An automatic calibration method and system for digital torque wrenches

Abstract

This invention discloses an automatic calibration method and system for digital torque wrenches. The method includes: a torque loading step, in which torque is applied by driving a transmission mechanism via a servo motor; a stability determination step, in which the torque is determined to be stable by monitoring the speed fluctuation rate and torque fluctuation rate; an image acquisition step, in which the wrench display image is acquired after the torque stabilizes; a standard data synchronization step, in which a standard torque value is obtained based on the sampling time of a standard torque tester; an error calculation and calibration determination step, in which the error between the wrench display value and the standard torque value is calculated and a calibration conclusion is output; and a data storage step, in which the calibration data is stored. This invention also proposes an inter-frame complementary splicing scheme for segmented LCD displays and a segmented variable acceleration torque loading strategy, thereby achieving automation, standardization, and traceability in the calibration of digital torque wrenches, and improving the accuracy and reliability of the calibration results.

Description

Technical Field

[0001] This invention relates to the field of measuring instrument calibration technology, specifically to an automatic calibration method and system for digital torque wrenches. Background Technology

[0002] A torque wrench is a tightening measuring instrument with a torque measuring mechanism, widely used in mechanical equipment assembly, maintenance, and quality inspection. The accuracy and reliability of the torque wrench directly affect the reliability and safety of fastener connections; therefore, it is necessary to periodically calibrate the torque wrench to ensure its readings are accurate.

[0003] With the development of manufacturing and the improvement of intelligent technology, torque wrenches are becoming increasingly diversified. Among them, digital torque wrenches are widely used in industrial settings due to their intuitive readings and convenient operation. The display component of digital torque wrenches mainly uses a segment LCD screen. Segment LCDs have advantages such as low power consumption, low cost, and long lifespan. However, their display method differs from the common seven-segment display; they use a segmented electrode driving method to display numbers and characters. When the displayed value changes, the segment LCD undergoes a periodic scanning and refreshing process, causing the displayed content to flicker, which poses a certain challenge to automatic recognition.

[0004] Currently, the calibration of digital torque wrenches typically employs manual operation or semi-automated auxiliary equipment. During manual calibration, the calibrator manually rotates the loading device and visually reads the value displayed on the digital torque wrench, comparing it to a standard torque meter. This method has the following problems: torque changes are unstable during manual loading; the loading speed is difficult to maintain uniformly when manually rotating the loading device, especially near the target torque value, where sudden speed changes can easily occur, leading to fluctuations in the applied torque and affecting the accuracy of the calibration results; manual reading of digital data is subject to visual errors; the segmented LCD display of the digital torque wrench exhibits scanning refresh flicker when displaying numerical changes, and the persistence of vision can easily lead to misreading or missed readings during manual visual reading.

[0005] Some existing technologies attempt to automate the calibration of digital torque wrenches by replacing manual readings with image acquisition. However, these existing technologies still have the following problems: Existing automated calibration devices lack effective dynamic control strategies during torque loading, and the loading speed cannot be adaptively adjusted according to the target torque value, easily leading to overshoot or undershoot; existing technologies are insufficient in controlling the timing of image acquisition, as image acquisition may occur before the torque has stabilized, resulting in the acquired wrench reading not being an accurate value from a stable state; for the display characteristics of segment-based LCD screens on digital torque wrenches, existing image recognition schemes have failed to effectively solve the display flickering problem caused by segment-based LCD scanning and refreshing, affecting the accuracy of digital recognition; furthermore, existing technologies are also insufficient in the synchronous matching of image acquisition data and standard torque data, making it difficult to ensure the correct correspondence between the wrench display value and the standard torque value at the same moment.

[0006] To address the aforementioned problems in existing technologies, there is a need to provide an automatic calibration method and system for digital torque wrenches to solve the problems of unstable torque loading, large reading errors, poor data synchronization, and difficulty in segment code LCD display recognition caused by manual operation, thereby achieving automation, standardization, and traceability of the calibration process. Summary of the Invention

[0007] The purpose of this invention is to provide an automatic calibration method and system for digital torque wrenches to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: an automatic calibration method for a digital torque wrench, comprising the following steps: In the torque loading step, torque is applied to the digital torque wrench by a servo motor driving the transmission mechanism. In the image acquisition step, after the torque value reaches a stable state, the display image of the digital torque wrench is acquired through a camera device; The standard data synchronization procedure uses the data sampling time of the standard torque tester as the time reference to synchronously acquire the corresponding standard torque value. The error calculation step involves aligning the wrench display value obtained from image recognition processing with the standard torque value over time to calculate the error value. The verification and judgment step outputs a verification conclusion based on the comparison result between the error value and the preset allowable error range; The data storage step involves storing the verification conclusions and related verification data into a database.

[0009] Furthermore, the image acquisition step is preceded by: The stable state determination step involves real-time monitoring of the servo motor's speed signal and the torque signal output by the torque sensor to calculate the speed fluctuation rate and torque fluctuation rate. When both the speed fluctuation rate and torque fluctuation rate are less than the corresponding preset threshold, the torque value is determined to have entered a stable state, triggering the execution of the image acquisition step. Wherein, the speed fluctuation rate is the ratio of the standard deviation of the speed sampling value per unit time to the current speed mean, and the torque fluctuation rate is the ratio of the standard deviation of the torque sampling value per unit time to the current torque mean.

[0010] Furthermore, the torque loading step specifically includes: During the acceleration phase, the servo motor is accelerated by a first preset acceleration value; During the deceleration switching phase, when the current torque value is detected to reach a preset ratio threshold of the preset target torque, the deceleration control is switched to a second preset acceleration value, wherein the second preset acceleration value is less than the first preset acceleration value. During the fine-tuning phase, once the torque value enters the preset approximation range of the preset target torque, the servo motor enters a low-speed closed-loop adjustment mode, adjusting the output torque through a proportional-integral-derivative control algorithm until the target torque value is reached.

[0011] Furthermore, when the digital torque wrench uses a segment-type LCD screen, the image acquisition step includes: The adaptive frame sampling sub-step sets the acquisition interval time for multiple consecutive frames of images according to the preset segment code liquid crystal refresh cycle, so that the acquisition interval time between two adjacent frames is less than the refresh cycle. The inter-frame complementary splicing sub-step involves performing overlap detection on the segment code display areas in the consecutive multi-frame images, extracting complete digital segments from each frame, and combining the digital segments into complete display values ​​using preset digital splicing rules.

[0012] Furthermore, the inter-frame complementary splicing sub-step specifically includes: Each frame of the image is binarized to obtain the pixel distribution map of the segment code display area; Based on the preset seven-segment digital partition template, identify whether each digital segment in each frame is lit up, and generate a digital segment lighting matrix for each frame. Perform a bitwise OR operation on the digital segment lighting matrices of adjacent frames to generate a spliced ​​digital segment lighting matrix; The complete display value is restored by lighting up the spliced ​​digital segments into a matrix.

