Control method and device of electric screw driver, electric screw driver and computer readable storage medium

Abstract

The invention discloses a control method and device of an electric screw driver, the electric screw driver and a computer readable storage medium. The control method comprises the steps that operation data of the electric screwdriver are obtained, and the operation data comprise torque data, angular velocity data and acceleration data; and respectively carrying out feature extraction on the torque data, the angular velocity data and the acceleration data to obtain a torque feature value, an angular velocity feature value and an acceleration feature value. And according to the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value and a preset torque learning model, the working state of the electric screw driver is determined. And according to the working state, operation parameters of the electric screwdriver are adjusted, so that the torque precision of the electric screwdriver is within a preset precision range. Through the arrangement, the torque of the electric screw driver can be accurately controlled according to the operation data of the electric screw driver, the stability and reliability in the torque control process are effectively improved, and the use requirements for the torque precision of the electric screw driver in different scenes are met.

Description

Technical Field

[0001] This invention relates to the field of electric screwdriver control technology, and in particular to a control method, device, electric screwdriver, and computer-readable storage medium for an electric screwdriver. Background Technology

[0002] With the rapid development of precision assembly industries such as miniaturization of electronic devices, integration of automotive parts, and high precision of aerospace components, the precision requirements for assembly processes in the assembly of precision parts and products are constantly increasing. Among them, electric screwdrivers, as core assembly tools, directly determine the stability of component connections, the reliability of product performance, and even the safety of use by the accuracy of their torque control. Therefore, the demand for precise torque control of electric screwdrivers is becoming increasingly prominent.

[0003] Existing torque control technologies for electric screwdrivers mainly rely on two methods: detecting motor current or using a mechanical torque clutch. Both methods suffer from insufficient torque accuracy.

[0004] Therefore, improving the torque accuracy of electric screwdrivers has become an urgent problem to be solved. Summary of the Invention

[0005] To address the aforementioned issues, embodiments of the present invention provide a control method, device, electric screwdriver, and computer-readable storage medium for an electric screwdriver. This method enables precise control of the electric screwdriver's torque based on its operating data, effectively improving the stability and reliability of the torque control process and meeting the torque accuracy requirements of electric screwdrivers in various scenarios.

[0006] According to an embodiment of the present invention, a method for controlling an electric screwdriver is provided, the method comprising:

[0007] The operating data of the electric screwdriver is acquired, wherein the operating data includes torque data, angular velocity data, and acceleration data;

[0008] Feature extraction is performed on the torque data, the angular velocity data, and the acceleration data respectively to obtain torque feature values, angular velocity feature values, and acceleration feature values;

[0009] The working state of the electric screwdriver is determined based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value, and the preset torque learning model.

[0010] Based on the operating state, adjust the operating parameters of the electric screwdriver to ensure that the torque accuracy of the electric screwdriver is within a preset accuracy range.

[0011] In one alternative approach, the torque characteristic values ​​include the mean torque, the torque variance, and the instantaneous rate of change of torque; the angular velocity characteristic values ​​include the peak angular velocity and the steady-state angular velocity; and the acceleration characteristic values ​​include the mean magnitude of the acceleration vector and the rate of change of acceleration.

[0012] In one alternative approach, the operating state includes an under-tightening state, and adjusting the operating parameters of the electric screwdriver according to the operating state includes:

[0013] If the working state is the under-tightening state, the current power supply current of the electric screwdriver is adjusted to the first power supply current, wherein the current value of the first power supply current is greater than the current value of the current power supply current.

[0014] In one optional embodiment, the working state includes an overtightening state and a disengaged state, and adjusting the operating parameters of the electric screwdriver according to the working state includes:

[0015] If the operating state is the over-tightening state or the unhooking state, then the current power supply to the electric screwdriver is cut off.

[0016] In an alternative embodiment, after adjusting the current supply current of the electric screwdriver to a first supply current if the operating state is the undertightening state, the method further includes:

[0017] Obtain the temperature data of the electric screwdriver;

[0018] If the temperature data exceeds the preset temperature range, the first power supply current of the electric screwdriver is adjusted to the second power supply current, wherein the current value of the second power supply current is greater than the current value of the first power supply current.

[0019] In an optional embodiment, before determining the working state of the electric screwdriver based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value, and a preset torque learning model, the method further includes:

[0020] Under multiple preset standard torque values, multiple reference torque data, multiple reference angular velocity data, and multiple reference acceleration data of the electric screwdriver are obtained;

[0021] Feature extraction is performed on multiple reference torque data, multiple reference angular velocity data, and multiple reference acceleration data respectively to obtain multiple reference torque feature values, multiple reference angular velocity feature values, and multiple reference acceleration feature values;

[0022] The general model is trained based on multiple reference torque feature values, multiple reference angular velocity feature values, and multiple reference acceleration feature values ​​to obtain a torque learning model.

[0023] In one alternative approach, after obtaining the torque learning model, the method further includes:

[0024] When the number of operations of the electric screwdriver exceeds the preset number of operations, the calibration torque is obtained;

[0025] The torque learning model is updated based on the calibration torque to obtain the calibrated torque learning model.

[0026] According to a second aspect of the present invention, a control device for an electric screwdriver is provided, the device comprising:

[0027] The acquisition module is used to acquire the operating data of the electric screwdriver, wherein the operating data includes torque data, angular velocity data and acceleration data;

[0028] The feature extraction module is used to extract features from the torque data, the angular velocity data and the acceleration data respectively to obtain torque feature values, angular velocity feature values ​​and acceleration feature values;

[0029] The determination module is used to determine the working state of the electric screwdriver based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value and the preset torque learning model;

[0030] The control module is used to adjust the operating parameters of the electric screwdriver according to the working state, so that the torque accuracy of the electric screwdriver is within a preset accuracy range.

[0031] According to a third aspect of the present invention, an electric screwdriver is provided, the electric screwdriver including the control method of the electric screwdriver described in any of the above aspects or any optional methods.

[0032] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, characterized in that the computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the control method of the electric screwdriver as described in any one of claims 1 to 7.

