end effector. The present invention uses a high-resolution multi-point tactile sensor to collect and accurately process tactile data in real time, and combines a PID control algorithm to realize dynamic adjustment of posture errors, thereby significantly improving the operation accuracy and stability of the robot in a complex dynamic environment.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a robot operation posture control method and system based on a tactile sensor. Background Art

[0002] With the rapid development of robotics technology, robots are increasingly used in industrial manufacturing, medical assistance, service industries and other fields. In these application scenarios, the precise operation capability of robots, especially in complex and dynamic environments, has become an important indicator to measure their performance. Traditional robot control methods mostly rely on visual sensors or torque sensors to perceive environmental information. Although these sensors can provide certain perception capabilities, they often face problems such as insufficient accuracy in visual information processing, torque perception lag and environmental interference in actual applications. In order to improve the operating accuracy and flexibility of robots, tactile sensors have gradually been introduced into robot control systems in recent years as an important supplementary sensor. Tactile sensors can obtain contact force, pressure distribution, contact area and torque information in real time, and provide detailed tactile feedback for the interaction between the robot's end effector and the target object, thereby helping the robot to more accurately perceive the state of objects and environmental changes when performing tasks.

[0003] However, existing tactile perception systems still have many problems in practical applications, especially in real-time posture control and precise feedback adjustment. In existing technologies, tactile data often need to go through complex data processing and feature extraction to provide valuable information for the control system, but these processing processes are usually limited by the efficiency and real-time performance of the algorithm, which makes it difficult for the system to meet the needs of high-speed motion and high-precision operation. In addition, traditional control systems often rely on a single control method, fail to make full use of tactile feedback information for dynamic adjustment, and lack fine-grained adjustment of the posture of the robot's end effector. Summary of the invention

[0004] Based on the above objectives, the present invention provides a robot operation posture control method and system based on tactile sensors.

[0005] A robot operation posture control method based on a tactile sensor comprises the following steps:

[0006] S1: Uses high-resolution multi-point tactile sensors to collect tactile data of the contact area between the robot end effector and the target object in real time, including pressure distribution, contact area and contact torque;

[0007] S2: Filter the tactile data collected by S1 to remove noise and interference;

[0008] S3: extract relevant tactile features from the data processed by S2, including the mean and variance of the pressure distribution and the direction and magnitude of the contact torque;

[0009] S4: Compare the tactile features extracted by S3 with the preset standard tactile state and calculate the posture error vector;

[0010] S5: Based on the posture error vector calculated by S4, the PID control algorithm is used to generate adjustment instructions to smoothly respond to the error change;

[0011] S6: Execute the adjustment command generated by S5 to adjust the position and posture of the robot end effector in real time, while continuously monitoring the feedback data from the tactile sensor, and input the new tactile data into S2 to form a control closed loop.

[0012] Optionally, the S1 specifically includes:

[0013] S11: Install high-resolution multi-point tactile sensors on the contact part of the robot end effector to ensure full coverage of the contact area with the target object;

[0014] S12: measuring and recording the pressure distribution of the contact area in real time through each sensing point in the tactile sensor array;

[0015] S13: Using the contact area detection function of the tactile sensor, the contact area between the robot end effector and the target object is calculated in real time. The calculation formula is: ,in, is the contact area, For the The pressure value measured at each sensing point, For the The area weight of each sensor point, is the total number of sensing points in the tactile sensor array;

[0016] S14: The direction and magnitude of the contact torque are monitored and recorded in real time through the built-in torque measurement function of the tactile sensor.

[0017] Optionally, the S2 specifically includes:

[0018] S21: De-noising the tactile data collected by S1, using a median filter algorithm, selecting a neighborhood range of the tactile data and replacing each data point with a median value within the range, thereby removing outliers caused by sensor noise;

[0019] S22: Smoothing the denoised tactile data, and performing weighted averaging on the data using a Gaussian filter algorithm to eliminate low-frequency fluctuations caused by insufficient sensor sampling frequency or environmental interference;

[0020] S23: Segment the tactile data processed by S21 and S22, and segment them according to the set time window based on the change characteristics of the data, so as to ensure that each segment of data has stable characteristics in the time domain;

[0021] S24: high-frequency noise is removed from each section of processed tactile data by using a bandpass filter to remove noise below a set frequency and above a set frequency, thereby ensuring that the remaining data conforms to the frequency spectrum characteristics of the target signal.

