Multi-sensor fusion calligraphy force detection intelligent writing practice pen

Abstract

The application relates to the technical field of calligraphy pens, and discloses a multi-sensing fusion calligraphy force detection intelligent calligraphy pen and a detection method, which comprises a pen shell, a pen core, a pen tip pressure sensing assembly, a three-finger partitioned grip force sensing assembly, a six-axis IMU inertial sensing module, a synchronous sampling MCU main control chip, a vibration feedback motor, a Bluetooth communication module and a power supply battery. The MCU synchronously collects multiple sensing data such as pen tip pressure, three-finger grip force, pen rod posture and pen operation acceleration by using a unified clock, compares the multiple sensing data with a standard stroke feature library after preprocessing and feature extraction, judges whether the grip pen force, the pen method strength, the pen rod posture and the pen operation rhythm are abnormal, drives the vibration motor to instantaneously feed back, and uploads the data to an external terminal through Bluetooth. The application realizes multi-dimensional synchronous collection and comprehensive evaluation in a writing process, and is suitable for standardized teaching and training of hard pen calligraphy.

Description

Technical Field

[0001] This invention relates to the field of calligraphy practice pen technology, specifically to a multi-sensor fusion intelligent calligraphy practice pen for detecting the force exertion in calligraphy. Background Technology

[0002] Calligraphy instruction is an important component of basic education, and proper writing posture and penmanship training play a fundamental role in cultivating children's writing habits. However, traditional calligraphy teaching has long relied on teachers' manual observation and experience-based judgment, resulting in a highly subjective teaching process with low standardization, making it difficult to achieve large-scale, standardized calligraphy instruction. With the development of educational informatization, various intelligent calligraphy practice pens and calligraphy teaching aids have emerged in recent years, aiming to collect and analyze data on the writing process through sensor technology.

[0003] The single sensing dimension cannot independently distinguish the gripping force of the thumb, index finger, and middle finger, making it impossible to determine if the pen grip force is unbalanced; the data from multiple sensors are not synchronized, with the timing of data such as pen tip pressure, finger grip force, pen posture, and pen acceleration being out of sync, making it impossible to fully reconstruct the writing dynamics process; the evaluation logic is simple and crude, relying solely on a single pressure threshold to determine the weight of the stroke, unable to distinguish the unique force curves and pause rhythms of different strokes, and unable to quantify the correctness of the penmanship; it lacks a real-time embedded feedback mechanism, and data can only be viewed on a mobile phone after writing, without timely reminders of deviations during writing, resulting in low correction efficiency.

[0004] Existing intelligent calligraphy pens and related calligraphy teaching equipment have significant shortcomings in terms of sensing dimensions, data synchronization, evaluation logic, and real-time feedback. They lack the support of multimodal fusion algorithms and cannot achieve synchronous collection and comprehensive evaluation of multiple dimensions such as finger grip force, pen stroke technique, pen tilt, and pen rhythm during the writing process, making it difficult to meet the actual needs of standardization and intelligentization in calligraphy teaching. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a multi-sensor fusion calligraphy force detection intelligent calligraphy pen, which solves the problems mentioned in the background above, such as the significant deficiencies of existing intelligent calligraphy pens in terms of sensing dimensions, data synchronization, evaluation logic, and real-time feedback, the lack of multi-modal fusion algorithm support, and the inability to meet the actual needs of standardized and intelligent calligraphy teaching.

[0006] This invention provides the following technical solution: a multi-sensor fusion calligraphy force detection intelligent calligraphy practice pen, comprising: The hollow pen casing is divided longitudinally into the pen tip area, the grip area, and the pen barrel area. The pen refill is replaceably disposed within the pen casing, and the pen tip of the pen refill extends from the pen tip area; A pen tip pressure sensing component is disposed inside the pen casing and is used to detect the writing pressure of the pen tip; The three-finger grip pressure sensing component is located in the grip area of ​​the pen shell and includes three independent pressure sensing plates. The three pressure sensing plates are arranged at intervals along the circumference of the pen shell and correspond to the contact areas of the thumb, index finger and middle finger respectively when gripping the pen. It is used to independently collect the real-time grip pressure of the three fingers. The six-axis IMU inertial sensing module integrates a three-axis accelerometer and a three-axis gyroscope to collect the pitch angle, yaw angle, roll angle of the pen shell, as well as the instantaneous acceleration and angular velocity of the pen movement in real time. The synchronous sampling MCU main control chip is electrically connected to the pen tip pressure sensing component, the three-finger zone grip force sensing component and the six-axis IMU inertial sensing module, respectively, for synchronous sampling with a unified clock and adding a unified timestamp to the collected sensing data; A vibration feedback motor is installed inside the pen shell and electrically connected to the synchronous sampling MCU main control chip. It is used to generate a vibration reminder when the synchronous sampling MCU main control chip determines that there is an abnormality in the writing. A Bluetooth communication module, located inside the pen casing, is used to transmit processed data to an external terminal; A power supply battery is located inside the pen casing to power various electrical devices.