[0033] The beneficial effects of this application are as follows: This application acquires torque, angular velocity, and acceleration data of an electric screwdriver, extracts features from each type of data to obtain corresponding feature values, and then combines this with a preset torque learning model to determine the working state of the electric screwdriver. Based on this working state, the operating parameters are adjusted to ensure that the torque accuracy remains within a preset range. This enables precise control of the electric screwdriver's torque, effectively improving the stability and reliability of the torque control process and meeting the torque accuracy requirements of electric screwdrivers in different scenarios. Simultaneously, through dynamic adjustment of operating parameters, it ensures that the electric screwdriver always operates in an appropriate state during work, reducing adverse situations caused by torque control deviations, improving the overall working performance and applicability of the electric screwdriver, and providing strong support for the efficient application of electric screwdrivers in various assembly operations. Attached Figure Description

[0034] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0035] Figure 1 This is a flowchart illustrating the steps of a control method for an electric screwdriver provided in this application;

[0036] Figure 2 This is a schematic diagram of the training steps of the torque learning model provided in this application;

[0037] Figure 3 This is a schematic diagram of the structure of a control device for an electric screwdriver provided by the present invention. Detailed Implementation

[0038] To enable those skilled in the art to better understand the technical solutions of this application, the application will be described in detail below with reference to the accompanying drawings and specific embodiments. Although the accompanying drawings and specific embodiments describe exemplary embodiments of the present invention, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0039] The terms "second" and similar words used in this application do not indicate any order, quantity, or importance, but are merely used for distinction. The terms "including" and similar words used in this application mean that the element preceding the word encompasses the elements listed after the word, and do not exclude the possibility of including other elements. The technical solutions of this application are not limited to the execution order described in the embodiments. The steps in the execution order can be combined, decomposed, or their order can be changed, as long as the logical relationship of the execution content is not affected.

[0040] All terms used in this application (including technical or scientific terms) have the same meaning as understood by one of ordinary skill in the art to which this application pertains, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant art, and not as idealized or highly formalized, unless expressly defined herein. Technologies and equipment known to one of ordinary skill in the art may not be discussed in detail, but where appropriate, such technologies and equipment should be considered part of the specification.

[0041] With the rapid development of precision assembly industries such as miniaturization of electronic devices, integration of automotive parts, and high precision of aerospace components, the precision requirements for assembly processes in the assembly of precision parts and products are constantly increasing. Among them, electric screwdrivers, as core assembly tools, directly determine the stability of component connections, the reliability of product performance, and even the safety of use by the accuracy of their torque control. Therefore, the demand for precise torque control of electric screwdrivers is becoming increasingly prominent.

[0042] Existing torque control technologies for electric screwdrivers mainly rely on two methods: detecting motor current or using a mechanical torque clutch. Both methods suffer from insufficient torque accuracy.

[0043] To address the aforementioned problems, this application provides a control method for an electric screwdriver, which can precisely control the torque of the electric screwdriver based on its operating data, effectively improving the stability and reliability of the torque control process and meeting the torque accuracy requirements of electric screwdrivers in different scenarios. Figure 1 As shown, Figure 1 This is a schematic diagram of the overall steps of the method described in this application. The method may include the following steps S110-S140.

[0044] S110: Obtain the operating data of the electric screwdriver, wherein the operating data includes torque data, angular velocity data and acceleration data.

[0045] In this embodiment, the operating data can be physical quantity data that is directly related to the torque control accuracy and is collected in real time during the tightening operation of the electric screwdriver.

[0046] It should be noted that in some other embodiments, the operating data may also include temperature data, vibration data, and other data, which are not specifically limited in this application embodiment.

[0047] Specifically, torque data can be acquired by embedding a high-precision torque sensor between the motor output shaft of the electric screwdriver and the bit adapter sleeve, directly collecting the real-time torque physical quantity output by the motor to the bit, avoiding the deviation of indirectly estimating torque through current.

[0048] Angular velocity data can be acquired by fixing a six-axis inertial measurement unit (IMU) sensor to the inner wall of the screwdriver bit housing. Its built-in gyroscope module collects the angular velocity data of the bit rotating around the central axis in real time, reflecting the stability of the bit rotation (such as whether there is jamming, free spin, etc.).

[0049] Acceleration data can also be acquired through the accelerometer module of the aforementioned six-axis IMU sensor, which collects linear acceleration data of the screwdriver bit in the X, Y, and Z axes to capture load impacts during screw tightening (such as instantaneous acceleration changes caused by screw jamming).

[0050] The collected torque, angular velocity, and acceleration data are transmitted in real time to the main control chip of the electric screwdriver via the Serial Peripheral Interface (SPI) communication interface (for example, forming a set of synchronized three-dimensional running data frames every 1ms), and are temporarily stored in the on-chip random access memory (RAM) of the main control chip.

[0051] S120: Perform feature extraction on the torque data, the angular velocity data, and the acceleration data respectively to obtain torque feature values, angular velocity feature values, and acceleration feature values.

[0052] In this embodiment, feature extraction can be carried out by filtering and calculating parameters from the original collected operating data that can reflect the essential laws of the data and are strongly correlated with the working state of the electric screwdriver (such as normal tightening, under-tightening, over-tightening, etc.) and eliminating noise interference in the operating data (such as instantaneous abnormal values ​​caused by electromagnetic interference).

[0053] Specifically, torque feature extraction involves calculating three core torque feature values—mean torque, torque variance, and instantaneous torque change rate—from multiple continuously collected torque data frames (e.g., 10, 20, or 50 frames). The mean torque is the arithmetic mean of the multiple torque data points, reflecting the average torque level within that time period. The instantaneous torque change rate is the ratio of the torque difference between two adjacent data frames to the time interval, reflecting the rate of torque change and used to determine if there is an over-tightening trend. For example, if the torque in frame n is 1.02 N·m and in frame n+1 it is 1.05 N·m, with a time interval of 1 ms, then the change rate is 30 N·m / s. The torque variance is the average of the sum of squared deviations of each torque data point from the mean torque within a certain time window, used to quantify the fluctuation of torque output. A smaller variance indicates more stable torque output; a sudden increase in variance or exceeding a threshold often indicates a hidden anomaly in the tightening process (non-instantaneous extreme value anomalies), which are difficult to detect using only the mean torque and instantaneous torque change rate. For example, in the scenario of tightening bolts in a new energy vehicle battery pack, if the average torque is stable at the target torque of 10 N·m (within the "normal range"), but the torque variance is higher than the usual 0.01 N·m... 2 The temperature rose sharply to 0.1 N·m. 2 This could be due to slight stripping of the screwdriver bit at the screw contact surface (not severe enough to cause disengagement) or minor jamming of the motor output shaft. In such cases, relying solely on the mean torque is insufficient to identify this hidden risk, while torque variance can serve as a key early warning indicator. For example, when tightening motherboard screws in electronic devices (target torque 0.5 N·m, preset accuracy ±0.3%), if the mean torque is stable at 0.501 N·m, but the variance consistently exceeds 0.0005 N·m², it indicates high-frequency fluctuations in torque output. This could lead to uneven force on the screws, potentially causing motherboard deformation over time. In this case, torque variance can help determine whether a seemingly normal tightening process harbors potential accuracy risks.