[0013] Furthermore, the standard data synchronization steps include: While the standard torque tester completes one torque sampling, it records the timestamp and sampled value of this sampling. After the image acquisition is completed, the standard torque value with the closest timestamp is retrieved from the stored standard torque sampling data according to the acquisition time corresponding to the image. The retrieved standard torque value is used as the standard torque value corresponding to the image acquisition time.

[0014] Furthermore, the verification data stored in the data storage step includes: The identification information and model specifications of the digital torque wrench under inspection; Ambient temperature, humidity, and atmospheric pressure during the verification process; The target torque value, the actual standard torque value under load, the wrench display value, the error value, and the verification conclusion at each calibration point; The sampling curves of rotational speed fluctuation rate and torque fluctuation rate during the steady state determination process.

[0015] The present invention also provides an automatic calibration system for digital torque wrenches, including a main control console, and further comprising: A servo drive unit, which is communicatively connected to the main control console, is used to receive torque control commands and output torque drive signals. A transmission mechanism is connected between the output end of the servo drive unit and the digital torque wrench to be tested, and is used to convert the torque drive signal into torque output; The standard torque tester is connected to the main control console via a physical signal interface, and is used to collect the standard torque value during the loading process in real time and output the sampling timestamp synchronously. The camera acquisition module is located in front of the display panel of the digital torque wrench and is used to perform image acquisition after receiving a stable trigger signal; The stability determination unit is connected to the servo drive unit and the standard torque tester, respectively, and is used to calculate the speed fluctuation rate and torque fluctuation rate based on the real-time acquired speed signal and torque signal, and output the stability trigger signal when the speed fluctuation rate and torque fluctuation rate are both less than a preset threshold.

[0016] Furthermore, the camera acquisition module includes: A camera device is positioned in front of the display panel of the digital torque wrench; A frame sampling controller, connected to the camera device, is used to set the acquisition interval of multiple consecutive frames of images according to the refresh cycle of the segment-type LCD. The image processing unit, connected to the frame sampling controller, is used to perform inter-frame complementary stitching processing on the continuous multi-frame images and output the complete wrench display value.

[0017] Furthermore, the main control console has a built-in coordination control program, which is used for: The servo drive unit is controlled to output torque according to a preset segmented variable acceleration strategy; The speed fluctuation rate and torque fluctuation rate data output by the stability determination unit are received and processed in real time. Upon receiving the stable trigger signal, the camera acquisition module is controlled to perform image acquisition and synchronously record the acquisition timestamp; The corresponding standard torque value is extracted from the sampling data of the standard torque tester based on the acquisition timestamp; The image data output by the image acquisition module is processed by calling an image recognition algorithm to obtain the wrench display value and calculate the error with the standard torque value.

[0018] Compared with existing technologies, the advantages of this invention are as follows: The use of a servo motor for torque loading solves the problem of unstable torque loading caused by manual operation; the dual-parameter stability determination mechanism ensures data collection after torque stabilization, improving data accuracy; the use of a segmented variable acceleration control strategy and PID closed-loop regulation balances loading efficiency and stability; a frame-complementary splicing scheme designed for segment-code LCD displays effectively solves the identification error problem caused by display flicker; precise matching of timestamps achieves a one-to-one correspondence between the wrench display value and the standard torque value; and the automation of the verification process and complete data storage achieve standardization and traceability of the verification process. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the automatic verification method according to an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 This invention discloses an automatic calibration method for digital torque wrenches, aiming to solve the problems of unstable torque loading, large reading errors, poor data synchronization, and difficulty in segment code LCD display recognition in existing technologies, thereby achieving automation, standardization, and traceability of the calibration process. The method includes the following steps: torque loading step, image acquisition step, standard data synchronization step, error calculation step, calibration judgment step, and data storage step.

[0022] The automatic calibration method for digital torque wrenches provided in this embodiment includes the following specific implementation steps: Regarding torque loading, a servo motor drives a transmission mechanism to apply torque to the digital torque wrench. Specifically, a series of target torque points can be preset. After receiving control commands, the servo motor drives the transmission mechanism to rotate using a segmented variable acceleration method.

[0023] Specifically, it includes: an acceleration phase in which the servo motor is accelerated by a first preset acceleration value, which can be 100 to 500 rpm / s, preferably 200 rpm / s; During the deceleration switching phase, when the current torque value is detected to reach a preset ratio threshold of the preset target torque, such as 80%, the deceleration control is switched to the second preset acceleration value. The second preset acceleration value can be 20 to 80 rpm / s, preferably 50 rpm / s, and the second preset acceleration value is less than the first preset acceleration value. During the fine-tuning phase, when the torque value enters the preset approximation range of the preset target torque, such as 95% to 105% of the target torque, the servo motor enters a low-speed closed-loop adjustment mode, and adjusts the output torque through a proportional-integral-derivative (PID) control algorithm until the target torque value is reached.

[0024] In the above fine-tuning stage, the parameter tuning of the proportional-integral-derivative control algorithm can be carried out using the following method: the initial value of the proportional coefficient Kp can be estimated based on the rated torque and target torque range of the servo motor, and is usually set to 1% to 10% of the ratio of the rated torque to the maximum deviation; The initial value of the integration time constant Ti can be set between 100ms and 500ms; The initial value of the differential coefficient Kd can be set to 1% to 5% of the product of Kp and the integration time.

[0025] During specific debugging, Ki=0 and Kd=0 can be set initially, and Kp can be gradually increased until the system exhibits continuous oscillation. Then, Kp can be decreased until the oscillation disappears. Finally, Ki can be gradually increased to eliminate steady-state error, and Kd can be added appropriately as needed to suppress overshoot. For this invention, considering the response characteristics of torque loading, it is recommended to use an incremental PID algorithm with a control period of 1ms to 5ms.

[0026] Regarding image acquisition, after the torque value stabilizes, the display image of the digital torque wrench is captured by a camera device. Specifically, by monitoring the speed signal of the servo motor and the torque signal output by the torque sensor in real time, the speed fluctuation rate and torque fluctuation rate are calculated. When both are less than the corresponding preset thresholds, the torque value is determined to have stabilized, triggering the camera device to acquire images. The speed fluctuation rate can be set from 0.5% to 2%, and the torque fluctuation rate can be set from 0.3% to 1%. Acquisition is triggered when both are below the preset thresholds.

[0027] The methods for determining the thresholds for speed fluctuation rate and torque fluctuation rate are as follows: The threshold settings are related to the accuracy class and range of the digital torque wrench under test. For high-precision wrenches, such as those with an allowable indication error range of ±1%, it is recommended to set the speed fluctuation rate threshold to 0.5% and the torque fluctuation rate threshold to 0.3%. For wrenches with a standard accuracy level, such as those with an allowable indication error range of ±4%, the thresholds can be appropriately relaxed to 2% for speed fluctuation and 1% for torque fluctuation. The thresholds should be determined using the following method: First, conduct multiple loading tests under the same conditions, recording the speed fluctuation and torque fluctuation data corresponding to the point of visual stability in each loading test; then, take the 90th percentile of the test data as the recommended threshold value. Furthermore, for torque wrenches with different ranges, the thresholds should be proportionally adjusted according to the range.