[0007] Preferably, the pen tip pressure sensing component is disposed within the pen barrel area, and includes: A strain gauge pressure sensor is located inside the pen barrel area, with the tail of the pen refill abutting against the strain gauge pressure sensor. The position of the strain gauge pressure sensor corresponds to the position of the tail of the pen refill. A return spring is sleeved on the surface of the pen refill, with one end of the return spring connected to the limiting part on the pen refill and the other end connected to the inner wall of the pen barrel area, for resetting the pen refill when the pen is lifted.

[0008] Preferably, all three pressure sensing sheets are FSR flexible pressure sensing sheets, and are evenly distributed at 120° intervals along the circumference of the pen shell. The pen grip area is provided with a soft rubber anti-slip liner. The three FSR flexible pressure sensor plates are connected to the pen grip area and are located in the interlayer between the soft rubber anti-slip liner and the pen grip area.

[0009] Preferably, a battery compartment is provided at the tail end of the pen barrel area, the power supply battery is disposed in the battery compartment, the battery compartment is provided with a charging interface, and the charging interface is electrically connected to the power supply battery.

[0010] Preferably, it also includes a flexible main control PCB, which is attached to the inner wall of the pen shell; The six-axis IMU inertial sensing module and the synchronous sampling MCU main control chip are both soldered on the flexible main control PCB. The synchronous sampling MCU main control chip is electrically connected to the strain gauge pressure sensor and the three FSR flexible pressure sensing plates respectively.

[0011] A detection method for a multi-sensor fusion intelligent calligraphy pen for detecting calligraphy force application, characterized by the following steps: S1. Synchronous Multi-Source Sensor Data Acquisition: The synchronous sampling MCU main control chip synchronously acquires four types of timing data streams at a fixed sampling period: Data stream A is the independent grip pressure sequence of the thumb, index finger, and middle finger acquired by the three-finger partition grip force sensing component; Data stream B is the pen tip writing pressure timing sequence acquired by the pen tip pressure sensing component; Data stream C1 is the pen barrel three-axis tilt angle sequence acquired by the six-axis IMU inertial sensing module; Data stream C2 is the pen barrel three-axis acceleration and three-axis angular velocity sequence acquired by the six-axis IMU inertial sensing module; and timestamps are uniformly added to all data streams to eliminate sensor sampling delay misalignment; S2. Sensor raw signal preprocessing: The synchronous sampling MCU main control chip performs sliding mean filtering to reduce noise in the pressure signals of data stream A and data stream B, and performs first-order complementary filtering on the attitude data in data stream C1 and data stream C2 to eliminate gyroscope drift and output a stable pen barrel three-dimensional attitude angle; and automatically splits each stroke into independent time-series data blocks with the pen tip pressure signal being greater than the preset contact threshold as the writing segment and the pressure returning to zero as the pen lifting segment; S3. Multimodal writing feature extraction: Synchronously sample the independent time-series data blocks of each stroke from the MCU main control chip to extract a four-dimensional feature set, including: pen grip force features, pen tip stroke force features, posture geometric features, and pen movement rhythm dynamic features. S4. Standard stroke feature library matching and comparison: The synchronous sampling MCU main control chip compares the extracted real-time features with the pre-stored standard stroke feature threshold range to determine whether each feature exceeds the threshold range. S5. Hierarchical real-time feedback and data output: When the synchronous sampling MCU main control chip determines that any feature exceeds the standard, it drives the vibration feedback motor to generate vibrations of different frequencies to distinguish different abnormality types; at the same time, the complete features of a single stroke and the abnormality determination results are uploaded to an external terminal via Bluetooth communication module.

[0012] Preferably, in step S2: the filtering and noise reduction of the pen tip pressure timing sequence and the three-finger independent grip pressure sequence is a sliding mean filter, wherein the sliding mean filter of the three-finger independent grip pressure sequence is used to eliminate the jitter noise of the FSR flexible pressure sensor sheet; the filtering of the pen barrel three-axis tilt angle sequence and the three-axis acceleration and three-axis angular velocity sequences is a first-order complementary filter.

[0013] Preferably, in step S3: The pen grip force characteristics include the average grip force of the three fingers, the pressure difference of the three fingers, and the fluctuation range of finger force, which are used to determine abnormal single-finger force and imbalance of three-finger force. The pen tip's pen technique strength characteristics include the starting pressure of the stroke, the average pressure in the middle section, the peak pressure at the end of the stroke, and the slope of the pressure change curve, which are used to distinguish incorrect pen techniques. The posture geometry features include the pen's pitch and tilt angle range and the left and right skewing amplitude throughout the stroke, which are used to determine incorrect pen grip posture. The dynamic characteristics of pen stroke rhythm include the total writing time of each stroke, peak acceleration, difference between pen lifting and pen falling speeds, and frequency of pauses, which are used to determine abnormal writing rhythms.

[0014] Preferably, in step S4: the pre-stored standard stroke feature threshold range includes the standard feature threshold of the basic strokes of the national standard hard pen, and the basic strokes of the national standard hard pen include at least horizontal, vertical, left-falling, right-falling, dot, hook, and fold; during comparison, if the grip force feature exceeds the threshold, it is determined that the grip force is abnormal; if the pressure curve deviates from the standard range, it is determined that the stroke force is not standardized; if the pen tilt angle exceeds the standard, it is determined that the grip posture is incorrect; if the pen movement time or acceleration exceeds the standard range, it is determined that the writing rhythm is disordered.