[0054] Angular velocity feature extraction can be performed on angular velocity data within the same time window, extracting two core angular velocity features: the peak angular velocity and the trough angular velocity. The peak angular velocity is the maximum value among multiple angular velocity data points, reflecting the highest rotational speed of the bit, typically occurring in the initial stage of tightening. The trough angular velocity is the minimum value among multiple angular velocity data points, reflecting the lowest rotational speed of the bit, typically occurring near the end of tightening.

[0055] It should be noted that in some other embodiments, the angular velocity feature value extraction can also extract the steady-state angular velocity, that is, the average value of the remaining angular velocity data after removing the peak and valley values ​​of the angular velocity, which is used to reflect the speed when the bit rotates stably.

[0056] Acceleration feature extraction can be performed on raw acceleration data along the X, Y, and Z axes. First, the acceleration vector magnitude of each frame is calculated to eliminate axial differences. Then, based on a time window, two core acceleration feature values ​​are extracted: the mean acceleration vector magnitude and the attitude angle deviation. The mean acceleration vector magnitude reflects the overall vibration level of the bit. The rate of change of acceleration, the ratio of the difference in vector magnitude between adjacent frames to the time interval, is used to identify sudden conditions such as decoupling; for example, during decoupling, the rate of change can suddenly increase from 0.5 g / ms to 5 g / ms.

[0057] It should be noted that in some other embodiments, the acceleration feature value extraction can also extract the attitude angle deviation. The attitude angle deviation refers to the difference between the current attitude angle (such as pitch angle and roll angle) of the bit, calculated by the six-axis IMU sensor by fusing angular velocity and acceleration data, and the initial calibration attitude angle. It reflects whether the bit is tilted (tilt will cause torque component error). For example, if the initial calibration attitude angle is 0° (perpendicular to the screw surface), and the currently calculated attitude angle is 2°, then the attitude angle deviation is 2°.

[0058] S130: Determine the working state of the electric screwdriver based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value, and the preset torque learning model.

[0059] In this embodiment, the preset torque learning model can be a data model pre-trained based on machine learning algorithms and stored in the main control chip of the electric screwdriver. Its core function is to establish a mapping relationship between multiple feature values ​​and the working state of the electric screwdriver. The torque learning model can be constructed using gradient boosting tree regression algorithms (such as the optimized distributed gradient boosting library algorithm (Extreme Gradient Boosting, XGBoost)), random forests, support vector machines, decision trees, and other model building algorithms. For example, this embodiment uses the gradient boosting tree regression algorithm as an example. Calibration data under a standard experimental environment (temperature 25℃±1℃, vibration ≤0.02mm) is used as training data. A standard torque calibrator (accuracy ±0.05%) can be used to set 8 target torques (0.1 N·m, 0.5 N·m, 1 N·m, 5 N·m, 10 N·m, 15 N·m, 20 N·m, 30 N·m). For each target torque, multiple feature values ​​are collected from 1000 tightening processes, and the corresponding actual working states are labeled (e.g., under a target torque of 1 N·m, feature vector A corresponds to the normal tightening state, and feature vector B corresponds to the over-tightening state). A total of 8000 training samples are obtained. After training, the accuracy of the model's state judgment is greatly improved (e.g., greater than or equal to 99.5%).

[0060] The working state can refer to the specific working conditions of the electric screwdriver during tightening operations, which can include four typical working conditions: normal tightening, under-tightening, over-tightening, and unhooking.

[0061] Specifically, the determination of the working state can be achieved by the main control chip inputting multiple feature values ​​obtained in step S120 into the aforementioned preset torque learning model. The torque learning model calculates the matching degree between the feature vector and the feature template of each typical working condition, and outputs the corresponding working state. For example, when the average torque value in the torque feature value reaches 95% to 105% of the target torque, the steady-state angular velocity in the angular velocity feature value is stable at 500° / s ± 50° / s, and the rate of change of acceleration in the acceleration feature value is ≤ 1g / ms, the model determines it as "normal tightening state". When the average torque value in the torque feature value is only 60% to 80% of the target torque, and the angular velocity feature value is continuously high (e.g., greater than 800° / s), the model determines it as "under-tightening state".

[0062] It should be noted that the feature values ​​of the input torque learning model can be the mean torque, steady-state angular velocity and acceleration change rate in the above embodiments, or at least one of the mean torque, torque variance and instantaneous torque change rate, and at least one of the steady-state angular velocity, peak angular velocity and valley angular velocity, and at least one of the acceleration change rate, mean acceleration vector magnitude and attitude angle deviation. The embodiments of this application do not specifically limit the features.

[0063] S140: Adjust the operating parameters of the electric screwdriver according to the working state so that the torque accuracy of the electric screwdriver is within the preset accuracy range.

[0064] In this embodiment, the operating parameters can be core control parameters that directly affect the torque output of the electric screwdriver, specifically the motor drive current, motor speed, and Pulse Width Modulation (PWM) signal duty cycle. The preset accuracy range is the allowable torque error range pre-set according to the application scenario of the electric screwdriver. The range varies in different scenarios (for example, the preset accuracy range is ±0.3% for electronic device motherboard assembly and ±0.5% for automotive chassis component assembly). It is set and stored by the user through the electric screwdriver's operation panel or the accompanying application (APP).

[0065] For example, if the working state is normal tightening, the main control chip keeps the current operating parameters stable. For instance, the motor drive current is maintained at 1.0A, the PWM duty cycle is maintained at 50%, and the motor speed is maintained at 600 r / min. By only fine-tuning the PWM duty cycle (each adjustment increment ±1%), the average torque is kept stable between 98% and 102% of the target torque, ensuring that the torque accuracy is within the preset range.

[0066] If the screw is under-tightened, the main control chip increases the motor drive current in increments of 0.05A per cycle (maximum increase to 1.5A to prevent excessive current from burning out the motor), while simultaneously monitoring the rate of change in the acceleration characteristic value. If the rate of change is less than or equal to 2g / ms, the current continues to increase. If the rate of change is greater than 2g / ms, it indicates screw jamming, pauses current adjustment, and triggers a soft buzzer alarm.