[0022] Optionally, the S3 specifically includes:

[0023] S31: Perform statistical analysis on the pressure distribution values ​​in the tactile data processed by S2, and calculate the arithmetic mean of the pressure value of each sensing point. The formula is: ,in, is the mean value of the pressure distribution, For the The pressure value measured at each sensing point, is the total number of sensing points in the tactile sensor array;

[0024] S32: Based on the mean value of the pressure distribution calculated in S31, the variance of the pressure distribution is calculated, and the formula is: ,in, is the variance of the pressure distribution;

[0025] S33: Calculate the direction of the contact torque using the torque component in the tactile data processed by S2, and determine its specific direction in two-dimensional or three-dimensional space by calculating the angle of the torque vector. The formula is: ,in, is the direction angle of the contact torque, and are the components of the moment on the X-axis and Y-axis respectively;

[0026] S34: According to the torque component in the tactile data processed by S2, the magnitude of the contact torque is calculated using the vector modulus calculation method, and the formula is: ,in, is the magnitude of the contact torque, are the components of the torque on the X-axis, Y-axis and Z-axis respectively.

[0027] Optionally, the S4 specifically includes:

[0028] S41: Compare the pressure distribution mean, pressure distribution variance, contact torque direction and magnitude extracted in S3 with the preset standard tactile features one by one, and calculate the difference value between each feature; suppose the standard values ​​of pressure distribution mean, pressure distribution variance, contact torque direction and contact torque magnitude are and , then the difference value comparison formula of each feature is: ; ; ; ,in, and They are the difference values ​​of the mean of pressure distribution, pressure distribution variance, contact moment direction and contact moment magnitude;

[0029] S42: According to the importance of each tactile feature to posture control, they are and Assign weight coefficient and ;

[0030] S43: Combine the weighted difference values ​​into vectors to form a posture error vector , and its calculation formula is: ,in, and They are the unit vector directions corresponding to the four features respectively.

[0031] Optionally, the S5 specifically includes:

[0032] S51: according to the posture error vector calculated in S4, decompose it into two parts: position error and posture error;

[0033] S52: Set the proportional coefficient of the PID controller , integral coefficient and the differential coefficient ;

[0034] S53: Calculate the proportional control amount according to the error vector decomposed in step S51; the calculation formula of the proportional control amount is: ,in, is the proportional control quantity, is the position error component, is the attitude error component;

[0035] S54: The integration result of position error and attitude error is combined with the integration coefficient Multiply them together to get the integral control amount. The calculation formula of the integral control amount is: ,in, is the integral control quantity, and are the functions of position error and attitude error changing with time respectively;

[0036] S55: The rate of error change and the differential coefficient Multiply them to get the differential control amount. The calculation formula of the differential control amount is: ,in, is the differential control quantity, is the rate of change of position error and attitude error over time;

[0037] S56: Add the proportional control amount, integral control amount and differential control amount calculated in S53 to S55 to obtain the final adjustment instruction.

[0038] Optionally, the S51 specifically includes:

[0039] S511: According to the pose error vector calculated in S4 , decompose it into two parts: position error and attitude error;

[0040] S512: By analyzing the first three components of the posture error vector calculated in S4, the position error is obtained. , the error vector represents the deviation between the current position of the robot end effector and the target position; the expression of the position error is: ,in, and They represent the robot end effector Position error in direction;

[0041] S513: By analyzing the last three components of the posture error vector calculated in S4, the posture error is obtained. , the error vector represents the deviation between the posture of the robot end effector and the target posture; the expression of posture error is: ,in, , and They represent the pitch angle error, yaw angle error and roll angle error of the robot end effector respectively.