[0015] Preferably, the process also includes step S6, sleep energy-saving control: when there is no pressure signal from the pen tip for 30 consecutive seconds, the synchronous sampling MCU main control chip reduces the sensor sampling frequency and enters a low-power mode.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention integrates three-finger zone independent grip force sensing, pen tip pressure sensing, and a six-axis IMU into a conventional pen barrel. Three FSR flexible pressure sensing plates are arranged along the circumference of the pen shell corresponding to the contact areas of the thumb, index finger, and middle finger. They are sampled synchronously with a unified clock. A single pen can completely capture all dynamic parameters of the writing process. The hardware structure is compact and does not affect the normal writing feel.

[0017] 2. This invention achieves a comprehensive and synchronous evaluation of the balance of finger force, the curve of pen stroke, the tilt of the pen, and the rhythm of pen stroke by precisely aligning data from various heterogeneous sensors in the time dimension and correlating them in the spatial dimension. This is achieved through precise alignment of data from multiple heterogeneous sensors, such as pen tip pressure, three-finger grip force, pen posture, and pen stroke acceleration, rather than through simple threshold comparisons in a single dimension. This allows for the fine differentiation of different strokes, such as horizontal, vertical, left-falling, right-falling, hook, and fold strokes, and the determination of correct or incorrect pen strokes.

[0018] 3. In this invention, by setting up a vibration feedback motor, the MCU can detect abnormalities in real time during the writing process and drive the vibration feedback motor to provide vibration warnings, thereby achieving instant error correction. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 This is a flowchart illustrating the overall process of the detection method of the present invention.

[0020] In the diagram: 1. Pen shell; 2. Pen refill; 31. Strain gauge pressure sensor; 32. Reset spring; 4. Pressure sensor plate; 5. Six-axis IMU inertial sensing module; 6. Synchronous sampling MCU main control chip; 7. Vibration feedback motor; 8. Bluetooth communication module; 9. Power supply battery; 10. Battery compartment; 11. Charging interface; 12. Flexible main control PCB. Detailed Implementation

[0021] 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.

[0022] Please see Figure 1 A multi-sensor fusion calligraphy force detection smart calligraphy pen includes a hollow pen shell 1 and a pen core 2, which are divided into a pen tip area, a pen grip area and a pen barrel area along the longitudinal direction. The whole pen is cylindrical and its size is similar to that of an ordinary writing pen, which makes it easy for users to hold and write without affecting the normal writing feel. The pen core 2 is replaceably set in the pen shell 1, and the pen tip of the pen core 2 extends from the pen tip area. A pen tip pressure sensing component is installed inside the pen casing 1 to detect the writing pressure of the pen tip; In a specific embodiment, the pen tip pressure sensing component is disposed inside the pen barrel area, including a strain gauge pressure sensor 31 and a return spring 32. The strain gauge pressure sensor 31 is disposed inside the pen barrel area, and the tail of the pen refill 2 abuts against the strain gauge pressure sensor 31. The position of the strain gauge pressure sensor 31 corresponds to the position of the tail of the pen refill 2. The return spring 32 is sleeved on the surface of the pen refill 2, and one end of the return spring 32 is connected to a limiting part on the pen refill 2, such as a protruding ring or a retaining plate of the pen refill, and the other end is connected to the inner wall of the pen barrel area. It is used to reset the pen refill 2 when the pen is lifted. When writing, the pen tip is pressed and drives the pen refill 2 to move slightly backward along the axis. The tail of the pen refill 2 pushes the strain gauge pressure sensor 31, and the sensor outputs a timing electrical signal corresponding to the writing pressure. When the pen is lifted, the return spring 32 pushes the pen refill 2 forward to reset by relying on the elastic restoring force. The tail of the pen refill 2 is removed from the push against the sensor, the pressure signal returns to zero, and the output analog voltage signal is amplified and sent to the analog-to-digital conversion channel of the synchronous sampling MCU main control chip 6.