[0028] Regarding standard data synchronization, the data sampling time of the standard torque tester is used as the time reference to synchronously acquire the corresponding standard torque value. The specific implementation method includes: firstly, recording the sampling timestamp and sampling value simultaneously when the standard torque tester completes one torque sampling; the sampling frequency can be set to 100 to 1000 Hz, preferably 500 Hz. The selection criteria for the sampling frequency are as follows: The determination of the sampling frequency should comprehensively consider the following factors: The bandwidth of a standard torque tester is typically above 1000Hz, therefore the sampling frequency should be lower than its bandwidth but higher than the highest frequency component of the torque signal. According to the Nyquist sampling theorem, the sampling frequency should be at least twice the highest frequency of the torque signal. Considering the dynamic characteristics during torque loading, the highest frequency of the torque fluctuation signal is usually no more than 50Hz, therefore a sampling frequency of 200Hz or higher is sufficient. However, to improve time synchronization accuracy, it is recommended to set the sampling frequency between 500Hz and 1000Hz, at which point the time matching error with the image acquisition time can be controlled within ±1ms.

[0029] Secondly, after the image acquisition is completed, the standard torque value with the closest timestamp is retrieved from the stored standard torque sampling data according to the acquisition time corresponding to the image. Finally, the retrieved standard torque value is used as the standard torque value corresponding to the image acquisition time.

[0030] In terms of error calculation, the error value is calculated by aligning the wrench display value obtained from image recognition processing with the standard torque value over time. Specifically, image recognition processing can use OCR technology to analyze the acquired image, extract the numbers on the digital torque wrench display, and then compare the recognized wrench display value with the standard torque value, calculating the difference between the two as the error value.

[0031] The error value is calculated as follows: Absolute error value = Wrench display value - Standard torque value; Relative error value = (Absolute error value / Standard torque value) × 100%. When multiple valid recognition results are obtained after inter-frame complementary stitching, the average value of the multiple recognition results is taken as the final wrench display value. If the standard deviation of multiple recognition results exceeds 0.5 minimum division values, the recognition is considered abnormal, and image acquisition is repeated. For the estimated digits displayed on the segment LCD screen, the recognition result should retain the value of that digit, and the confidence level of the estimated digit should be marked when storing the data.

[0032] In terms of verification and judgment, the verification conclusion is output based on the comparison between the error value and the preset allowable error range. Specifically, the upper and lower limits of the allowable error can be preset, and the calculated error value is compared with the allowable error range. If it is within the allowable range, it is judged as qualified; if it exceeds the range, it is judged as unqualified.

[0033] In terms of data storage, the verification conclusions and related verification data are stored in the database. Specifically, information such as the target torque value, standard torque value, wrench display value, error value, and verification conclusion for each verification can be recorded in the database in a structured manner, including fields such as verification date, verification personnel, wrench model, and verification result.

[0034] The environmental parameters are collected as follows: Environmental parameters are collected using the following methods: A platinum resistance thermometer with an accuracy of ±0.5℃ is recommended for the temperature sensor, installed inside the calibration system cabinet near the wrench being tested to accurately reflect the actual temperature around the wrench. A capacitive humidity sensor with an accuracy of ±3%RH is recommended. A piezoresistive pressure sensor with an accuracy of ±0.1kPa is recommended. Environmental parameters are collected as follows: the environmental sensors are activated 5 minutes before the start of each batch of calibration, and environmental parameters are collected every 30 seconds, with the average value calculated as the recorded environmental parameter value for this calibration.

[0035] This method solves the problem of unstable torque loading caused by manual operation by introducing a servo motor for stable torque loading. Image acquisition after the torque is stable avoids data errors in unstable states. Data synchronization based on the sampling time of the standard torque tester ensures the accurate correspondence between the wrench display value and the standard torque value. The automation of error calculation and verification judgment eliminates the subjectivity of manual judgment. Verification data is stored in the database to realize the traceability of the verification process.

[0036] In some of the solutions described above in this application, image acquisition is performed only after the torque value has reached a stable state, thus ensuring that the acquired digital torque wrench reading is the accurate value after loading. However, this process does not clearly define how to accurately determine whether the torque has reached a stable state. If an unstable torque state is mistakenly identified as a stable state, the image acquisition timing will be premature, and the acquired wrench reading will not be the true stable torque value, thereby affecting the accuracy of subsequent verification error calculations and compromising the reliability of the verification results. Relying solely on experience or a single parameter to determine the stable state is also prone to misjudgment. Therefore, a more accurate and reliable method for determining the stable state is needed to determine the triggering timing of image acquisition.

[0037] To address this, this application further proposes a stable state determination step prior to the image acquisition step. This involves real-time monitoring of the servo motor's speed signal and the torque signal output from the torque sensor to calculate the speed fluctuation rate and torque fluctuation rate. When both the speed fluctuation rate and torque fluctuation rate are less than their respective preset thresholds, the torque value is determined to have entered a stable state, triggering the image acquisition step. Specifically, the speed fluctuation rate is the ratio of the standard deviation of the speed sampling values ​​per unit time to the current average speed, and the torque fluctuation rate is the ratio of the standard deviation of the torque sampling values ​​per unit time to the current average torque.

[0038] The stability determination step aims to ensure that the applied torque has reached a stable state before image acquisition of the digital torque wrench. Its core function is to avoid image acquisition when the torque fluctuates or is not fully stable, thus ensuring that the acquired wrench display value is accurate and reliable. This step can be implemented by setting up a separate hardware module or integrating a software module into the main controller. This module continuously receives data from the sensor and executes a preset algorithm to determine stability. Alternatively, a state machine can be defined in the calibration system's control software. When specific stability conditions are met, the state machine switches from loading to stability, triggering subsequent image acquisition operations.

[0039] Real-time monitoring of the servo motor's speed signal and the torque signal output from the torque sensor is crucial. Real-time monitoring refers to continuously acquiring these signals or acquiring them at high frequency to promptly reflect dynamic changes during torque loading. The servo motor's speed signal reflects the motion state of the power output end of the loading device, and its stability directly affects the stability of the applied torque. The torque signal output from the torque sensor directly reflects the actual torque applied to the digital torque wrench and its fluctuations. Simultaneous monitoring of these two signals allows for a comprehensive assessment of the system's stability from both the power source and actual force dimensions. Servo motors are typically equipped with encoders or Hall effect sensors, which can directly output speed signals. Torque sensors convert torque into electrical signals using principles such as strain gauges. These signals can be acquired by an analog-to-digital converter (ADC) and fed into the control system. Alternatively, the servo motor's speed data can be directly acquired from the servo driver via an industrial bus (such as EtherCAT or Profinet), and the torque signal can be obtained through a dedicated torque sensor interface module, ensuring real-time and accurate data transmission.