[0067] If the working state is over-tightening, that is, the average torque exceeds 105% of the target torque, the main control chip promptly sends a high-level signal to the relay in the lithium battery power supply circuit to cut off the motor power output. At the same time, it records data such as the torque peak and feature vector at the moment of over-tightening to ensure that the torque accuracy no longer deviates from the preset range.

[0068] If the working state is unhooked (the rate of change of the acceleration characteristic value is greater than 5g / ms and the angular velocity characteristic value drops sharply by more than 80%), the main control chip immediately cuts off the motor power output and triggers an alarm combination of a continuous buzzer and a flashing red LED to prevent the bit from spinning and wearing out. The operation can be resumed after the user re-clamps the screw.

[0069] Through the above adjustments, the torque accuracy of the electric screwdriver can be stably maintained within the preset accuracy range, meeting the needs of different precision assembly scenarios.

[0070] This invention acquires torque, angular velocity, and acceleration data from an electric screwdriver, extracts features from each data type to obtain corresponding feature values, and then combines this with a preset torque learning model to determine the working state of the electric screwdriver. Based on this working state, the operating parameters are adjusted to ensure torque accuracy remains within a preset range. This enables precise torque control of the electric screwdriver, effectively improving the stability and reliability of the torque control process and meeting the torque accuracy requirements of electric screwdrivers in various scenarios. Furthermore, by dynamically adjusting the operating parameters, it ensures that the electric screwdriver always operates in an appropriate state during work, reducing adverse situations caused by torque control deviations, improving the overall working performance and applicability of the electric screwdriver, and providing strong support for the efficient application of electric screwdrivers in various assembly operations.

[0071] In some embodiments, the torque characteristic values ​​include the mean torque, the torque variance, and the instantaneous rate of change of torque; the angular velocity characteristic values ​​include the peak angular velocity and the steady-state angular velocity; and the acceleration characteristic values ​​include the mean magnitude of the acceleration vector and the rate of change of acceleration.

[0072] In this embodiment of the application, the determination of the above working state is based on the collaborative operation of the above 7-dimensional feature values ​​(3 torque feature values, 2 angular velocity feature values, and 2 acceleration feature values) and the preset torque learning model. The accurate determination is achieved by matching the feature values ​​with the quantitative feature templates of typical working conditions.

[0073] Specifically, the preset torque learning model can be a "feature value-working state" mapping model trained using a random forest algorithm (120 decision trees, maximum depth 8). Its core function is to establish a quantitative correlation between 7-dimensional feature values ​​(mean torque, torque variance, instantaneous rate of change of torque, peak angular velocity, steady-state angular velocity, mean magnitude of acceleration vector, and rate of change of acceleration) and four types of working states (normal tightening, undertightening, overtightening, and unhooking). The model analyzes the combination patterns of feature values ​​and outputs the most matching working state. The decision logic is based on judgment rules formed from a large amount of calibration data learned during the training phase.

[0074] For example, the logic for determining the four types of working states is that the model comprehensively determines the working state by comparing the real-time 7-dimensional feature values ​​with preset typical working condition feature thresholds.

[0075] When the 7-dimensional feature values ​​simultaneously meet the following conditions, it is determined to be in a normal tightening state:

[0076] The average torque is between 98% and 102% of the target torque (e.g., when the target torque is 1 N·m, the average is between 0.98 and 1.02 N·m).

[0077] The torque variance is less than or equal to 0.00002 N·m 2 (This reflects stable torque output without drastic fluctuations);

[0078] The instantaneous rate of change of torque is less than or equal to 30 N·m / s (the torque change is gradual, without sudden increases / decreases);

[0079] The peak angular velocity is less than or equal to 800° / s, and the steady-state angular velocity is between 500 and 700° / s (the bit rotation speed is stable, without idling or jamming);

[0080] The average magnitude of the acceleration vector is less than or equal to 1g (no obvious vibration or impact on the bit);

[0081] Rate of change of acceleration: less than or equal to 500g / s (no sudden change in acceleration, and uniform force on the screw).

[0082] When the 7-dimensional feature values ​​meet the following core conditions, it is determined to be in an under-tightening state:

[0083] The average torque is less than 90% of the target torque (e.g., when the target torque is 1 N·m, the average torque is less than 0.9 N·m).

[0084] Peak angular velocity greater than 850° / s (the bit is rotating too fast and has not reached the effective tightening load);

[0085] Steady-state angular velocity greater than 750° / s (continuous high-speed rotation without deceleration due to increased load);

[0086] The rate of change of acceleration is less than 200g / s (no obvious load impact, screw not fully attached to workpiece).

[0087] When the 7-dimensional feature values ​​meet the following core conditions, it is determined to be in an over-twisted state:

[0088] The average torque is greater than 105% of the target torque (e.g., when the target torque is 1 N·m, the average torque is greater than 1.05 N·m).

[0089] The instantaneous rate of change of torque is greater than 50 N·m / s (torque increases rapidly, with a tendency to overtighten);

[0090] The torque variance is greater than 0.00003 N·m 2 (Unstable torque output due to load fluctuations caused by over-tightening);

[0091] The steady-state angular velocity is less than 400° / s (due to excessive load, the bit speed is forced to decrease).

[0092] When the 7-dimensional feature values ​​meet the following sudden conditions, it is determined to be in a decoupling state:

[0093] Peak angular velocity: A sudden drop of more than 80% (e.g., from 800° / s to below 150° / s within 10ms, the bit suddenly stalls).

[0094] Rate of change of acceleration: greater than 2000g / s (a violent impact occurs the moment the bit detaches from the screw);

[0095] Average torque: A sudden drop of more than 50% (the load suddenly disappears, and the torque drops rapidly).

[0096] The torque learning model's decision-making process involves the main control chip inputting real-time collected 7-dimensional feature values ​​into a pre-defined torque learning model. The model then uses 120 decision trees to perform parallel judgments on these feature values. Each decision tree outputs an independent state judgment result based on different feature combinations (such as "mean torque + steady-state angular velocity", "rate of change of acceleration + instantaneous rate of change of torque", etc.). The current operating state is ultimately determined through a "voting method" (the state supported by the majority of decision trees is the final result). For example, if 95 out of the 120 decision trees determine an under-tightening state, the model will ultimately output the under-tightening state, ensuring the robustness and reliability of the judgment results.