[0023] The three-finger grip pressure sensing component is set in the grip area of ​​the pen shell 1. It includes three independent pressure sensing plates 4. The three pressure sensing plates 4 are arranged at intervals along the circumference of the pen shell 1 and correspond to the contact areas of the thumb, index finger and middle finger respectively in the grip state. It is used to independently collect the real-time grip pressure of the three fingers. In a specific embodiment, all three pressure sensor pieces 4 are FSR flexible pressure sensor pieces 4, which output pressure analog signals from 0 to 1024 levels. Each sensor piece, together with a pull-up resistor, forms a voltage divider circuit to convert resistance changes into analog voltage signal output. They are evenly distributed at 120° intervals along the circumference of the pen shell 1. The pen grip area of ​​the pen shell 1 is provided with a soft rubber anti-slip liner. The three FSR flexible pressure sensor pieces 4 are connected to the shell of the pen grip area and are located in the sandwich between the soft rubber anti-slip liner and the shell of the pen grip area. The soft rubber anti-slip liner is made of silicone or thermoplastic polyurethane material to ensure the user's pen grip comfort and anti-slip performance. The six-axis IMU inertial sensing module 5 integrates a three-axis accelerometer and a three-axis gyroscope to collect the pitch angle, yaw angle, roll angle, and instantaneous acceleration and angular velocity of the pen shell 1 in real time. The preferred model is MPU6050 or ICM-42605. It has a built-in 16-bit analog-to-digital converter, an acceleration range of ±16 times the gravitational acceleration, and an angular velocity range of ±2000 degrees per second. It communicates with the synchronous sampling MCU main control chip 6 through the integrated circuit bus interface. The synchronous sampling MCU main control chip 6 is electrically connected to the pen tip pressure sensing component, the three-finger grip force sensing component, and the six-axis IMU inertial sensing module 5, respectively. It is used for synchronous sampling with a unified clock and adding a unified timestamp to the acquired sensor data. The synchronous sampling MCU main control chip 6 uses a low-power processor with an ARM Cortex-M4 or M7 core, such as the STM32L4 series, and has a built-in multi-channel 12-bit analog-to-digital converter and hardware timer. The synchronous sampling MCU main control chip 6 synchronously acquires analog or digital signals from all sensors with a fixed sampling period, preferably 10 milliseconds (100 Hz sampling rate), and adds a unified hardware timestamp to all data to achieve millisecond-level timing alignment. The synchronous sampling MCU main control chip 6 integrates signal preprocessing, feature extraction, and anomaly detection algorithm modules. In one specific embodiment, a flexible main control PCB 12 is installed on the inner wall of the pen shell 1. A six-axis IMU inertial sensing module 5, a synchronous sampling MCU main control chip 6 and a Bluetooth communication module 8 are soldered on the flexible main control PCB 12. The synchronous sampling MCU main control chip 6 is electrically connected to the strain gauge pressure sensor 31 and three FSR flexible pressure sensing sheets 4 respectively. The vibration feedback motor 7 is located inside the pen shell 1 and is electrically connected to the synchronous sampling MCU main control chip. It is used to generate vibration reminders when the synchronous sampling MCU main control chip determines that there is an abnormality in writing. In one specific embodiment, the vibration feedback motor 7 is located in the pen barrel area and is directly electrically connected to the input / output port of the synchronous sampling MCU main control chip 6. The vibration feedback motor 7 is a miniature flat vibration motor, such as a 1020 or 1030 type linear vibration motor, with a diameter of about 10 mm, a thickness of about 3 mm, an operating voltage of 3 volts, and a response time of less than 50 milliseconds. When the synchronous sampling MCU main control chip 6 determines an abnormal writing condition, it drives the vibration feedback motor 7 to generate vibrations of different frequencies to distinguish different types of writing errors. Bluetooth communication module 8, located inside pen shell 1, is used to transmit processed data to an external terminal; In one specific embodiment, the Bluetooth communication module 8 is soldered onto the flexible main control PCB 12 and electrically connected to the synchronous sampling MCU main control chip 6. The Bluetooth communication module 8 uses a low-power Bluetooth 5.0 chip such as the nRF52832 or CC2640, supporting wireless data transmission with external terminals such as mobile phones and tablets. The Bluetooth communication module 8 uploads the multimodal fusion data, feature extraction results, and anomaly detection results processed by the synchronous sampling MCU main control chip 6 to the external terminal. And the power supply battery 9, which is located inside the pen casing 1, is used to power various electrical devices; Specifically, the power supply battery 9 is located in the battery compartment 10 at the end of the pen body area, providing power to all sensing, control, and communication components. The power supply battery 9 is preferably a rechargeable lithium polymer battery with a capacity of 100 mAh to 200 mAh and an operating voltage of 3.7 volts. The battery compartment 10 is located at the end of the pen body area, and the power supply battery 9 is located in the battery compartment 10. The battery compartment 10 is provided with a charging interface 11, which is electrically connected to the power supply battery 9. The charging interface 11 is preferably a Type-C interface, supporting 5V / 1A charging.