[0042] Optionally, the S56 specifically includes:

[0043] S561: The proportional control amount calculated in S53 , the integral control amount calculated in S54 And the differential control amount calculated in S55 Summarize and generate final adjustment instructions ; The specific calculation formula is: ,in, For the final adjustment instructions, They are proportional, integral and differential control quantities respectively.

[0044] Optionally, the S6 specifically includes:

[0045] S61: receiving the adjustment instruction generated in step S5 , and convert it into a specific robot end effector posture adjustment operation; adjustment instruction It includes position and attitude corrections, which are applied to each degree of freedom of the robot's end effector;

[0046] S62: According to the adjustment instructions converted in S61, the movement of the robot end effector is controlled in real time, and the position and posture are dynamically adjusted by driving the motor;

[0047] S63: while executing S62 to adjust the posture, continuously collect real-time tactile data of the contact area between the robot end effector and the target object through a high-resolution multi-point tactile sensor;

[0048] S64: The new tactile data collected in S63 is input into step S2 in real time for filtering processing to form a closed-loop control.

[0049] A robot operation posture control system based on a tactile sensor, used to implement the above-mentioned robot operation posture control method based on a tactile sensor, includes the following modules:

[0050] Acquisition module: used to collect tactile data of the contact area between the robot end effector and the target object, the tactile data including pressure distribution, contact area and contact torque;

[0051] Data processing module: used to filter the tactile data collected by the acquisition module to remove noise and interference;

[0052] Feature extraction module: used to extract tactile features from the data processed by the data processing module, including the mean and variance of pressure distribution and the direction and magnitude of contact torque;

[0053] A posture error calculation module: used to compare the tactile features extracted by the feature extraction module with the preset standard tactile state and calculate the posture error vector;

[0054] Control module: used to generate adjustment instructions using PID control algorithm according to the posture error vector calculated by the posture error calculation module, and adjust the posture of the robot end effector in real time according to the adjustment instructions;

[0055] Feedback monitoring module: used to continuously monitor the feedback data of the tactile sensor and input the new tactile data into the data processing module for reprocessing to form a control closed loop.

[0056] Beneficial effects of the present invention:

[0057] The present invention, by adopting high-resolution multi-point tactile sensors, collects and processes tactile data of the contact area between the robot end effector and the target object in real time, and can accurately reflect the contact state between the object and the robot; these tactile data, including pressure distribution, contact area and contact torque, provide richer and more comprehensive feedback information than traditional vision or torque sensors, which helps the robot to perform more precise operations in complex environments and provides a reliable basis for subsequent posture adjustments.

[0058] The present invention, by adopting a posture adjustment method based on a PID control algorithm, can achieve efficient posture correction by accurately calculating the posture error vector and combining the actual feedback of the robot's end effector; by dynamically processing the error vector, a smooth response of the control command is ensured, and over-adjustment or oscillation is effectively avoided, thereby achieving precise control while ensuring the operational stability of the robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0060] Figure 1 A schematic diagram of a robot operation posture control method according to an embodiment of the present invention;

[0061] Figure 2 Schematic diagram of a robot operation posture control system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0062] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0063] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0064] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0065] like Figure 1 As shown, a robot operation posture control method based on a tactile sensor comprises the following steps:

[0066] S1: Use high-resolution multi-point tactile sensors to collect tactile data of the contact area between the robot end effector and the target object in real time, including pressure distribution, contact area and contact torque, to fully reflect the contact status;

[0067] S2: Filter the tactile data collected by S1 to remove noise and interference to ensure the stability and accuracy of subsequent calculations;

[0068] S3: Extract relevant tactile features from the data processed by S2, including the mean and variance of the pressure distribution and the direction and magnitude of the contact torque, to provide a basis for subsequent control;

[0069] S4: Compare the tactile features extracted by S3 with the preset standard tactile state and calculate the posture error vector;

[0070] S5: Based on the posture error vector calculated by S4, the PID control algorithm is used to generate adjustment instructions to smoothly respond to error changes and reduce oscillation;

[0071] S6: Execute the adjustment command generated by S5 to adjust the posture of the robot end effector in real time, while continuously monitoring the feedback data from the tactile sensor, and input the new tactile data into S2 to form a control closed loop to achieve continuously optimized posture control.