[0040] Calculate the fluctuation rate of rotational speed and torque, where fluctuation rate is a quantitative indicator measuring the stability of a signal over a period of time. By calculating the fluctuation rate of rotational speed and torque, the abstract concept of stability can be transformed into concrete numerical values, facilitating objective judgment. In the control system, a sliding time window can be set, within which N rotational speed samples and N torque samples are continuously collected. The sampling frequency can be set to 100 to 1000 Hz, and the sliding time window width can be set to 0.5 to 2 seconds. Then, the fluctuation rate is calculated based on the ratio of the standard deviation to the mean. Alternatively, a digital filter can be used to preprocess the original signal to remove high-frequency noise, and then the percentage of the difference between the maximum and minimum values ​​of the samples relative to the mean is calculated within a fixed sampling period as an approximation of the fluctuation rate.

[0041] Using the ratio of standard deviation to mean as volatility has normalization properties, meaning it is a relative value and is not affected by the absolute value of speed or torque. This implies that regardless of whether the target torque is high or low, as long as the fluctuation level is relatively stable, the volatility can be reflected, making the preset threshold more universal. Typical preset thresholds, determined experimentally, are 0.5% to 2% for speed volatility and 0.3% to 1% for torque volatility.

[0042] When both the speed fluctuation rate and torque fluctuation rate are less than their respective preset thresholds, the torque value is determined to have entered a stable state, triggering the image acquisition step. This is the core logic of the stable state determination. Only when the fluctuations of the two key parameters are sufficiently small, i.e., when the system's power output and actual force tend to stabilize, is the torque loading considered complete and the system considered to have entered a stable state. This dual determination mechanism improves the accuracy and reliability of the determination. In the control software, two independent preset thresholds can be set, one for speed fluctuation rate and one for torque fluctuation rate. The system continuously compares the calculated fluctuation rate with its respective threshold. When both simultaneously meet the condition of being less than the threshold, a trigger signal is sent to the image acquisition module. Alternatively, the two fluctuation rates can be weighted and averaged or logically combined to form a comprehensive stability index, which is then compared with a single preset threshold. Once the condition is met, image acquisition is triggered through an interrupt or event notification mechanism. When both the speed fluctuation rate and torque fluctuation rate simultaneously meet the condition of being less than the preset threshold, the stability determination unit notifies the main control console through an interrupt signal, and the coordination control program then sends a trigger command to the camera acquisition module. The camera device completes exposure and image acquisition within a fixed delay after receiving the trigger command, such as 10 to 50 milliseconds.

[0043] Through the above technical solution, this application introduces a stable state determination step. Before image acquisition, the system monitors the servo motor's speed signal and the torque signal output by the torque sensor in real time. The system calculates the speed fluctuation rate and torque fluctuation rate of these two signals per unit time and compares them with preset thresholds. Only when both the speed fluctuation rate and torque fluctuation rate are less than the corresponding preset thresholds is the torque value determined to have entered a stable state, triggering the image acquisition step. This dual-dimensional, quantitative determination method ensures that image acquisition is performed after the torque has truly stabilized, thus solving the problem of inaccurate readings caused by torque fluctuations and improving the accuracy of subsequent error calculations. Compared to single parameter or empirical judgment, simultaneously monitoring the fluctuation rates of speed and torque can more comprehensively and objectively reflect the stable state of the system, avoiding misjudgments and making the verification results more reliable. Furthermore, using the ratio of standard deviation to mean as the fluctuation rate gives the preset threshold a normalization characteristic, adapting to different target torque values ​​in verification conditions without frequent parameter adjustments, thus improving the method's versatility and ease of operation. This solution provides a precise triggering mechanism for image acquisition, further improving the automation level of the automatic calibration method for digital torque wrenches and reducing manual intervention and potential human error.

[0044] In some of the solutions described above in this application, a torque loading step is proposed to apply torque to a digital torque wrench via a servo motor-driven transmission mechanism, completing the torque input before verification. However, during the torque loading process, traditional loading methods lack a dynamic speed adjustment strategy to adapt to the loading process. The loading speed is constant throughout or improperly adjusted. Loading too slowly when far from the target torque will reduce verification efficiency, while loading too quickly when approaching the target torque can easily cause torque overshoot. It is difficult to ensure that the torque reaches the target value smoothly and accurately, which can easily generate unnecessary torque fluctuations. This not only affects the subsequent determination of the stable state but also reduces the accuracy of the final verification result.

[0045] In response, this application further proposes that the aforementioned torque loading steps specifically include: an acceleration phase, in which the servo motor is accelerated by a first preset acceleration value; a deceleration switching phase, in which, when the current torque value is detected to reach a preset proportional threshold of the preset target torque, the deceleration control is switched to a second preset acceleration value, the second preset acceleration value being less than the first preset acceleration value; and a fine adjustment phase, in which, when the torque value enters a preset approximation range of the preset target torque, the servo motor enters a low-speed closed-loop adjustment mode, and the output torque is adjusted through a proportional-integral-derivative control algorithm until the target torque value is reached.

[0046] The acceleration phase refers to the initial stage of the torque loading process, at which point there is a significant gap between the current torque value and the preset target torque value. During this phase, the servo motor is accelerated using a first preset acceleration value to rapidly increase the torque. The first preset acceleration value is a pre-set acceleration parameter, the magnitude of which can be determined experimentally or through simulation based on the performance of the servo motor and transmission mechanism, as well as the range of the torque wrench under test. For example, it can be set to a relatively high constant acceleration, typically 100 to 500 rpm / s, preferably 200 rpm / s. The servo motor, as the power source, can precisely control angular displacement, angular velocity, and torque. It can be implemented using an AC servo motor or a DC servo motor, typically equipped with an encoder for position and speed feedback to achieve precise control. Acceleration control increases the output torque or speed of the servo motor according to a preset acceleration curve by controlling the input signals, such as voltage, current, or pulse frequency. This can be achieved through open-loop control or simple closed-loop speed control.

[0047] The deceleration switching phase is a transitional phase implemented to avoid overshoot when the torque value approaches the preset target torque. When the current torque value is detected to have reached a preset proportional threshold of the preset target torque, the system switches to a second preset acceleration value for deceleration control. The preset proportional threshold is a proportional coefficient between 0 and 1, used to define the critical point for switching from the acceleration phase to the deceleration switching phase, typically set to 70% to 90%. The second preset acceleration value is a pre-set acceleration parameter whose absolute value is less than the first preset acceleration value, used to reduce the loading speed when the torque approaches the target value; a typical value is 20 to 80 rpm / s, preferably 50 rpm / s. Deceleration control is achieved by adjusting the drive current or pulse frequency of the servo motor to ensure smooth deceleration when the torque loading approaches the target value.