[0024] Please see Figure 2 A detection method for a multi-sensor fusion intelligent calligraphy practice pen for detecting calligraphy force includes the following steps: S1. Synchronous multi-source sensor data acquisition: When the user holds the pen and writes, the pen tip contacts the paper, triggering the system to enter full-power sampling mode. The MCU main control chip simultaneously collects four types of timing data streams at a fixed sampling period of 10 milliseconds per sampling: Data stream A: Independent grip pressure sequences of the thumb, index finger, and middle finger are collected by three flexible piezoresistive thin-film pressure sensor sheets 4, and each channel outputs 0 to 1024 levels of pressure analog signals; Data stream B: The timing sequence of pen tip writing pressure acquired by strain gauge pressure sensor 31, outputting an analog voltage signal corresponding to the writing pressure; Data stream C1: The three-axis tilt angle sequence of pen shell 1, including pitch angle, yaw angle, and roll angle, acquired by the six-axis inertial measurement unit inertial sensing module 5; Data stream C2: The three-axis acceleration sequence and three-axis angular velocity sequence of the pen shell 1 acquired by the six-axis inertial measurement unit inertial sensing module 5; The synchronous sampling MCU main control chip 6 uses a built-in hardware timer to uniformly add millisecond-level timestamps to all data streams, eliminating data misalignment caused by differences in sampling start time and signal transmission delay among various sensors, and ensuring accurate correspondence of various data in the time dimension in subsequent multimodal fusion analysis; S2. Preprocessing of raw sensor signals: The synchronous sampling MCU main control chip performs sliding mean filtering to reduce noise in the pressure signals of data stream A (three-finger grip pressure sequence) and data stream B (pen tip writing pressure sequence). The sliding window size is preferably 5 sampling points, corresponding to a 50-millisecond time window, which effectively eliminates high-frequency noise caused by factors such as contact jitter and circuit noise in the flexible piezoresistive thin film pressure sensor 4, and outputs a smooth pressure curve. Inertial Measurement Unit Attitude Compensation: First-order complementary filtering is performed on the attitude data in the three-axis tilt angle of data stream C1 and the three-axis acceleration and three-axis angular velocity of data stream C2. The low-frequency stability of the accelerometer is used to correct the long-term integral drift of the gyroscope, eliminating the angle drift of the gyroscope caused by factors such as temperature changes and integral accumulation error, and outputting a stable penholder three-dimensional attitude angle. The optimal complementary filtering coefficients are α=0.98 gyroscope weight and β=0.02 accelerometer weight, achieving a good balance between dynamic response and static stability. Writing action segmentation: The writing segment is determined by the pen tip pressure signal being greater than the preset contact threshold, preferably greater than 5 grams, and the pen lifting segment is determined by the pressure returning to zero. The MCU main control chip 6, which is synchronously sampled, automatically identifies the start and end times of each stroke based on the temporal changes of the pen tip pressure signal, and splits the continuous writing data stream into independent temporal data blocks for each stroke, which are then used for subsequent stroke-by-stroke feature extraction. S3. Multimodal writing feature extraction: The MCU main control chip synchronously samples the independent time-series data blocks of each stroke and extracts a four-dimensional feature set, including: pen grip force features, pen tip stroke force features, posture geometric features, and pen movement rhythm dynamic features. Pen grip force characteristics: The independent grip force sequence of three fingers in data stream A is calculated to extract the average grip force of the three fingers, the pressure difference between each pair of the three fingers, and the fluctuation range of the grip force of each finger throughout the stroke period. By the pressure difference and fluctuation range of the three fingers, it can be determined whether there are pen grip problems such as excessive or insufficient force of a single finger or imbalance of force of the three fingers.

[0025] Pen tip pressure characteristics: Analyze the pen tip pressure time series curve in data stream B to extract the pressure value at the beginning of the stroke, the average pressure value in the middle of the stroke, the pressure peak at the end of the stroke, and the slope characteristics of the entire pressure change curve. The combination of the above features can be used to distinguish different pen techniques such as light start, heavy pause, full pressure, and no pause. Posture geometric features: Analyze the three-dimensional tilt angle sequence of the pen in data stream C1, extract the range of changes in the pen's pitch and tilt angles throughout the stroke, i.e. the maximum and minimum values, as well as the magnitude of the pen's left and right tilt. Through the above geometric features, it can be determined whether there are posture errors such as holding the pen too low or excessive tilting of the pen.

[0026] The dynamic characteristics of pen stroke rhythm: Analyze the triaxial acceleration and angular velocity sequences in data stream C2, and combine them with timestamp information to extract the total writing time, peak acceleration, speed difference between lifting and lowering the pen, and frequency of pauses during the writing process. These characteristics can be used to determine problems such as writing too fast, sluggish writing, and irregular and arrhythmic writing. S4. Standard stroke feature library matching and comparison: The synchronous sampling MCU main control chip compares the extracted real-time features with the pre-stored standard stroke feature threshold range to determine whether each feature exceeds the threshold range. The data is compared item by item with the standard stroke feature threshold ranges pre-stored in local storage or on the mobile device. The standard stroke feature library covers the basic strokes of the national standard hard pen, including at least seven basic stroke categories: horizontal, vertical, left-falling, right-falling, dot, hook, and turn. Each stroke category has a standard feature threshold range calibrated by calligraphy experts, specifically including: Horizontal stroke: The standard penholder tilt angle is 8 to 15 degrees. The stroke starts with a slightly heavier pressure, the pressure is kept moderate in the middle, and the stroke ends with a slight pause. Vertical strokes: The standard penholder tilt angle is less than or equal to 15 degrees, the pressure at the end of the stroke reaches its peak, and the pressure difference between the three fingers is less than 200 levels; The stroke with a left-falling stroke: The standard starting pressure value is relatively high, such as 0.8 Newtons, and the pressure gradually decreases to a low value, such as 0.3 Newtons, at the end of the stroke; The standard curve for the downward stroke is to start lightly, gradually increase pressure in the middle, and end with a heavy pause.