[0072] S1 specifically includes:

[0073] S11: Install high-resolution multi-point tactile sensors on the contact part of the robot end effector to ensure full coverage of the contact area with the target object;

[0074] S12: Through each sensing point in the tactile sensor array, the pressure distribution of the contact area is measured and recorded in real time to ensure high accuracy and real-time data;

[0075] S13: Using the contact area detection function of the tactile sensor, the contact area between the robot end effector and the target object is calculated in real time. The calculation formula is: ,in, is the contact area, For the The pressure value measured at each sensing point, For the The area weight of each sensor point, is the total number of sensing points in the tactile sensor array;

[0076] S14: Through the built-in torque measurement function of the tactile sensor, the direction and magnitude of the contact torque are monitored and recorded in real time to obtain comprehensive tactile data; through the above steps S11 to S15, the multi-dimensional tactile data between the robot end effector and the target object can be efficiently and accurately collected. The real-time collection and transmission of such high-resolution multi-point tactile data ensures that the robot can obtain comprehensive and accurate contact information during operation.

[0077] S2 specifically includes:

[0078] S21: De-noising the tactile data collected by S1, using a median filter algorithm, selecting a neighborhood range of the tactile data and replacing each data point with a median value within the range, thereby removing outliers caused by sensor noise;

[0079] S22: Smoothing the denoised tactile data, and performing weighted averaging on the data using a Gaussian filter algorithm to eliminate low-frequency fluctuations caused by insufficient sensor sampling frequency or environmental interference;

[0080] S23: Segment the tactile data processed by S21 and S22, and segment them according to the change characteristics of the data and the set time window. For example, the time window can be set to 100 milliseconds to ensure that each segment of data has stable characteristics in the time domain, and each segment of data is processed independently;

[0081] S24: high-frequency noise is removed from each section of processed tactile data, and a bandpass filter is used to remove noise below the set frequency (20 Hz) and above the set frequency (200 Hz), ensuring that the remaining data conforms to the spectral characteristics of the target signal; the above steps S21 to S24 effectively remove noise and interference in the tactile data through multiple filtering techniques; the combination of median filtering, Gaussian filtering and bandpass filtering ensures the accuracy and stability of the tactile data, and avoids errors caused by environmental interference or limitations of the sensor itself.

[0082] S3 specifically includes:

[0083] S31: Perform statistical analysis on the pressure distribution values ​​in the tactile data processed by S2, and calculate the arithmetic mean of the pressure value of each sensing point. The formula is: ,in, is the mean value of the pressure distribution, For the The pressure value measured at each sensing point, is the total number of sensing points in the tactile sensor array;

[0084] S32: Based on the mean value of the pressure distribution calculated in S31, the variance of the pressure distribution is calculated to evaluate the degree of dispersion of the pressure value. The formula is: ,in, is the variance of pressure distribution, other symbols are the same as S31;

[0085] S33: Calculate the direction of the contact torque using the torque component in the tactile data processed by S2, and determine its specific direction in two-dimensional or three-dimensional space by calculating the angle of the torque vector. The formula is: ,in, is the direction angle of the contact torque, and are the components of the moment on the X-axis and Y-axis respectively;

[0086] S34: According to the torque component in the tactile data processed by S2, the magnitude of the contact torque is calculated using the vector modulus calculation method, and the formula is: ,in, is the magnitude of the contact torque, They are the components of the torque on the X-axis, Y-axis and Z-axis respectively; through steps S31 to S34, key tactile features can be accurately extracted from the filtered tactile data, including the mean and variance of the pressure distribution and the direction and magnitude of the contact torque. These precise tactile features provide a reliable data basis for subsequent posture error calculation and control instruction generation.