[0048] The fine-tuning stage is the final stage of torque loading, designed to precisely stabilize the torque at the target value. Once the torque value enters the preset approximation range of the target torque, the servo motor enters a low-speed closed-loop adjustment mode. The proportional-integral-derivative (PID) control algorithm adjusts the output torque until the target torque value is reached. The preset approximation range is a small interval centered on the preset target torque, such as 95% to 105% of the target torque value. The low-speed closed-loop adjustment mode refers to the servo motor operating at lower speeds or torque outputs, comparing the real-time feedback torque value with the target torque value to form a closed-loop control. In this mode, the system continuously monitors the actual torque and makes fine adjustments based on deviations. The proportional-integral-derivative (PID) control algorithm is a classic closed-loop control algorithm. It achieves precise, stable, and rapid adjustment of the output torque by using the proportional term (P) to quickly respond to errors, the integral term (I) to eliminate steady-state errors, and the derivative term (D) to suppress overshoot and oscillations. PID parameters can be determined using conventional debugging methods based on specific operating conditions.

[0049] Through the above technical solution, this application divides the torque loading process into an acceleration phase, a deceleration switching phase, and a fine adjustment phase, and adopts different control strategies for each phase, effectively solving the problems of low efficiency, easy overshoot or undershoot, and large torque fluctuations in traditional loading methods. In the acceleration phase, a larger first preset acceleration value is used to accelerate the servo motor, enabling the torque to quickly approach the target value, significantly improving the overall calibration efficiency. When the torque value approaches the preset target torque, the system enters the deceleration switching phase. By switching the acceleration to a second preset acceleration value less than the first preset acceleration value for deceleration control, the loading inertia is effectively reduced, avoiding torque overshoot of the target value and reducing overshoot. Subsequently, after the torque value enters the preset approximation range of the preset target torque, the system enters the fine adjustment phase. The servo motor switches to a low-speed closed-loop adjustment mode, and a proportional-integral-derivative control algorithm is used to precisely adjust the output torque. This closed-loop control method can eliminate minute deviations in real time, ensuring that the torque smoothly and accurately reaches and stabilizes at the target value, avoiding undershoot or slight overshoot. This provides a highly stable and accurate torque loading foundation for subsequent stability determination, image acquisition, and error calculation. Overall, this staged variable acceleration control strategy balances loading efficiency and loading stability, significantly improving the accuracy and reliability of the automatic calibration method for digital torque wrenches.

[0050] In some of the solutions described above in this application, an image acquisition step is proposed to acquire the display image of the digital torque wrench after the torque has stabilized, for subsequent identification to obtain the wrench display value and complete the verification. However, when the digital torque wrench uses a segment LCD screen, the segment LCD has a periodic scanning and refreshing process. A single image acquisition often captures incomplete images during the refresh process, failing to obtain complete display content, which can easily lead to errors in subsequent digital recognition and thus affect the accuracy of the verification results. Existing image acquisition methods cannot adapt to the refresh characteristics of segment LCDs, thereby failing to solve the problem of incomplete display.

[0051] In response, this application further proposes that when a digital torque wrench uses a segment-type LCD screen, the image acquisition step includes: an adaptive frame sampling sub-step, which sets the acquisition interval time of multiple consecutive frames of images according to a preset segment-type LCD refresh cycle, so that the acquisition interval time between two adjacent frames is less than the refresh cycle; and an inter-frame complementary splicing sub-step, which performs overlap detection on the segment display areas in multiple consecutive frames of images, extracts the complete digital segments displayed in each frame, and combines the digital segments into a complete display value through a preset digital splicing rule.

[0052] The specific implementation of the preset digital splicing rules is as follows: The preset digital splicing rules include the following processing logic: Arrange the digital segment illumination matrix of N consecutive frames (e.g., N=5 to 10 frames) in chronological order. For each segment position of each digital character, count the number of times it is illuminated (i.e., the number of times its grayscale value is 0) within N frames. If a segment is illuminated more than N / 2 times within N frames, it is determined that the segment is in an illuminated state when fully displayed; otherwise, it is determined to be in an off state. This majority-voting-based splicing rule can effectively handle the problem of incomplete display content in each frame during segment LCD refresh.

[0053] The adaptive frame sampling sub-step aims to optimize the image acquisition process based on the refresh characteristics of the segment LCD used in the digital torque wrench, ensuring complete display information is captured. Specifically, it sets the acquisition interval for multiple consecutive frames based on a preset segment LCD refresh cycle, ensuring the interval between adjacent frames is less than the refresh cycle. A typical segment LCD refresh cycle is 50 to 100 milliseconds, therefore the acquisition interval between adjacent frames should be less than the refresh cycle. One implementation involves obtaining the segment LCD refresh cycle during system initialization by reading the specifications of the digital torque wrench or its display module, and then dynamically adjusting the frame rate or trigger frequency of the camera device based on this refresh cycle. Another implementation involves using a high-speed camera to continuously acquire images at a fixed frame rate significantly higher than any known refresh cycle, and then using a software algorithm to filter out image frames from the continuously acquired image stream that meet the condition that the acquisition interval between adjacent frames is less than the refresh cycle.

[0054] The inter-frame complementary stitching sub-step aims to utilize the continuous multi-frame images obtained from the adaptive frame sampling sub-step to intelligently combine potentially incomplete display segments from different frames into a complete display value, thereby overcoming the display incompleteness problem caused by segment code LCD refresh. Overlap detection is performed on the segment code display areas in the continuous multi-frame images. This operation ensures that the corresponding digit display areas in different frames are accurately aligned during image stitching. One implementation method is to predefine a region of interest (ROI) containing the segment code display areas before image processing, and then perform image registration on the ROI of each frame to determine the relative displacement between frames, thereby correcting the image. Another implementation method is to set specific reference markers around the segment code display areas, identify these markers in each frame, and use them as a reference to perform geometric transformations and alignment of the segment code display areas.

[0055] When extracting complete digital segments from each frame, this operation aims to identify the segments that are fully illuminated at the current moment in each frame. One approach is to segment the segment display area of ​​each frame, identify the individual segments that make up the digit (e.g., the seven segments of a seven-segment display), and then perform brightness analysis or connectivity detection on each segment to determine whether it is fully illuminated. Another approach is to use machine learning-based image recognition techniques to train a model to identify the illumination status of each segment in the segment display area.

[0056] The inter-frame complementary stitching sub-step specifically includes: performing binarization processing on each frame image to obtain the pixel distribution map of the segment code display area; The specific method for image binarization is as follows: An adaptive thresholding method is used for image binarization. The specific steps are as follows: Calculate the global grayscale mean and global grayscale standard deviation for the grayscale image of the segment code display area. Use the global grayscale mean minus 0.5 to 1.5 times the standard deviation as the adaptive threshold. Pixels with grayscale values ​​greater than the threshold are considered background (grayscale 255), while those less than or equal to the threshold are considered illuminated areas (grayscale 0). For cases of uneven illumination, the image can be first processed for illumination correction, using homomorphic filtering or the Retinex algorithm to eliminate the influence of illumination before binarization.

[0057] Based on the preset seven-segment digital partition template, it identifies whether each digital segment in each frame is lit up, and generates a digital segment lighting matrix for each frame.