[0027] The comparison and judgment rules are as follows: If the three-finger pressure difference or single-finger pressure value exceeds the threshold range, it is judged as abnormal pen grip force; if the pen tip pressure curve deviates from the standard pressure range of the corresponding stroke, such as the starting pressure, middle pressure, and peak pressure of the pause, it is judged as non-standard stroke force; if the pen tilt angle, pitch angle, or yaw angle exceeds the standard threshold, it is judged as incorrect pen grip posture; if the pen stroke duration or acceleration peak exceeds the standard range, it is judged as disordered writing rhythm. S5. Hierarchical real-time feedback and data output: When the synchronous sampling MCU main control chip determines that any feature exceeds the standard, it drives the vibration feedback motor 7 to generate vibrations of different frequencies to distinguish different abnormal types; at the same time, the complete features of a single stroke and the abnormal judgment results are uploaded to an external terminal through the Bluetooth communication module 8. Embedded real-time feedback: When the synchronous sampling MCU main control chip 6 determines that any feature exceeds the standard in step S4, it immediately drives the vibration feedback motor 7 to generate vibrations of the corresponding frequency. The vibration encoding scheme is as follows: Abnormal pen grip force, imbalance of three fingers or abnormal force of one finger: two consecutive short vibrations lasting 0.2 seconds, with an interval of 0.3 seconds; Incorrect pen pressure and improper penmanship: Three consecutive short vibrations lasting 0.15 seconds, with an interval of 0.2 seconds; Incorrect pen grip posture, excessive pen tilt: one long vibration lasting 0.8 seconds; Writing rhythm is disordered, too fast, dragging, or has abnormal pauses: short vibration 1 vibration 0.3 seconds.

[0028] Users can instantly perceive vibration alerts during the writing process and immediately adjust their grip strength, pen posture, or writing rhythm to achieve real-time error correction.

[0029] Data Packaging Output: After each stroke is detected, the synchronous sampling MCU main control chip 6 packages the complete four-dimensional feature data of a single stroke, including grip strength features, pen force features, posture geometric features, pen rhythm features, and anomaly judgment results of each feature, and uploads them to the mobile application via Bluetooth communication module 8. After receiving the data, the mobile application summarizes the detection results of all strokes and automatically generates a comprehensive handwriting score, statistical analysis reports of various errors, and targeted correction suggestions. S6. Sleep Energy Saving Control: When the synchronous sampling MCU main control chip 6 detects no pressure signal from the pen tip for 30 consecutive seconds, and the pen tip pressure value is lower than the contact threshold, it automatically executes a low-power management strategy: reducing the sensor sampling frequency from 100 Hz to 1 Hz, switching the high-performance mode of the six-axis inertial measurement unit inertial sensing module 5 to low-power mode, reducing the broadcast frequency of the Bluetooth communication module 8, and retaining only the low-power wake-up monitoring function of the pen tip pressure sensor. After the system enters low-power sleep mode, the overall power consumption drops to below 50 microamps. When the user puts pen down again and the pen tip pressure signal exceeds the contact threshold, the synchronous sampling MCU main control chip 6 is immediately woken up, and the system recovers to full-power sampling mode within 50 milliseconds, allowing the user to continue writing without any additional operation.

[0030] The user holds a pen and writes the Chinese character "Yong" on ordinary paper. At the moment when the pen tip touches the paper surface, the pen tip pressure signal triggers the system to wake up from the sleep mode, and the synchronous sampling MCU main control chip 6 synchronously collects four channels of sensing data at a period of 10 milliseconds. When writing the first stroke "dot" of the character "Yong", the pen tip pressure curve presents the typical characteristics of "light start - gradually increasing pressure - pause and end"; the three-finger grip force sensor synchronously records the real-time clamping forces of the thumb, index finger and middle finger; the inertial measurement unit records that the pitch angle of the pen barrel is about 12 degrees and the yaw angle is about 0 degrees. The synchronous sampling MCU main control chip 6 extracts the features of this stroke, such as the starting pressure of about level 50, the average pressure in the middle section of about level 300, and the peak pressure at the pause of about level 600. After comparing with the standard "dot" stroke feature threshold, all features are qualified, and no vibration reminder is issued.

[0031] When writing the second stroke "horizontal", if the pitch angle of the user's pen barrel reaches 28 degrees while the standard threshold is less than or equal to 15 degrees, the synchronous sampling MCU main control chip 6 determines that the posture is wrong, and immediately drives the motor to vibrate for 0.8 seconds for reminder. After perceiving the vibration, the user adjusts the pen barrel angle to a reasonable range and continues writing. When writing the "right-falling" stroke, if the user exerts uniform pressure throughout the process without a heavy pause at the end, the synchronous sampling MCU main control chip 6 determines that the stroke force application method is wrong, and drives the motor to vibrate 3 times in short bursts for reminder.

[0032] After completing all strokes of the character "Yong", the synchronous sampling MCU main control chip 6 uploads the four-dimensional feature data and error marks of all strokes to the mobile application through the Bluetooth communication module 8. The mobile application generates a calligraphy practice report: the total score is 78 points, the "dot" stroke is excellent, there is 1 posture error in the "horizontal" stroke which has been corrected immediately, the force application method of the "right-falling" stroke is not standard and needs more practice, and the report is attached with targeted correction suggestion animations and demonstration videos. Parents or teachers can remotely check the child's calligraphy practice progress and common error types through the mobile application, so as to realize standardized and data-based calligraphy teaching management.