[0087] S4 specifically includes:

[0088] S41: Compare the pressure distribution mean, pressure distribution variance, contact torque direction and magnitude extracted in S3 with the preset standard tactile features one by one, and calculate the difference value between each feature; suppose the standard values ​​of pressure distribution mean, pressure distribution variance, contact torque direction and contact torque magnitude are and , then the difference value comparison formula of each feature is: ; ; ; ,in, and They are the difference values ​​of the mean of pressure distribution, pressure distribution variance, contact moment direction and contact moment magnitude;

[0089] S42: According to the importance of each tactile feature to posture control, they are and Assign weight coefficient and ; Weight coefficients are assigned based on pre-set importance to ensure that key features have a greater impact on pose errors: ;

[0090] S43: Combine the weighted difference values ​​into vectors to form a posture error vector , and its calculation formula is: ,in, and They are the unit vector directions corresponding to the four features respectively; through steps S41 to S43, the extracted tactile features can be accurately compared with the preset ideal tactile state, and the posture error can be quantified through weight allocation and error vector calculation. This process ensures the reasonable contribution of each tactile feature in the posture error calculation and improves the accuracy and representativeness of the error vector.

[0091] S5 specifically includes:

[0092] S51: according to the position and posture error vector calculated in S4, decompose it into two parts: position error and attitude error, wherein the position error is a component related to the position in the error vector, and the attitude error is a component related to the rotation;

[0093] S52: Set the proportional coefficient of the PID controller , integral coefficient and the differential coefficient , these coefficients determine the sensitivity of the PID controller to the error response, among which the proportional coefficient is used to adjust the system's response strength to the current error; the integral coefficient is used to correct the long-term accumulated error; the differential coefficient is used to predict the trend of the error change and prevent over-adjustment; the parameters in this step can be optimized by experiment or automatic adjustment; S53: According to the error vector decomposed in step S51, calculate the proportional control amount. The proportional control is based on the product of the current error value, so by multiplying the error component and the proportional coefficient By multiplying, the corresponding proportional adjustment amount can be obtained; the calculation formula of the proportional control amount is: ,in, is the proportional control quantity, is the position error component, is the attitude error component;

[0094] S54: The integral control calculation considers the cumulative effect of error over time. In this step, the integral results of position error and attitude error are compared with the integral coefficient Multiply them together to get the integral control amount. The calculation formula of the integral control amount is: ,in, is the integral control quantity, and are functions of position error and attitude error over time respectively; this integral control quantity helps to eliminate small errors that exist in the system for a long time and avoid deviations in the robot's operating accuracy over time;

[0095] S55: The differential control calculation predicts the speed of error change, helping the system respond to error changes more quickly. Specifically, the rate of error change is calculated by comparing the differential coefficient Multiply them to get the differential control amount. The calculation formula of the differential control amount is: ,in, is the differential control quantity, is the rate of change of position error and attitude error over time; the role of the differential control amount is to suppress the sudden change of the error and avoid instability caused by excessive adjustment of the robot's action;

[0096] S56: Add the proportional control quantity, integral control quantity and differential control quantity calculated in S53 to S55 to obtain the final adjustment instruction. The adjustment instruction is a comprehensive response based on the posture error. After adjustment by the PID controller, it can effectively control the posture adjustment of the robot's end effector to ensure that the robot's end effector accurately reaches the target position and posture; through steps S51 to S56, the adjustment instruction can be accurately generated based on the PID control algorithm, thereby realizing precise posture control of the robot's end effector; through the combined effects of proportion, integration and differentiation, the PID control algorithm can not only cope with the current error, but also correct the long-term accumulated error and predict the future error change trend, thereby optimizing the robot's response speed and stability.

