Board racket track optimization method for real-time monitoring of board ball and self-driven board racket

Abstract

The invention provides a racket track optimization method for real-time monitoring of a cricket ball and a self-driven racket, and relates to the technical field of motion track planning, and the method comprises the following steps: S1, collecting the actual position, friction induction current and pressure induction current of a hitting point in real time; s2, converting the collected information into a digital signal and transmitting the digital signal to an upper computer; s3, inputting multiple groups of sample data into the action learning model, establishing a corresponding mapping relation, and outputting a real-time moving track; s4, comparing the real-time moving trajectory with a preset ideal trajectory to obtain a trajectory deviation value; and S5, reversely optimizing the control parameters of the ball hitting action, and outputting an optimized parameter set. According to the method, the real-time moving track is compared with the preset ideal moving track, the control parameters of the ball hitting action are reversely optimized according to the deviation of the real-time moving track and the preset ideal track, and the adjusted control parameters are output, so that the moving track of the bat can be optimized.

Description

Technical Field

[0001] This invention relates to the field of sports training monitoring technology, and in particular to a method for optimizing cricket racket trajectory for real-time monitoring and a self-driven cricket racket. Background Technology

[0002] In cricket, factors such as batting position, racket movement speed, posture control, power output, and body coordination directly affect batting accuracy and scoring efficiency, and are crucial in determining the outcome of a match. Cricket training urgently needs to shift from traditional experience-driven methods to data-driven, intelligent approaches. However, current cricket training still faces problems such as insufficient quantification of technical movements and a lack of real-time feedback methods, making the development of a real-time monitoring system based on intelligent sensors a pressing need.

[0003] Current motion monitoring technologies primarily rely on optical motion capture systems and inertial measurement units (IMUs). Optical systems use multiple cameras to track reflective markers, acquiring high-precision three-dimensional kinematic data. However, these systems are expensive, limited by site conditions, and difficult to adapt to outdoor real-time monitoring scenarios. While IMUs offer advantages in portability and low cost, they require the deployment of sensors on multiple parts of the user's body, leading to high wearing complexity, poor comfort, and complex multi-sensor data fusion algorithms, limiting practical applications. In recent years, wearable flexible sensors, such as textile-based sensors and conductive polymer materials, have shown potential in the field of motion monitoring. However, their complex manufacturing processes and reliance on external power systems result in low reliability, limiting large-scale applications. Furthermore, existing motion monitoring systems based on wearable flexible sensors mostly focus on sensing single motion parameters, lacking targeted designs for sports like cricket, which require simultaneous analysis of hit position and grip dynamics.

[0004] In conclusion, there is an urgent need in the field of sports training monitoring for a self-driven intelligent cricket racket for real-time training monitoring in cricket, which can achieve dynamic tracking of the hitting position and real-time perception of grip force changes. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a real-time cricket action system that can collect control parameters of the user when hitting the cricket in real time, thereby calculating the real-time trajectory of the cricket after being hit, comparing it with the preset ideal trajectory, and outputting a judgment result on the current hitting action category; then, based on the deviation between the real-time trajectory and the preset ideal trajectory, the control parameters of the hitting action are optimized in reverse, and the optimized control parameters are output, thereby obtaining the ideal cricket racket movement trajectory.

[0006] Specifically, the present invention provides a method for optimizing cricket racket trajectory for real-time cricket monitoring, which includes the following steps: S1. Real-time acquisition of the actual position of the hitting point on the cricket racket at the moment of impact, the friction sensing current generated by the friction sensing module on the cricket racket when the leather surface of the cricket ball contacts the thermal film on the cricket racket, and the pressure sensing current generated synchronously by the piezoelectric sensing module on the handle. S2. Convert the collected actual position, friction induced current and pressure induced current into digital signals and transmit them to the host computer; S3. Input multiple sets of sample data of the hitting action into the hitting action learning model in the host computer, establish the mapping relationship between the temporal characteristics of the multi-channel voltage signal and the hitting action category, and output the real-time trajectory of the cricket ball after being hit: : ; In the formula, For the moment of cricket kicking, The time from launch to the next collision; Let be the component of the launch velocity in the coordinate system. Let be the angle between the velocity direction and the xz plane. Let be the angle between the projection of the velocity onto the xz plane and the positive z-axis. For Magnus force The angle with the xz plane, The coordinates of the cricket's launch position in space; Where m is air resistance and m is the mass of the cricket ball. Where D is the air density and D is the diameter of the cricket ball; S4. Through the signal analysis and comparison unit in the host computer, the type of hitting action is determined according to the mapping relationship in S3, and the real-time trajectory after the ball is hit is compared with the preset ideal trajectory of the corresponding hitting action type to obtain the trajectory deviation value. S5. Based on the trajectory deviation value, optimize the control parameters of the hitting action in reverse and output the optimized parameter set. The control parameters include: racket face incident angle. , velocity vector of the shot Tangential force of the racket face , contact point of the racket face and grip strength .