[0097] By using the above-mentioned 7-dimensional multi-feature fusion and model decision-making method, various working conditions of the electric screwdriver during the tightening process can be accurately identified, providing a clear basis for the dynamic adjustment of subsequent operating parameters and ensuring that the torque control accuracy is always within the preset range.

[0098] In some embodiments, the working state includes an under-tightening state, and adjusting the operating parameters of the electric screwdriver according to the working state includes: if the working state is the under-tightening state, adjusting the current power supply current of the electric screwdriver to a first power supply current, wherein the current value of the first power supply current is greater than the current value of the current power supply current.

[0099] In this embodiment, the current supply current is the real-time current value of the motor driven by the electric screwdriver before it is determined to be under-tightened (collected in real time by the main control chip through a current sensor). The first supply current can be a target current value set to eliminate the under-tightening state, which is greater than the current supply current. Its adjustment is in a step-by-step manner and a safety threshold is set to ensure that motor overload is avoided while increasing torque output.

[0100] Specifically, the initial current adjustment is achieved by the main control chip reading the current supply current value (e.g., 1.0A) and calculating the initial value of the first supply current: First supply current = current supply current + 0.1A (the increment can be preset through the parameter configuration interface, and the range can be from 0.05A to 0.2A; in this embodiment, 0.1A is selected), that is, the initial first supply current is 1.1A.

[0101] Next, a step-by-step incremental mechanism is implemented. After adjusting to the initial first supply current, the main control chip re-collects torque characteristic values ​​at preset time intervals (e.g., every 5ms), focusing on monitoring the average torque. If the average torque is still less than 95% of the target torque, the judgment point is not improved, and the first supply current is increased by 0.1A per increment (e.g., 1.2A, 1.3A, 1.4A, etc.). If the average torque reaches 95% to 100% of the target torque, it is determined that the improvement has been achieved, and the current first supply current is maintained.

[0102] It should be noted that a safety threshold limit needs to be set for the first power supply current. Set a maximum threshold for the first power supply current (e.g., 1.5 times the motor's rated current; if the motor's rated current is 2A, then the maximum threshold is 3A). When the first power supply current reaches the maximum threshold, even if the under-tightening condition has not been completely eliminated, the current increase will stop to prevent the motor from overheating and burning out. Simultaneously, a buzzer alarm will be triggered (e.g., sounding at 1-second intervals) to prompt the user to check the condition of the screws or bits (e.g., whether the threads are stripped, whether the bit type is compatible).

[0103] This solution directly increases the motor's output torque by specifically increasing the supply current when the screws are under-tightened, quickly eliminating the under-tightening phenomenon and ensuring the screw connection strength meets standards. The stepped, incremental current adjustment method avoids the risk of overshoot caused by sudden torque increases. Combined with safety threshold limits, it ensures the effectiveness of under-tightening correction while protecting the motor and workpiece from damage. Simultaneously, the adjustment process is linked to real-time monitoring of torque characteristic values, dynamically adapting to the tightening resistance of different screws. This enhances the electric screwdriver's adaptability under complex working conditions, further ensuring the accuracy of torque control and operational efficiency.

[0104] In some embodiments, the working state includes an overtightening state and a disengaged state, and adjusting the operating parameters of the electric screwdriver according to the working state includes: if the working state is the overtightening state or the disengaged state, then cutting off the current power supply current of the electric screwdriver.

[0105] In this embodiment, the current power supply current can be the real-time current driving the motor before the electric screwdriver is determined to be in an over-tightening or disengaged state. Cutting off the current power supply current can be achieved by controlling a relay in the power supply circuit to disconnect via the main control chip, causing the motor to momentarily lose power input and stop torque output.

[0106] Specifically, when an overtightening condition is detected, the main control chip immediately sends a high-level control signal to the relay, triggering the relay contacts to open and cutting off the power supply circuit between the lithium battery and the motor. This ensures a short response time from state detection to current cutoff (e.g., less than or equal to 5ms, including signal transmission delay less than or equal to 2ms and relay action delay less than or equal to 3ms). After the current is cut off, the main control chip continuously monitors the torque decay curve through a torque sensor until the torque drops to 0 N·m. It records data such as the peak overtightening torque (e.g., 5.3 N·m) and the duration of overtightening (e.g., 8ms), storing these data in the memory for subsequent quality traceability. Simultaneously, an alarm is triggered, causing the red LED indicator to remain constantly lit and the buzzer to sound continuously at a 2kHz frequency for 2 seconds, alerting the user to the occurrence of overtightening.

[0107] When the bit is detected as disengaged, the main control chip sends a high-level control signal to the relay, cutting off the power supply within a short time (e.g., less than or equal to 5ms) to prevent the bit from running idle and causing wear or scratches on the workpiece. After cutting off the current, the motor starting circuit is locked (the user needs to press the "reset" button to unlock it) to prevent accidental restart. An alarm is triggered, causing the yellow LED indicator to flash and the buzzer to sound intermittently at a fixed frequency, prompting the user to address the disengagement issue (e.g., re-clamping the bit or screw).

[0108] This embodiment, by promptly cutting off the power supply current in over-tightening or unhooking states, can quickly terminate the motor's torque output, fundamentally avoiding workpiece damage caused by over-tightening (such as bolt breakage or housing deformation) or equipment wear caused by unhooking (such as bit chipping), significantly improving operational safety. The short response time ensures intervention and protection in the early stages of abnormal conditions, minimizing losses. Simultaneously, combined with an alarm mechanism and data logging function, it can promptly alert users to handle abnormalities and provide data support for subsequent process optimization, further enhancing the reliability and controllability of the electric screwdriver in precision assembly scenarios.

[0109] In some embodiments, after adjusting the current power supply current of the electric screwdriver to a first power supply current if the working state is the undertightening state, the method further includes: acquiring temperature data of the electric screwdriver; if the temperature data exceeds a preset temperature range, adjusting the first power supply current of the electric screwdriver to a second power supply current, wherein the current value of the second power supply current is greater than the current value of the first power supply current.

[0110] In this embodiment, the temperature data can be the real-time operating temperature of the electric screwdriver motor windings (which directly affects the motor's output torque characteristics). This data is collected by a digital temperature sensor mounted close to the motor housing, with the sampling frequency synchronized with the current adjustment cycle (e.g., sampling once every 10ms). The data is transmitted to the main control chip via a communication protocol to ensure timely temperature monitoring. In this embodiment, the preset temperature range can be a stable torque output range set based on the motor's rated operating parameters and experimental data, for example, set to 25℃ to 60℃ (if the temperature exceeds 60℃, the motor winding resistance increases, and the torque output decreases by approximately 5% to 8% under the same current). The second supply current can be a target current value, greater than the first supply current, set to compensate for torque loss at high temperatures. Its adjustment logic can be based on the principle that the degree of temperature exceedance is positively correlated with the degree of current compensation.