[0033] After the intelligent calligraphy practice pen of the present invention is fully charged, the capacity of the power supply battery 9 is 150 mAh, and the endurance test is carried out by simulating the continuous calligraphy practice scenario of children.

[0034] Full-power working mode: under the continuous writing state, the synchronous sampling MCU main control chip 6 synchronously collects all sensor data at a sampling rate of 100 Hz, the Bluetooth communication module uploads data at the frequency of once per stroke, and the vibration feedback motor is triggered on demand. The measured working current of the whole device is about 25 mA, of which the synchronous sampling MCU main control chip 6 accounts for about 8 mA, the inertial measurement unit about 4 mA, the Bluetooth transmission about 10 mA, the flexible piezoresistive film pressure sensor sheet about 2 mA, and the vibration feedback motor in standby state about 1 mA, and the endurance for continuous writing is about 6 hours.

[0035] Low-power sleep mode: After 30 seconds of no writing signal, the system automatically enters low-power sleep mode: the sampling frequency drops to 1 Hz, the inertial measurement unit switches to low-power mode and the current drops to 0.5 mA, and the Bluetooth communication module enters deep sleep mode and the current drops to 2 μA. The overall standby current drops to approximately 50 μA, and the standby battery life can reach several months.

[0036] Intermittent writing scenario: Simulating the actual usage habit of children practicing calligraphy for 30 minutes and resting for 15 minutes each time, the system automatically enters low-power sleep mode during writing breaks. Comprehensive testing shows that a single charge can meet the daily calligraphy practice needs of about 4 to 5 days with 1 hour of calligraphy practice per day, far exceeding the battery life of similar products.

[0037] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-sensor fusion intelligent calligraphy practice pen for detecting calligraphy force, characterized in that, include: The hollow pen casing (1) is divided into the pen tip area, the pen grip area and the pen barrel area along the longitudinal direction; The pen refill (2) is replaceably disposed inside the pen casing (1), and the pen tip of the pen refill (2) extends from the pen tip area; A pen tip pressure sensing component is disposed inside the pen shell (1) and is used to detect the writing pressure of the pen tip; The three-finger grip force sensing component is set in the grip area of ​​the pen shell (1), including three independent pressure sensing plates (4). The three pressure sensing plates (4) are arranged at intervals along the circumference of the pen shell (1) and correspond to the contact areas of the thumb, index finger and middle finger respectively in the grip state, for independently collecting the real-time clamping pressure of the three fingers. The six-axis IMU inertial sensing module (5) integrates a three-axis accelerometer and a three-axis gyroscope, which are used to collect the pitch angle, yaw angle, roll angle and instantaneous acceleration and angular velocity of the pen shell (1) in real time. The synchronous sampling MCU main control chip (6) is electrically connected to the pen tip pressure sensing component, the three-finger partition grip force sensing component and the six-axis IMU inertial sensing module (5) respectively, for synchronous sampling with a unified clock and adding a unified timestamp to the collected sensing data; A vibration feedback motor (7) is installed inside the pen shell (1) and electrically connected to the synchronous sampling MCU main control chip (6). It is used to generate a vibration reminder when the synchronous sampling MCU main control chip (6) determines that the writing is abnormal. A Bluetooth communication module (8) is installed inside the pen shell (1) and is used to transmit the processed data to an external terminal; And a power supply battery (9) is disposed inside the pen casing (1) for supplying power to various electrical devices.

2. The intelligent calligraphy practice pen with multi-sensor fusion for detecting calligraphy force as described in claim 1, characterized in that, The pen tip pressure sensing component is disposed within the pen barrel area and includes: A strain gauge pressure sensor (31) is located inside the pen barrel area. The tail of the pen refill (2) abuts against the strain gauge pressure sensor (31). The position of the strain gauge pressure sensor (31) corresponds to the position of the tail of the pen refill (2). A reset spring (32) is sleeved on the surface of the pen refill (2), and one end of the reset spring (32) is connected to the limiting part on the pen refill (2), and the other end is connected to the inner wall of the pen barrel area, for resetting the pen refill (2) when the pen is lifted.

3. The intelligent calligraphy practice pen with multi-sensor fusion for detecting calligraphy force as described in claim 2, characterized in that, All three pressure sensor sheets (4) are FSR flexible pressure sensor sheets (4), and are evenly distributed at 120° intervals along the circumference of the pen shell (1). The pen holder (1) has a soft rubber anti-slip liner in the grip area. The three FSR flexible pressure sensor sheets (4) are connected to the pen holder housing and are located in the interlayer between the soft rubber anti-slip liner and the pen holder housing.

4. The intelligent calligraphy practice pen with multi-sensor fusion for detecting calligraphy force as described in claim 1, characterized in that, A battery compartment (10) is provided at the end of the pen barrel area. The power supply battery (9) is located in the battery compartment (10). The battery compartment (10) is provided with a charging interface (11). The charging interface (11) is electrically connected to the power supply battery (9).