[0097] S51 specifically includes:

[0098] S511: According to the pose error vector calculated in S4 , decompose it into two parts: position error and attitude error; the position and attitude error vector is a six-dimensional vector containing position error and attitude error, where the first three components correspond to the position error and the last three components correspond to the attitude error;

[0099] S512: By analyzing the first three components of the posture error vector calculated in S4, the position error is obtained. , the error vector represents the deviation between the current position of the robot end effector and the target position; the expression of the position error is: ,in, and They represent the robot end effector Position error in direction;

[0100] S513: By analyzing the last three components of the posture error vector calculated in S4, the posture error is obtained. , the error vector represents the deviation between the posture of the robot end effector and the target posture, usually using Euler angles (such as pitch angle, yaw angle, roll angle) or quaternions to represent the posture error; the expression of the posture error is: ,in, , and They respectively represent the pitch angle error, yaw angle error and roll angle error of the robot's end effector; through steps S511 to S513, the posture error vector can be accurately decomposed into two parts: position error and attitude error. The decomposition of position error and attitude error not only helps to understand the robot's operating accuracy deviation, but also enables the control system to optimize and adjust the position and attitude respectively.

[0101] S56 specifically includes:

[0102] S561: The proportional control amount calculated in S53 , the integral control amount calculated in S54 And the differential control amount calculated in S55 Summarize and generate final adjustment instructions ; The specific calculation formula is: ,in, For the final adjustment instructions, They are proportional, integral and differential control quantities respectively.

[0103] S61: receiving the adjustment instruction generated in step S5 , and convert it into a specific robot end effector posture adjustment operation; adjustment instruction Including position and attitude corrections, which are applied to each degree of freedom of the robot's end effector (such as position adjustment of the X, Y, and Z axes and attitude adjustment of the pitch angle, yaw angle, and roll angle) to ensure that the end effector moves and rotates accurately in the direction of the target posture;

[0104] S62: According to the adjustment instructions converted in S61, the movement of the robot end effector is controlled in real time, and the position and posture are dynamically adjusted by driving the motor;

[0105] S63: while executing S62 to adjust the posture, continuously collect real-time tactile data of the contact area between the robot end effector and the target object through a high-resolution multi-point tactile sensor;

[0106] S64: The new tactile data collected in S63 is input into step S2 in real time for filtering and processing to form a closed-loop control; through steps S61 to S64, the real-time adjustment and continuous monitoring of the position of the robot's end effector are realized, forming an efficient control closed loop. This closed-loop control mechanism significantly improves the adaptability and reliability of the robot's operation, ensuring that it can complete tasks stably and efficiently in various application scenarios.

[0107] like Figure 2 As shown, a robot operation posture control system based on a tactile sensor is used to implement the above-mentioned robot operation posture control method based on a tactile sensor, and includes the following modules:

[0108] Acquisition module: used to collect tactile data of the contact area between the robot end effector and the target object. The tactile data includes pressure distribution, contact area and contact torque.

[0109] Data processing module: used to filter the tactile data collected by the acquisition module, remove noise and interference, and ensure the accuracy and stability of the data;

[0110] Feature extraction module: used to extract tactile features from the data processed by the data processing module, including the mean and variance of pressure distribution and the direction and magnitude of contact torque;

[0111] A posture error calculation module: used to compare the tactile features extracted by the feature extraction module with the preset standard tactile state and calculate the posture error vector;

[0112] Control module: used to generate adjustment instructions using PID control algorithm according to the posture error vector calculated by the posture error calculation module, and adjust the posture of the robot end effector in real time according to the adjustment instructions;

[0113] Feedback monitoring module: used to continuously monitor the feedback data of the tactile sensor and input the new tactile data into the data processing module for reprocessing to form a control closed loop.