[0111] The specific adjustment methods may be as follows:

[0112] First, a preset temperature threshold is established. When the temperature data is less than or equal to 60℃, it is determined that "the preset range has not been exceeded," and the first power supply current remains unchanged; when the temperature data is greater than 60℃, it is determined that "the preset range has been exceeded," and the second power supply current adjustment is triggered.

[0113] Secondly, the second power supply current is calculated. The compensation coefficient is determined based on the degree of temperature exceedance. For example, for every 1°C that the temperature exceeds the preset temperature threshold, the current compensation increases by 0.5%, that is, the second power supply current = the first power supply current × (1 + (actual temperature - 60°C) × 0.5%), and the single adjustment range does not exceed 10% of the first power supply current (to avoid a sudden increase in current).

[0114] Finally, implement safe current limiting. The maximum threshold for the second power supply current can be set to 1.2 times the motor's rated current (e.g., if the motor's rated current is 2A, then the maximum threshold is 2.4A). Even if the temperature continues to rise, the second power supply current will not exceed this threshold, and at the same time, the orange LED indicator will flash to remind the user that the motor is in a high-temperature operating condition.

[0115] This solution effectively compensates for the decrease in motor torque output under high-temperature conditions by introducing temperature data monitoring when the screw is under-tightened and further increasing the power supply current to a second power supply current when the temperature exceeds the limit. This ensures that even in complex working conditions with fluctuating temperatures, the under-tightening phenomenon can be quickly eliminated, guaranteeing the consistency of screw connection strength. The temperature and current correlation adjustment logic not only achieves precise compensation but also avoids the risk of motor overload through safety threshold limits, improving the adaptability and reliability of the electric screwdriver in wide temperature range operating scenarios. At the same time, it reduces the rework rate caused by temperature factors, indirectly improving assembly efficiency.

[0116] In some embodiments, such as Figure 2 As shown, Figure 2 This is a schematic diagram of the training steps of the torque learning model provided in this application. Before determining the working state of the electric screwdriver based on the torque feature value, the angular velocity feature value, the acceleration feature value, and the preset torque learning model, the method further includes the following steps S610-S630.

[0117] S610: Under multiple preset standard torque values, acquire multiple reference torque data, multiple reference angular velocity data, and multiple reference acceleration data of the electric screwdriver.

[0118] In this embodiment, the preset standard torque values ​​can be typical torque values ​​covering the commonly used working range of electric screwdrivers. These values ​​are set according to the application scenario (such as electronic equipment or automotive component assembly), for example, selecting seven standard values: 0.1 N·m, 0.5 N·m, 1 N·m, 5 N·m, 10 N·m, 20 N·m, and 30 N·m, covering low, medium, and high torque ranges. The reference torque data, reference angular velocity data, and reference acceleration data can be raw data collected when the electric screwdriver completes a standard tightening operation at each standard torque value, used to reflect the normal operating state under that torque.

[0119] The specific data acquisition method can be carried out in a calibration laboratory with constant temperature (25℃±1℃) and no strong vibration (amplitude ≤0.02mm) to avoid environmental interference. Standard equipment such as the FLUKE 830 torque calibrator (accuracy ±0.05%) can be used as a standard load to simulate the tightening resistance of different screws. Data acquisition can be performed by controlling an electric screwdriver to repeat 500 tightening operations for each standard torque value (each operation is the entire process from bit contact with the screw to completion of tightening), simultaneously collecting data through a torque sensor and a six-axis IMU sensor. The sampling frequency can be 2000Hz. Each completed operation generates a set of reference data sequences containing torque, angular velocity, and acceleration (approximately 500 to 1000 data points in length, varying with tightening time). Data filtering can remove abnormal data sets caused by operational errors (such as bit slippage) (e.g., retaining 450 valid data sets for each standard torque value) to ensure the reliability of the reference data.

[0120] S620: Perform feature extraction on the multiple reference torque data, multiple reference angular velocity data and multiple reference acceleration data respectively to obtain multiple reference torque feature values, multiple reference angular velocity feature values ​​and multiple reference acceleration feature values.

[0121] The feature extraction method in this step is consistent with the above embodiment. For each set of benchmark data sequences under each standard torque value, the corresponding feature values ​​are calculated using a preset time period (e.g., 10ms) as the time window. The benchmark torque feature values ​​may include the mean torque, torque variance, and instantaneous rate of change of torque for each time window (e.g., under a standard torque of 0.5 N·m, the mean torque for a certain window is 0.498 N·m, the variance is 0.00001 N·m², and the instantaneous rate of change is 20 N·m / s). The benchmark angular velocity feature values ​​may include the peak angular velocity and steady-state angular velocity for each time window (e.g., under a standard torque of 10 N·m, the peak angular velocity for a certain window is 650° / s, and the steady-state angular velocity is 580° / s). The benchmark acceleration feature values ​​may include the mean acceleration vector magnitude and rate of change of acceleration for each time window (e.g., under a standard torque of 20 N·m, the mean acceleration magnitude for a certain window is 0.8g, and the rate of change is 300g / s).

[0122] Ultimately, the 450 sets of benchmark data sequences for each standard torque value generate approximately 4,500 benchmark feature value combinations (7-dimensional feature values) (each sequence contains 10 windows), and the 7 standard torque values ​​generate a total of approximately 31,500 benchmark feature value combinations, which serve as sample data for model training.

[0123] S630: The general model is trained based on multiple reference torque feature values, multiple reference angular velocity feature values ​​and multiple reference acceleration feature values ​​to obtain a torque learning model.

[0124] In this embodiment, the general model can be an untrained basic machine learning model (such as a random forest or gradient boosting tree). This embodiment selects a gradient boosting tree as the general model, with initial parameters set to a learning rate of 0.1, a tree depth of 6, and 200 basic evaluators. Training can be performed by adjusting the model parameters using sample data, enabling the model to accurately output the corresponding working state (such as normal, under-tightening, over-tightening, or unhooking) based on the input feature values.

[0125] The specific training process can be sample labeling - dataset partitioning - model training - model solidification.