[0007] Furthermore: In S5, the set of output optimization parameters includes: By minimizing the error function between the actual trajectory of the cricket ball and the ideal trajectory, the cricket racket control parameters related to the hitting action are optimized in reverse. The coordinates of the cricket racket's striking point are obtained by weighting the voltages of multiple triboelectric channels. The cricket racket face posture is calculated by voltage gradient; By combining machine learning models, piezoelectric waveforms are mapped to specific hitting action categories and launch velocity ranges; Based on the trajectory deviation value obtained in S4, and using the error gradient, the racket face parameters at the time of the collision event are corrected: ; in, The amount of change in the motion parameters that need to be adjusted; The intensity of the adjustment of the action parameters is determined by the learning gain matrix; This represents the partial derivatives of the trajectory error with respect to each parameter; Adjust the mapping library in the host computer: ; Iterative updates are used: ; in, This is the parameter vector for each shot in the current collision event; The learning rate is used to control the adjustment range each time and output an optimized solution. This represents the trajectory deviation value.

[0008] Further: Step S2 includes the following sub-steps: S21. Friction-induced current and pressure-induced current flow through the connected external load, thereby generating corresponding voltage signals; S22. The voltage signals output by each sensor are synchronously acquired through the signal acquisition module; S23. The analog voltage signal is converted into a digital signal by the analog-to-digital converter built into the signal acquisition module; S24. Transmit the acquired digital signals to the host computer in the computer system through the communication interface.

[0009] Furthermore: In step S3, the process of transmitting the signal generated by the real-time hitting action to the hitting action learning model for calculation and analysis includes: S31: Estimate the velocity of the ball upon release based on the four-channel voltage signal output from the piezoelectric sensing module and the ball-hitting motion learning model. and the direction of velocity; S32: The motion learning model maps voltage timing features to different speed ranges corresponding to different types of hitting actions through machine learning; S33: Based on the type of striking action and the characteristics of piezoelectric signals, the tangential force of the striking action is obtained through linear fitting of the action model, and the spin angular velocity of the cricket ball at the time of launch is estimated based on the tangential force of the striking action. Rotation direction and rotation axis attitude data; S34: Calculate the real-time trajectory of the cricket ball after it is struck. .

[0010] Furthermore: In step S34, the trajectory of the cricket ball after being struck is calculated. Includes the following sub-steps: S341: Calculate the gravitational force acting on the cricket ball during the next ball-collision event. air resistance And Magnus : S342: Calculate the cricket velocity components before the next collision event. : ; S343: By integrating the velocity components in S342, the real-time trajectory of the cricket ball after being struck is obtained.

[0011] Further: In step S341, air resistance The calculation method is as follows: ; in, air density, The cross-sectional area of ​​the cricket ball. The instantaneous velocity of a cricket ball. The drag coefficient is calculated based on the Reynolds number. calculate: ; Among them, Reynolds number : ; In the formula, The cross-sectional area of ​​the cricket ball. is the air viscosity coefficient.

[0012] Further: Step S1 involves acquiring the actual position of the striking point on the cricket racket at the moment of impact, including the following sub-steps: S11. The triboelectric signal transmitted to the host computer is in the form of real-time updated multi-channel data, and the cricket racket with multiple hitting positions is set on the host computer interface. S12. Set a column of feature rectangles of the same size in the host computer; S13. Set up a comparison function in the host computer to analyze multi-channel data in real time, select the largest one in each group of data, and change the color of the rectangle block at its corresponding position. S14. Set critical conditions in the host computer so that the rectangular block at the corresponding position changes color only when the maximum voltage in each set of data exceeds 0.5V, thereby controlling the detection error and effectively reducing the impact of other interferences on the detection results.