[0114] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0115] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A robot operation posture control method based on tactile sensor, characterized in that: The following steps are involved: S1: Uses high-resolution multi-point tactile sensors to collect tactile data of the contact area between the robot end effector and the target object in real time, including pressure distribution, contact area and contact torque; S2: Filter the tactile data collected by S1 to remove noise and interference; S3: extract relevant tactile features from the data processed by S2, including the mean and variance of the pressure distribution and the direction and magnitude of the contact torque; S4: Compare the tactile features extracted by S3 with the preset standard tactile state and calculate the posture error vector; S5: Based on the posture error vector calculated by S4, the PID control algorithm is used to generate adjustment instructions to smoothly respond to the error change; S6: Execute the adjustment command generated by S5 to adjust the position and posture of the robot end effector in real time, while continuously monitoring the feedback data from the tactile sensor, and input the new tactile data into S2 to form a control closed loop.

2. A robot operation posture control method based on a tactile sensor according to claim 1, characterized in that: The S1 specifically includes: S11: Install high-resolution multi-point tactile sensors on the contact part of the robot end effector to ensure full coverage of the contact area with the target object; S12: measuring and recording the pressure distribution of the contact area in real time through each sensing point in the tactile sensor array; S13: Using the contact area detection function of the tactile sensor, the contact area between the robot end effector and the target object is calculated in real time. The calculation formula is: ,in, is the contact area, For the The pressure value measured at each sensing point, For the The area weight of each sensor point, is the total number of sensing points in the tactile sensor array; S14: The direction and magnitude of the contact torque are monitored and recorded in real time through the built-in torque measurement function of the tactile sensor.

3. The robot operation posture control method based on tactile sensor according to claim 1 is characterized in that: The S2 specifically includes: S21: De-noising the tactile data collected by S1, using a median filter algorithm, selecting a neighborhood range of the tactile data and replacing each data point with a median value within the range, thereby removing outliers caused by sensor noise; S22: Smoothing the denoised tactile data, and performing weighted averaging on the data using a Gaussian filter algorithm to eliminate low-frequency fluctuations caused by insufficient sensor sampling frequency or environmental interference; S23: Segment the tactile data processed by S21 and S22, and segment them according to the set time window based on the change characteristics of the data, so as to ensure that each segment of data has stable characteristics in the time domain; S24: high-frequency noise is removed from each section of processed tactile data by using a bandpass filter to remove noise below a set frequency and above a set frequency, thereby ensuring that the remaining data conforms to the frequency spectrum characteristics of the target signal.

4. The robot operation posture control method based on tactile sensor according to claim 1 is characterized in that: The S3 specifically includes: S31: Perform statistical analysis on the pressure distribution values ​​in the tactile data processed by S2, and calculate the arithmetic mean of the pressure value of each sensing point. The formula is: ,in, is the mean value of the pressure distribution, For the The pressure value measured at each sensing point, is the total number of sensing points in the tactile sensor array; S32: Based on the mean value of the pressure distribution calculated in S31, the variance of the pressure distribution is calculated, and the formula is: ,in, is the variance of the pressure distribution; S33: Calculate the direction of the contact torque using the torque component in the tactile data processed by S2, and determine its specific direction in two-dimensional or three-dimensional space by calculating the angle of the torque vector. The formula is: ,in is the direction angle of the contact torque, and are the components of the moment on the X-axis and Y-axis respectively; S34: According to the torque component in the tactile data processed by S2, the magnitude of the contact torque is calculated using the vector modulus calculation method, and the formula is: ,in, is the magnitude of the contact torque, , , are the components of the torque on the X-axis, Y-axis and Z-axis respectively.

5. The robot operation posture control method based on tactile sensor according to claim 4 is characterized in that: The S4 specifically includes: S41: Compare the pressure distribution mean, pressure distribution variance, contact torque direction and magnitude extracted in S3 with the preset standard tactile features one by one, and calculate the difference value between each feature; set the standard values ​​of pressure distribution mean, pressure distribution variance, contact torque direction and contact torque magnitude to be and , then the difference value comparison formula of each feature is: ; ; ; ,in, and They are the difference values ​​of the mean of pressure distribution, pressure distribution variance, contact moment direction and contact moment magnitude; S42: According to the importance of each tactile feature to posture control, they are and Assign weight coefficient and ; S43: Combine the weighted difference values ​​into vectors to form a posture error vector , and its calculation formula is: ,in, and They are the unit vector directions corresponding to the four features respectively.