[0126] Sample labeling can label the corresponding working state for each benchmark feature value combination, based on the standard torque value and the operation process. For example, under a standard torque of 0.5 N·m, feature value combinations with an average torque of 0.49 to 0.51 N·m are labeled as "normal state", and those with an average torque of less than 0.45 N·m are labeled as "under-tightened state".

[0127] The dataset can be split into a training set (25,200 samples) and a validation set (6,300 samples) in an 8:2 ratio.

[0128] Model training can involve inputting the training set into a general model, adjusting the model parameters through iterative optimization (e.g., 100 iterations), evaluating the model accuracy (the correctness in judging the working state) using a validation set after each iteration, and stopping training when the accuracy stabilizes above 99.5%.

[0129] Model solidification can be achieved by converting the trained model parameters (such as tree structure and node thresholds) into C language array format and burning them into the FLASH storage area of ​​the electric screwdriver's main control chip (occupying less than or equal to 1MB of space), forming a torque learning model that can be directly called.

[0130] This solution collects benchmark data and extracts feature values ​​under multiple standard torque values, enabling the trained torque learning model to cover the main working range of electric screwdrivers and improving its adaptability to different torque scenarios. The training process based on a large number of labeled samples ensures that the torque learning model can accurately identify the correlation between feature values ​​and working states, reducing the risk of misjudgment. Simultaneously, the standardized torque learning model construction process facilitates the reuse and optimization of the torque learning model during mass production, reducing the difficulty of technology implementation. Ultimately, the high-precision torque learning model provides reliable support for the precise control of electric screwdrivers.

[0131] In some embodiments, after obtaining the torque learning model, the method further includes: when the number of operations of the electric screwdriver is greater than a preset number of operations, obtaining a calibration torque; updating the torque learning model based on the calibration torque to obtain a calibrated torque learning model.

[0132] In this embodiment, the number of electric screwdriver operations can be the cumulative number of times the electric screwdriver completes a full tightening operation (from the moment the bit contacts the screw until the torque reaches the preset accuracy range and stops outputting torque). This number is automatically recorded by the main control chip through a counting module and stored in non-volatile memory to ensure that the data is not lost after power failure. The preset number of operations can be a threshold set according to the usage frequency, wear characteristics, and accuracy requirements of the electric screwdriver, for example, set to 500 times. When the cumulative number of operations exceeds this threshold, a calibration process is triggered.

[0133] It should be noted that the preset number of times can be adjusted through the accompanying software, and can be set according to the usage scenario. For example, it can be set to 300 times for high-frequency usage scenarios and 1000 times for low-frequency usage scenarios, etc. This application embodiment does not make specific limitations here.

[0134] The calibration torque can be the actual torque value measured by a high-precision external device when an electric screwdriver completes a standard tightening operation under standard calibration conditions. It reflects the true torque output under the current equipment condition. Updating the torque learning model can be done by fine-tuning the parameters of the original torque learning model using the calibration torque and corresponding feature values ​​in the calibration dataset of the above embodiment, rather than retraining the model. This is to correct the deviation caused by long-term use while retaining the generalization ability of the original model.

[0135] The specific update process includes feature extraction, sample construction, model fine-tuning, and model replacement.

[0136] Feature extraction can be performed on torque data, angular velocity data, and acceleration data collected during calibration operations, and feature values ​​(such as average torque of 4.98 N·m, peak angular velocity of 620° / s, and average acceleration modulus of 0.7g) can be extracted according to the steps in the above embodiments.

[0137] Sample construction can be achieved by combining the extracted feature values ​​with the calibration torque (5.02 N·m) and the corresponding working status label (such as "normal tightening") to form a set of calibration samples.

[0138] Model fine-tuning can be achieved through incremental training. The calibration samples are input into the original torque learning model (such as a random forest model), 90% of the basic decision tree parameters in the model are frozen, and only the node thresholds of the 10% of decision trees with the highest correlation to the calibration torque are adjusted (e.g., the torque mean is adjusted, and the judgment threshold of "greater than 4.95 N·m" is adjusted to "greater than 4.97 N·m"). This reduces the model's prediction error for the calibration torque from ±0.15 N·m to within ±0.05 N·m.

[0139] Model replacement can be achieved by writing the fine-tuned model parameters into the storage area of ​​the main control chip, overwriting the original torque learning model, and forming a calibrated torque learning model, which is then used to determine the working status in subsequent operations.

[0140] This solution introduces a calibration torque when the electric screwdriver has been used more than a preset number of times, and updates the torque learning model based on the calibration data. This effectively compensates for the decrease in model accuracy caused by motor wear and sensor drift during long-term use, ensuring that the model always matches the actual state of the equipment and maintaining the long-term stability of torque control. The incremental update method avoids model retraining, reduces the demand for hardware computing power, and the calibration process is simple and easy to operate, requiring no professional intervention. This significantly improves the practicality and maintenance convenience of electric screwdrivers in mass production scenarios, and indirectly ensures the consistency of product assembly quality.

[0141] This application also provides a control device for an electric screwdriver, such as... Figure 3 As shown, Figure 3 This is a schematic diagram of the structure of a control device for an electric screwdriver provided by the present invention. The device includes:

[0142] The acquisition module is used to acquire the operating data of the electric screwdriver, wherein the operating data includes torque data, angular velocity data and acceleration data;

[0143] The feature extraction module is used to extract features from the torque data, the angular velocity data and the acceleration data respectively to obtain torque feature values, angular velocity feature values ​​and acceleration feature values;

[0144] The determination module is used to determine the working state of the electric screwdriver based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value and the preset torque learning model;

[0145] The control module is used to adjust the operating parameters of the electric screwdriver according to the working state, so that the torque accuracy of the electric screwdriver is within a preset accuracy range.

[0146] This solution constructs a well-defined and highly efficient electric screwdriver control device by setting up an acquisition module, a feature extraction module, a determination module, and a control module. The acquisition module accurately collects operational data directly related to torque control, providing a reliable data foundation for subsequent processing. The feature extraction module extracts key feature values ​​reflecting the electric screwdriver's operating status from the raw data, eliminating redundant information and noise interference. The determination module, using a preset torque learning model, can quickly and accurately determine the current working state of the electric screwdriver, avoiding the lag and subjectivity of manual judgment. The control module dynamically adjusts operating parameters according to the working state, ensuring that torque accuracy remains stable within the preset range. This design not only makes the device's functional logic clear, facilitating later maintenance, upgrades, and troubleshooting, but also significantly improves the accuracy and response speed of electric screwdriver torque control through the efficient collaboration of each module. It effectively adapts to the torque control needs of different scenarios such as precision assembly of electronic equipment and assembly of core automotive components, ensuring the consistency and reliability of product assembly quality, while reducing the risk of workpiece damage or rework due to torque control deviations, and improving overall work efficiency.