5. The intelligent calligraphy practice pen with multi-sensor fusion for detecting calligraphy force according to claim 3, characterized in that, It also includes a flexible main control PCB (12), which is attached to the inner wall of the pen shell (1); The six-axis IMU inertial sensing module (5) and the synchronous sampling MCU main control chip (6) are both soldered on the flexible main control PCB (12). The synchronous sampling MCU main control chip (6) is electrically connected to the strain pressure sensor (31) and the three FSR flexible pressure sensing sheets (4).

6. A detection method for a multi-sensor fusion calligraphy force detection intelligent calligraphy pen according to any one of claims 1-5, characterized in that, Includes the following steps: S1. Synchronous multi-source sensor data acquisition: The MCU main control chip (6) synchronously collects four types of time-series data streams at a fixed sampling period: Data stream A is the independent grip pressure sequence of the thumb, index finger, and middle finger collected by the three-finger partition grip pressure sensing component; Data stream B is the pen tip writing pressure time sequence collected by the pen tip pressure sensing component; Data stream C1 is the pen barrel three-axis tilt angle sequence collected by the six-axis IMU inertial sensing module (5); Data stream C2 is the pen barrel three-axis acceleration and three-axis angular velocity sequence collected by the six-axis IMU inertial sensing module (5); and timestamps are uniformly added to all data streams to eliminate sensor sampling delay misalignment; S2. Sensor raw signal preprocessing: The synchronous sampling MCU main control chip (6) performs sliding mean filtering to reduce noise of the pressure signals in data stream A and data stream B, and performs first-order complementary filtering on the attitude data in data stream C1 and data stream C2 to eliminate gyroscope drift and output a stable pen barrel three-dimensional attitude angle; and automatically splits each stroke into independent time sequence data blocks with the pen tip pressure signal being greater than the preset contact threshold as the writing segment and the pressure returning to zero as the pen lifting segment; S3. Multimodal writing feature extraction: The MCU main control chip (6) extracts a four-dimensional feature set for each stroke's independent time-series data block, including: pen grip force features, pen tip pen technique force features, posture geometric features, and pen movement rhythm dynamic features. S4. Standard stroke feature library matching and comparison: The synchronous sampling MCU main control chip (6) compares the extracted real-time features with the pre-stored standard stroke feature threshold range to determine whether each feature exceeds the threshold range. S5. Hierarchical real-time feedback and data output: When the synchronous sampling MCU main control chip (6) determines that any feature exceeds the standard, it drives the vibration feedback motor (7) to generate vibrations of different frequencies to distinguish different abnormal types; at the same time, the complete features of a single stroke and the abnormal judgment results are uploaded to the external terminal through the Bluetooth communication module (8).

7. The detection method for a multi-sensor fusion calligraphy force detection intelligent calligraphy pen according to claim 6, characterized in that, In step S2: the filtering and noise reduction of the pen tip pressure timing sequence and the three-finger independent grip pressure sequence is sliding mean filtering, wherein the sliding mean filtering of the three-finger independent grip pressure sequence is used to eliminate the jitter noise of the FSR flexible pressure sensor (4); the filtering of the pen barrel three-axis tilt angle sequence and the three-axis acceleration and three-axis angular velocity sequence is first-order complementary filtering.

8. The detection method for a multi-sensor fusion calligraphy force detection intelligent calligraphy pen according to claim 6, characterized in that, In step S3: The pen grip force characteristics include the average grip force of the three fingers, the pressure difference of the three fingers, and the fluctuation range of finger force, which are used to determine abnormal single-finger force and imbalance of three-finger force. The pen tip's pen technique strength characteristics include the starting pressure of the stroke, the average pressure in the middle section, the peak pressure at the end of the stroke, and the slope of the pressure change curve, which are used to distinguish incorrect pen techniques. The posture geometry features include the pen's pitch and tilt angle range and the left and right skewing amplitude throughout the stroke, which are used to determine incorrect pen grip posture. The dynamic characteristics of pen stroke rhythm include the total writing time of each stroke, peak acceleration, difference between pen lifting and pen falling speeds, and frequency of pauses, which are used to determine abnormal writing rhythms.

9. The detection method for a multi-sensor fusion calligraphy force detection intelligent calligraphy pen according to claim 6, characterized in that, In step S4: the pre-stored standard stroke feature threshold range includes the standard feature threshold of the basic strokes of the national standard hard pen, and the basic strokes of the national standard hard pen include at least horizontal, vertical, left-falling, right-falling, dot, hook, and turn; during comparison, if the grip force feature exceeds the threshold, it is determined that the grip force is abnormal; if the pressure curve deviates from the standard range, it is determined that the stroke force is not standardized; if the pen tilt angle exceeds the standard, it is determined that the grip posture is incorrect; if the pen movement time or acceleration exceeds the standard range, it is determined that the writing rhythm is disordered.

10. The detection method for a multi-sensor fusion calligraphy force detection intelligent calligraphy pen according to claim 6, characterized in that, It also includes step S6 sleep energy saving control: when there is no pressure signal from the pen tip for 30 consecutive seconds, the synchronous sampling MCU main control chip (6) reduces the sensor sampling frequency and enters low power mode.

Images