6. The robot operation posture control method based on tactile sensor according to claim 1 is characterized in that: The S5 specifically includes: S51: according to the posture error vector calculated in S4, decompose it into two parts: position error and posture error; S52: Set the proportional coefficient of the PID controller , integral coefficient and the differential coefficient ; S53: Calculate the proportional control amount according to the error vector decomposed in step S51. The calculation formula of the proportional control amount is: ,in, is the proportional control quantity, is the position error component, is the attitude error component; S54: The integration result of position error and attitude error is combined with the integration coefficient Multiply them together to get the integral control amount. The calculation formula of the integral control amount is: ,in, is the integral control quantity, and are the functions of position error and attitude error changing with time respectively; S55: The rate of error change and the differential coefficient Multiply them to get the differential control amount. The calculation formula of the differential control amount is: in, is the differential control quantity, is the rate of change of position error and attitude error over time; S56: Add the proportional control amount, integral control amount and differential control amount calculated in S53 to S55 to obtain the final adjustment instruction.

7. The robot operation posture control method based on tactile sensor according to claim 6 is characterized in that: The S51 specifically includes: S511: According to the pose error vector calculated in S4 , decompose it into two parts: position error and attitude error; S512: By analyzing the first three components of the posture error vector calculated in S4, the position error is obtained. , the error vector represents the deviation between the current position of the robot end effector and the target position; the expression of the position error is: ,in, and They represent the robot end effector Position error in direction; S513: By analyzing the last three components of the posture error vector calculated in S4, the posture error is obtained. , the error vector represents the deviation between the posture of the robot end effector and the target posture; the expression of posture error is: ,in, , and They represent the pitch angle error, yaw angle error and roll angle error of the robot end effector respectively.

8. The robot operation posture control method based on tactile sensor according to claim 7 is characterized in that: The S56 specifically includes: S561: converting the proportional control amount calculated in S53 into , the integral control amount calculated in S54 And the differential control amount calculated in S55 Summarize and generate final adjustment instructions ; The specific calculation formula is: ,in, For the final adjustment instructions, They are proportional, integral and differential control quantities respectively.

9. The robot operation posture control method based on tactile sensor according to claim 1 is characterized in that: The S6 specifically includes: S61: receiving the adjustment instruction generated in step S5 , and convert it into a specific robot end effector posture adjustment operation; adjustment instruction It includes position and attitude corrections, which are applied to each degree of freedom of the robot's end effector; S62: According to the adjustment instructions converted in S61, the movement of the robot end effector is controlled in real time, and the position and posture are dynamically adjusted by driving the motor; S63: while executing S62 to adjust the posture, continuously collect real-time tactile data of the contact area between the robot end effector and the target object through a high-resolution multi-point tactile sensor; S64: The new tactile data collected in S63 is input into step S2 in real time for filtering processing to form a closed-loop control.

10. A robot operation posture control system based on a tactile sensor, used to implement a robot operation posture control method based on a tactile sensor as claimed in any one of claims 1 to 9, characterized in that: Includes the following modules: Acquisition module: used to collect tactile data of the contact area between the robot end effector and the target object, the tactile data including pressure distribution, contact area and contact torque; Data processing module: used to filter the tactile data collected by the acquisition module to remove noise and interference; Feature extraction module: used to extract tactile features from the data processed by the data processing module, including the mean and variance of pressure distribution and the direction and magnitude of contact torque; A posture error calculation module: used to compare the tactile features extracted by the feature extraction module with the preset standard tactile state and calculate the posture error vector; Control module: used to generate adjustment instructions using PID control algorithm according to the posture error vector calculated by the posture error calculation module, and adjust the posture of the robot end effector in real time according to the adjustment instructions; Feedback monitoring module: used to continuously monitor the feedback data of the tactile sensor and input the new tactile data into the data processing module for reprocessing to form a control closed loop.

Images