[0147] This application also provides an electric screwdriver, which is used to control the electric screwdriver by the control method of the electric screwdriver described in any of the above embodiments or implementations.

[0148] The electric screwdriver in this solution, by applying the control method described in the above embodiments, can fully leverage the technical advantages of this control method in terms of precise torque control, adaptive adjustment under working conditions, and long-term accuracy maintenance, effectively solving the problems of insufficient torque control accuracy, weak anti-interference ability, and cumbersome calibration of traditional electric screwdrivers. Specifically, it can accurately identify different working states such as under-tightening, over-tightening, and disengagement by collecting and extracting multi-dimensional operating data, combined with a preset torque learning model, and adjust operating parameters such as power supply current accordingly. For example, when under-tightening occurs, the current is increased in stages to enhance torque output; when over-tightening or disengagement occurs, the current is quickly cut off to avoid damage to the workpiece. At the same time, it can also dynamically compensate for torque decay by combining temperature data and offset the accuracy drift over long-term use by periodically calibrating and updating the model.

[0149] The in-depth application of this control method enables the electric screwdriver to not only meet the stringent torque accuracy requirements in various scenarios such as precision assembly of electronic equipment, assembly of core automotive components, and fixation of lightweight aerospace components (e.g., the control error can be stabilized within ±0.3%), but also to have strong environmental adaptability and long-term reliability. This reduces rework and workpiece damage caused by torque control issues, significantly improves the efficiency and quality consistency of assembly operations, and provides strong support for precision assembly in the high-end manufacturing field.

[0150] This application also provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the control method of the electric screwdriver described in any of the above embodiments or implementations.

[0151] It should be noted that the main control chip and processor in the embodiments of this application can be a microcontroller unit (MCU), microprocessor unit (MPU), system on chip (SOC), digital signal processor (DSP), graphics processing unit (GPU), or other integrated circuits used to execute program instructions, process data, and control system operation.

[0152] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system, or other device. Furthermore, the embodiments of this invention are not directed to any particular programming language.

[0153] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. Similarly, for the sake of brevity and to aid in understanding one or more aspects of the invention, in the description of exemplary embodiments of the invention above, various features of the embodiments are sometimes grouped together in a single embodiment, figure, or description thereof. The claims, which follow the detailed description, are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of the invention.

[0154] Those skilled in the art will understand that the modules in the apparatus of the embodiments can be adaptively changed and placed in one or more apparatuses different from those of the embodiments. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components, except that at least some of such features and / or processes or units are mutually exclusive.

[0155] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware.

Claims

1. A control method for an electric screwdriver, characterized in that, The method includes: The operating data of the electric screwdriver is acquired, wherein the operating data includes torque data, angular velocity data, and acceleration data; Feature extraction is performed on the torque data, the angular velocity data, and the acceleration data respectively to obtain torque feature values, angular velocity feature values, and acceleration feature values; The working state of the electric screwdriver is determined based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value, and the preset torque learning model. Based on the operating state, adjust the operating parameters of the electric screwdriver to ensure that the torque accuracy of the electric screwdriver is within a preset accuracy range.

2. The method according to claim 1, characterized in that, The torque characteristic values ​​include the mean torque, the torque variance, and the instantaneous rate of change of torque; the angular velocity characteristic values ​​include the peak angular velocity and the steady-state angular velocity; and the acceleration characteristic values ​​include the mean magnitude of the acceleration vector and the rate of change of acceleration.

3. The method according to claim 1, characterized in that, The working state includes an under-tightening state, and adjusting the operating parameters of the electric screwdriver according to the working state includes: If the working state is the under-tightening state, the current power supply current of the electric screwdriver is adjusted to the first power supply current, wherein the current value of the first power supply current is greater than the current value of the current power supply current.

4. The method according to claim 1, characterized in that, The working states include overtightening and unhooking states. Adjusting the operating parameters of the electric screwdriver according to the working states includes: If the operating state is the over-tightening state or the unhooking state, then the current power supply to the electric screwdriver is cut off.

5. The method according to claim 3, characterized in that, After adjusting the current power supply current of the electric screwdriver to the first power supply current if the working state is the undertightening state, the method further includes: Obtain the temperature data of the electric screwdriver; If the temperature data exceeds the preset temperature range, the first power supply current of the electric screwdriver is adjusted to the second power supply current, wherein the current value of the second power supply current is greater than the current value of the first power supply current.

6. The method according to claim 1, characterized in that, Before determining the working state of the electric screwdriver based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value, and the preset torque learning model, the method further includes: Under multiple preset standard torque values, multiple reference torque data, multiple reference angular velocity data, and multiple reference acceleration data of the electric screwdriver are obtained; Feature extraction is performed on multiple reference torque data, multiple reference angular velocity data, and multiple reference acceleration data to obtain multiple reference torque feature values, multiple reference angular velocity feature values, and multiple reference acceleration feature values; The general model is trained based on multiple reference torque feature values, multiple reference angular velocity feature values, and multiple reference acceleration feature values ​​to obtain a torque learning model.

7. The method according to claim 6, characterized in that, After obtaining the torque learning model, the following is also included: When the number of operations of the electric screwdriver exceeds the preset number of operations, the calibration torque is obtained; The torque learning model is updated based on the calibration torque to obtain the calibrated torque learning model.

8. A control device for an electric screwdriver, characterized in that, include: The acquisition module is used to acquire the operating data of the electric screwdriver, wherein the operating data includes torque data, angular velocity data and acceleration data; The feature extraction module is used to extract features from the torque data, the angular velocity data and the acceleration data respectively to obtain torque feature values, angular velocity feature values ​​and acceleration feature values; The determination module is used to determine the working state of the electric screwdriver based on the torque characteristic value, the angular velocity characteristic value, the acceleration characteristic value and the preset torque learning model; The control module is used to adjust the operating parameters of the electric screwdriver according to the working state, so that the torque accuracy of the electric screwdriver is within a preset accuracy range.

9. An electric screwdriver, characterized in that, The electric screwdriver is applied to the control method of the electric screwdriver according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the control method for the electric screwdriver as described in any one of claims 1 to 7.

Images