[0058] The specific content of the preset seven-segment digital partition template is as follows: The seven-segment digital partition template defines the pixel area coordinates of the seven independent display segments of each digital character on the segment LCD screen in the image. The template data structure includes: the number of characters N, usually 3 to 5 digits, the coordinates of the region of interest (ROI) corresponding to each character, and the coordinates or mask image of the dividing lines of the seven segments within each character. For different models of digital torque wrenches, the display position and character size of the segment LCD may be different. Therefore, the template can be automatically generated during system initialization in the following way: the image of the standard digits displayed by the digital torque wrench is acquired, and then the boundary of each digital character and the separation position of the internal segments are automatically identified through image processing algorithms to generate and store the template file corresponding to the wrench model.

[0059] The digital segment illumination matrices of adjacent frames are bitwise ORed to generate a concatenated digital segment illumination matrix; the complete display values ​​are then reconstructed based on this concatenated matrix. The image processing unit can be implemented using a DSP or GPU, or it can be completed by running image processing algorithms on a general-purpose CPU.

[0060] Through the above technical solution, this application provides an efficient and accurate image acquisition and processing mechanism to address the periodic scanning refresh and flickering problems of segment-type LCD displays commonly used in digital torque wrenches when displaying numerical changes. The camera device captures images of the display panel, ensuring the quality of the original image. The frame sampling controller precisely sets the acquisition interval for multiple consecutive frames based on the refresh cycle of the segment-type LCD, enabling the capture of multiple frames containing different segment illumination states within a complete refresh cycle of the LCD display. This multi-frame acquisition strategy effectively avoids the problem of incomplete digital display due to incomplete refresh during a single acquisition. Subsequently, the image processing unit performs inter-frame complementary stitching processing on these consecutive multi-frame images, intelligently combining the fully displayed digital segments from different frames, thereby overcoming the recognition difficulties caused by the flickering of the segment-type LCD display and ultimately outputting a complete and accurate wrench display value. This fundamentally solves the problem of low accuracy of traditional image recognition methods when processing segment LCD displays, significantly improving the reliability and accuracy of digital torque wrench value acquisition, providing a solid data foundation for subsequent error calculation and verification, and thus ensuring the accuracy and effectiveness of the entire automatic verification method.

[0061] In some of the solutions described above in this application, a standard data synchronization step is proposed to align the wrench display value with the standard torque value for error calculation. However, in this process, existing solutions cannot accurately match the standard torque value corresponding to the image acquisition moment, which easily leads to time misalignment between the wrench display value and the standard torque value. This results in inaccurate error calculation results, cannot guarantee the reliability of the verification conclusion, and makes it difficult to solve the problem of poor synchronization between standard torque data and image acquisition data in the prior art.

[0062] In this regard, this application further proposes standard data synchronization steps, which specifically include the following: First, while the standard torque tester completes a torque sampling, it records the timestamp and sampled value of that sampling. This step aims to ensure that the standard torque tester accurately records the corresponding time information while acquiring the torque sampled value. This simultaneous recording mechanism is the foundation for achieving accurate data synchronization. Specifically, this can be achieved in the following ways: One way is that the standard torque tester integrates a high-precision clock module, which immediately binds the sampled torque value with the timestamp provided by the current clock module and outputs it after each torque sampling; another way is that when the standard torque tester outputs a torque sampling signal, the external control system responds in real time through a hardware interrupt or a high-priority task, and obtains the current time from the system's unified high-precision clock source, associating and recording this time with the received torque sampled value. The sampling frequency can be set from 100 to 1000 Hz, preferably 500 Hz.

[0063] Secondly, after image acquisition is completed, the standard torque value with the closest timestamp is retrieved from the stored standard torque sampling data based on the acquisition time corresponding to the image. This step is used to find the standard torque value that best matches the actual acquisition time of the image from the pre-stored standard torque sampling data after image acquisition. Specifically, the following methods can be used: One method is to store all standard torque sampling data and their corresponding timestamps in chronological order. When the image acquisition time is obtained, a binary search algorithm or a linear search algorithm is used to quickly locate the standard torque sampling point with the smallest time difference from the image acquisition time in the stored dataset. Another method is to set a time matching window, search for all standard torque sampling points within a certain time range before and after the image acquisition time, and select the one with the closest timestamp to the image acquisition time.

[0064] Finally, the retrieved standard torque value is used as the standard torque value corresponding to the image acquisition time. This step clarifies the purpose of the standard torque value obtained through the above retrieval mechanism: it serves as a standard reference value that precisely corresponds to the image acquisition time. Specifically, the retrieved standard torque value can be directly used as the benchmark value for error calculation based on the digital torque wrench reading at the image acquisition time. Alternatively, if there is still a slight deviation between the retrieved timestamp and the image acquisition time, further interpolation algorithms can be used, such as linear or polynomial interpolation based on the retrieved standard torque value and its adjacent standard torque values, to obtain a more accurate standard torque value at the image acquisition time.

[0065] Through the above technical solution, this application effectively solves the problem of poor time synchronization between image acquisition data and standard torque data during the automatic calibration of digital torque wrenches. By recording the timestamp and sampled value simultaneously with torque sampling on the standard torque tester, it ensures that each standard torque data has a precise timestamp, avoiding deviations caused by time lag or advance recording. Subsequently, after image acquisition is completed, based on the actual acquisition time of the image, the closest standard torque value in time is retrieved from the stored time-stamped standard torque data and used as the standard torque value corresponding to the image acquisition time. This precise matching mechanism based on timestamps can minimize the time misalignment between the wrench display value and the standard torque value, thus providing accurate and aligned benchmark data for subsequent error calculation. In view of this, this application can significantly improve the accuracy and reliability of calibration results, ensure the validity of calibration conclusions, and thereby realize the automation, standardization, and traceability of the digital torque wrench calibration process.

[0066] In some of the solutions described above in this application, an automatic calibration system for digital torque wrenches is proposed, comprising a main control console, a servo drive unit, a transmission mechanism, a standard torque tester, a camera acquisition module, and a stability judgment unit, among other functional units, to achieve automated calibration of digital torque wrenches. However, in this process, the specific working content of the control program coordinating the operation of each functional unit in the main control console is not clearly defined. The workflow between the various units lacks unified and orderly scheduling and coordination, which easily leads to problems such as inadequate execution of torque loading strategies, untimely processing of stability judgment data, asynchronous acquisition of image data and standard torque data, and inability to automatically complete error calculation. This fails to guarantee the orderly and accurate operation of the entire automated calibration process, affecting the accuracy of the calibration results and the degree of automation in the calibration process.

[0067] In response, this application proposes an automatic calibration system for digital torque wrenches, including a main control console, a servo drive unit, a transmission mechanism, a standard torque tester, a camera acquisition module, and a stability judgment unit.

[0068] The central control console is the core controller of the system, with a built-in coordination control program responsible for the unified management, scheduling, and synchronization of all hardware units and software functions throughout the system. The coordination control program can adopt a state machine-based or event-driven architecture, coordinating the operation of each module by defining different verification stages and state transition conditions.

[0069] The servo drive unit communicates with the main control console to receive torque control commands and output torque drive signals. The transmission mechanism connects the output of the servo drive unit to the digital torque wrench under test, converting the torque drive signal into torque output.

[0070] The standard torque tester is connected to the main control console via a physical signal interface to collect standard torque values ​​during the loading process in real time and output sampling timestamps synchronously.

[0071] The image acquisition module is located in front of the display panel of the digital torque wrench and includes a camera device, a frame sampling controller, and an image processing unit. The camera device is used to capture images from the display panel; the frame sampling controller is used to set the acquisition interval according to the refresh cycle of the segment LCD; and the image processing unit is used to perform inter-frame complementary stitching processing on multiple consecutive images to output the complete wrench display value.

[0072] The stability judgment unit is connected to the servo drive unit and the standard torque tester respectively. It is used to calculate the speed fluctuation rate and torque fluctuation rate based on the real-time acquired speed signal and torque signal, and outputs a stability trigger signal when both are less than the preset threshold.

[0073] The coordination control program is specifically used to: control the servo drive unit to output torque according to a preset segmented variable acceleration strategy; receive and process the speed fluctuation rate and torque fluctuation rate data output by the stability judgment unit in real time; after receiving the stability trigger signal, control the camera acquisition module to perform image acquisition and synchronously record the acquisition timestamp; extract the corresponding standard torque value from the sampling data of the standard torque tester according to the acquisition timestamp; call the image recognition algorithm to process the image data output by the image acquisition module, obtain the wrench display value, and calculate the error with the standard torque value.

[0074] During system operation, the main control console sends torque control commands to the servo drive unit according to the preset verification process; the stability judgment unit continuously monitors the speed and torque signals, and determines the fluctuation rate by calculating the deviation rate; when the speed fluctuation rate and torque fluctuation rate simultaneously meet the stability conditions, the camera acquisition module is immediately triggered to capture the image on the display panel; simultaneously, the standard torque tester transmits the standard torque value corresponding to the sampling timestamp to the main control console; finally, the system performs time matching between the image recognition result and the standard torque value, calculates the error, and stores the complete verification data. Compared with the basic scheme, this application adopts a dual-parameter stability judgment mechanism, which significantly improves the reliability of stable state recognition and avoids false triggering that may occur due to single-parameter judgment; the triggering time of the camera acquisition module is strictly limited to the stable state, effectively overcoming the recognition difficulties caused by the flickering of the segment code LCD; the timestamp synchronization mechanism of the standard torque tester ensures the accuracy of data matching and eliminates the subjective error introduced by manual intervention.

[0075] Through the above technical solution, this application achieves fully automated operation of the verification process, eliminating the instability of manual loading and errors in visual readings, ensuring the accuracy of segment code LCD display recognition, and establishing a strict time correspondence between the wrench display value and the standard torque value. Overall, this system not only improves verification efficiency and result reliability, but also achieves complete traceability of the verification process through data storage, providing solid technical support for the standardized verification of digital torque wrenches.

[0076] In some of the solutions described above in this application, a data storage step is proposed to store the verification conclusions and related verification data in a database. However, in their implementation, only the storage of verification conclusions and related verification data is mentioned in general terms, without clearly specifying the specific scope of verification data that needs to be stored. This can easily lead to the problem of missing key verification information, failing to meet the requirements of metrological verification for traceability of the verification process and reproducibility of verification data, and also not complying with the relevant norms for standardized management of metrological verification work.

[0077] In this regard, this application further proposes that the verification data stored in the data storage step includes: the identification information and model specifications of the digital torque wrench under test; the ambient temperature, humidity and atmospheric pressure during the verification process; the target torque value, the actual standard torque value, the wrench display value, the error value and the verification conclusion at each verification point; and the speed fluctuation rate and torque fluctuation rate sampling curves during the steady state determination process.

[0078] The identification information for the tested digital torque wrench refers to a unique code or name used to identify it, such as a serial number or asset number. This ensures a one-to-one correspondence between calibration data and a specific wrench, preventing data confusion. The model specification indicates the wrench's specific type, range, accuracy class, and other technical parameters, such as "XX brand - YY model - 100 N.m range." This facilitates the classification and performance evaluation of calibration results. This information can be obtained and stored through various methods, including manual input, scanning barcodes or QR codes, or RFID reading.

[0079] The ambient temperature, humidity, and atmospheric pressure during the calibration process refer to the air temperature, water vapor content, and air pressure at the calibration site. These environmental parameters have a significant impact on the performance and measurement results of precision measuring instruments. For example, temperature changes can cause materials to expand and contract with temperature changes, affecting the accuracy of torque sensors. Therefore, recording these parameters is crucial for ensuring the accuracy and reproducibility of calibration results. Environmental parameters can be acquired and recorded in real time using environmental sensors integrated into the calibration system, such as temperature and humidity sensors and atmospheric pressure sensors, or input manually by reading data from standard environmental monitoring equipment.

[0080] The target torque value at each calibration point refers to the preset torque setpoint that needs to be applied to the digital torque wrench during the calibration process, such as 20 N·m, 50 N·m, etc. The actual applied standard torque value refers to the torque value actually measured at the target torque point using a standard torque tester; it is the benchmark for measuring the accuracy of the wrench's displayed value. The wrench display value refers to the value read from the digital torque wrench's own display screen when the actual standard torque value is applied. The error value is the deviation calculated based on the wrench display value and the actual applied standard torque value. The calibration conclusion is the conclusion determining whether the wrench is qualified based on the comparison between the error value and the preset allowable error range, such as "qualified," "unqualified," or "needs calibration." These data are the core outputs of the calibration process, which can be automatically calculated and generated by the calibration system and stored in the database.

[0081] In the steady-state determination process, the speed fluctuation rate refers to the ratio of the standard deviation of the speed sampling value per unit time to the current average speed during torque loading of the servo motor, reflecting the smoothness of the loading process. The torque fluctuation rate refers to the ratio of the standard deviation of the torque sampling value per unit time to the current average torque of the torque signal output by the torque sensor, reflecting the stability of the applied torque. The sampling curve refers to a continuous record of these fluctuation rates changing over time. Storing this curve data can be used to analyze the dynamic characteristics and stability of the torque loading process. For example, by plotting curves to visually demonstrate the torque from loading to stabilization, or by evaluating the performance of the loading control algorithm through data analysis. This data can be acquired in real time by the servo drive unit and torque sensor, and calculated and recorded by the stability determination unit.

[0082] Through the above technical solution, this application clarifies the various types of verification data that need to be stored during the automatic verification process of digital torque wrenches, thus effectively solving the problem of missing key information caused by only storing verification data in a general way. Specifically, storing the identification information and model specifications of the digital torque wrench under test enables the verification data to be accurately traced to a specific tested object, avoiding data confusion and greatly facilitating subsequent data retrieval, management, and result verification. At the same time, recording environmental parameters such as ambient temperature, humidity, and atmospheric pressure during the verification process complies with metrological verification specifications, providing important environmental evidence for subsequent verification of the validity of the verification results and helping to identify the potential impact of environmental factors on the accuracy of the verification. In addition, storing in detail the target torque value, the actual standard torque value, the wrench display value, the error value, and the verification conclusion at each verification point constitutes the core data chain of the verification process, enabling the entire verification process to be fully reproduced and verified, providing a solid foundation for metrological management and data analysis. Furthermore, storing the speed fluctuation rate and torque fluctuation rate sampling curves during the steady-state determination process enables retrospective analysis of the stabilization process after torque loading, verifying the compliance of the steady-state determination and thus ensuring the reliability of the verification results. In summary, this application, through a standardized and refined data storage strategy, significantly improves the traceability, data reproducibility, and standardization level of the automatic verification process of digital torque wrenches, ensuring the accuracy and authority of the verification results.

[0083] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. An automatic calibration method for a digital torque wrench, characterized in that, Includes the following steps: In the torque loading step, torque is applied to the digital torque wrench by a servo motor driving the transmission mechanism. In the image acquisition step, after the torque value reaches a stable state, the display image of the digital torque wrench is acquired through a camera device; The standard data synchronization procedure uses the data sampling time of the standard torque tester as the time reference to synchronously acquire the corresponding standard torque value. The error calculation step involves aligning the wrench display value obtained from image recognition processing with the standard torque value over time to calculate the error value. The verification and judgment step outputs a verification conclusion based on the comparison result between the error value and the preset allowable error range; The data storage step involves storing the verification conclusions and related verification data into a database.

2. The automatic calibration method for digital torque wrenches according to claim 1, characterized in that, The image acquisition step is preceded by: The stable state determination step involves real-time monitoring of the servo motor's speed signal and the torque signal output by the torque sensor to calculate the speed fluctuation rate and torque fluctuation rate. When both the speed fluctuation rate and torque fluctuation rate are less than the corresponding preset threshold, the torque value is determined to have entered a stable state, triggering the execution of the image acquisition step. Wherein, the speed fluctuation rate is the ratio of the standard deviation of the speed sampling value per unit time to the current speed mean, and the torque fluctuation rate is the ratio of the standard deviation of the torque sampling value per unit time to the current torque mean.

3. The automatic calibration method for a digital torque wrench according to claim 2, characterized in that, The torque loading step specifically includes: During the acceleration phase, the servo motor is accelerated by a first preset acceleration value; During the deceleration switching phase, when the current torque value is detected to reach a preset ratio threshold of the preset target torque, the deceleration control is switched to a second preset acceleration value, wherein the second preset acceleration value is less than the first preset acceleration value. During the fine-tuning phase, once the torque value enters the preset approximation range of the preset target torque, the servo motor enters a low-speed closed-loop adjustment mode, adjusting the output torque through a proportional-integral-derivative control algorithm until the target torque value is reached.

4. The automatic calibration method for digital torque wrenches according to claim 1, characterized in that, When the digital torque wrench uses a segment-type LCD screen, the image acquisition steps include: The adaptive frame sampling sub-step sets the acquisition interval time for multiple consecutive frames of images according to the preset segment code liquid crystal refresh cycle, so that the acquisition interval time between two adjacent frames is less than the refresh cycle. The inter-frame complementary splicing sub-step involves performing overlap detection on the segment code display areas in the consecutive multi-frame images, extracting complete digital segments from each frame, and combining the digital segments into complete display values ​​using preset digital splicing rules.

5. The automatic calibration method for a digital torque wrench according to claim 4, characterized in that, The inter-frame complementary splicing sub-step specifically includes: Each frame of the image is binarized to obtain the pixel distribution map of the segment code display area; Based on the preset seven-segment digital partition template, identify whether each digital segment in each frame is lit up, and generate a digital segment lighting matrix for each frame. Perform a bitwise OR operation on the digital segment lighting matrices of adjacent frames to generate a spliced ​​digital segment lighting matrix; The complete display value is restored by lighting up the spliced ​​digital segments into a matrix.

6. The automatic calibration method for a digital torque wrench according to claim 1, characterized in that, The standard data synchronization steps include: While the standard torque tester completes one torque sampling, it records the timestamp and sampled value of this sampling. After the image acquisition is completed, the standard torque value with the closest timestamp is retrieved from the stored standard torque sampling data according to the acquisition time corresponding to the image. The retrieved standard torque value is used as the standard torque value corresponding to the image acquisition time.

7. The automatic calibration method for a digital torque wrench according to any one of claims 1 to 6, characterized in that, The verification data stored in the data storage step includes: The identification information and model specifications of the digital torque wrench under inspection; Ambient temperature, humidity, and atmospheric pressure during the verification process; The target torque value, the actual standard torque value under load, the wrench display value, the error value, and the verification conclusion at each calibration point; The sampling curves of rotational speed fluctuation rate and torque fluctuation rate during the steady state determination process.

8. An automatic calibration system for digital torque wrenches, comprising a main control console, characterized in that, Also includes: A servo drive unit, which is communicatively connected to the main control console, is used to receive torque control commands and output torque drive signals. A transmission mechanism is connected between the output end of the servo drive unit and the digital torque wrench to be tested, and is used to convert the torque drive signal into torque output; The standard torque tester is connected to the main control console via a physical signal interface, and is used to collect the standard torque value during the loading process in real time and output the sampling timestamp synchronously. The camera acquisition module is located in front of the display panel of the digital torque wrench and is used to perform image acquisition after receiving a stable trigger signal; The stability determination unit is connected to the servo drive unit and the standard torque tester, respectively, and is used to calculate the speed fluctuation rate and torque fluctuation rate based on the real-time acquired speed signal and torque signal, and output the stability trigger signal when the speed fluctuation rate and torque fluctuation rate are both less than a preset threshold.

9. The automatic calibration system for digital torque wrenches according to claim 8, characterized in that, The camera acquisition module includes: A camera device is positioned in front of the display panel of the digital torque wrench; A frame sampling controller, connected to the camera device, is used to set the acquisition interval of multiple consecutive frames of images according to the refresh cycle of the segment-type LCD. The image processing unit, connected to the frame sampling controller, is used to perform inter-frame complementary stitching processing on the continuous multi-frame images and output the complete wrench display value.

10. The automatic calibration system for digital torque wrenches according to claim 8, characterized in that, The main control console has a built-in coordination control program, which is used for: The servo drive unit is controlled to output torque according to a preset segmented variable acceleration strategy; The system receives and processes the speed fluctuation rate and torque fluctuation rate data output by the stability determination unit in real time. Upon receiving the stable trigger signal, the camera acquisition module is controlled to perform image acquisition and synchronously record the acquisition timestamp; The corresponding standard torque value is extracted from the sampling data of the standard torque tester based on the acquisition timestamp; The image data output by the image acquisition module is processed by calling an image recognition algorithm to obtain the wrench display value and calculate the error with the standard torque value.

Images