[0013] Further: Step S4 includes the following sub-steps: S41. The piezoelectric signal transmitted to the host computer is in the format of real-time updated four-channel data, and a ball-hitting action display module is set on the host computer interface. S42. By comparing real-time data features with the feature patterns stored in the learning model of the hitting action, output the judgment result of the current hitting action category.

[0014] The present invention also provides a self-driven racket for implementing the aforementioned trajectory optimization method, which includes a racket body equipped with a friction sensing module and a handle equipped with a piezoelectric sensing module and connected to the racket body. The friction sensing module includes multiple copper foils and a cover layer disposed on the first end face of the cricket racket. The copper foils are evenly spaced apart. A signal acquisition circuit board is provided on the second end face of the cricket racket. Each copper foil is connected to the signal acquisition circuit board through a first wire. The cover layer is a thermal film, which covers the copper foils and wraps around the cricket racket. Multiple copper foils cover the striking area of ​​the cricket racket. When the cricket ball hits one of the copper foils, the signal acquisition circuit board identifies the local coordinates of the striking point through multi-channel voltage signals. Combined with the real-time posture data of the cricket racket captured by the measurement unit, the local coordinates are converted into global coordinates. ); The piezoelectric sensing module includes piezoelectric films and an insulating layer. The piezoelectric films include a first piezoelectric film, a second piezoelectric film, a third piezoelectric film, and a fourth piezoelectric film. The first and third piezoelectric films are located on the first end face of the handle, and the second and fourth piezoelectric films are located on the second end face of the handle. The first and second piezoelectric films are perpendicularly aligned, as are the third and fourth piezoelectric films. Each piezoelectric film has a silver-plated layer on both ends. Each piezoelectric film is connected to a signal acquisition circuit board via a second wire. The insulating layer is insulating tape cut to a preset shape and size. The insulating tape is wrapped around the four piezoelectric films and then wrapped around the handle.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The self-driven cricket device and trajectory judgment method for real-time monitoring of cricket movement provided by the present invention have a friction sensing module on the cricket racket, which can be used to accurately lock the hitting position, and a piezoelectric sensing module on the handle, which can be used to sense the dynamic grip posture and action analysis during the hitting process in real time. There is no need for external bulky sensors and power supply, minimizing interference with the user's natural movements.

[0016] 2. The self-driven cricket device and trajectory judgment method for real-time monitoring of cricket movements provided by this invention can realize real-time monitoring of cricket striking actions. Through continuous learning of striking actions, it can improve the accuracy of users in the three core actions of forward defense, smash, and cross-court press, providing strong guidance for the learning and training of cricket techniques.

[0017] 3. This invention provides a method for judging the trajectory of a cricket ball. It comprehensively considers factors such as the ball's trajectory after being hit, its rotation, and air resistance. It can accurately determine the trajectory of the ball after being hit and compare it with a pre-set standard trajectory to determine whether the user's hitting action and dynamic grip are standard. It also helps to optimize the cricket racket's motion trajectory by outputting an optimized parameter set.

[0018] 4. This invention compares the real-time trajectory of the cricket racket after being struck with the preset ideal trajectory, and optimizes the control parameters of the striking action based on the deviation between the real-time trajectory and the preset ideal trajectory. It also outputs adjustment suggestions for the control parameters, such as tilting the racket face forward by 5°, increasing the racket movement speed by 3%, and shifting the striking point to the right by 10mm. This helps users optimize the movement trajectory of the cricket racket and the striking point of the cricket racket, thereby effectively improving the accuracy of the strike and the scoring efficiency. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the overall structure of the self-driving racket of the present invention; Figure 2 This is a schematic diagram illustrating the manufacturing process of the friction sensing module of the present invention; Figure 3 This is a photograph of the friction sensing module of the present invention installed on a cricket racket; Figure 4 This is a schematic diagram illustrating the working principle of the friction sensing module of the present invention; Figure 5 This is a photograph of the piezoelectric sensing module of the present invention mounted on the handle; Figure 6 This is a schematic diagram illustrating the working principle of the piezoelectric sensing module of the present invention; Figure 7 This is a diagram showing the output voltage of the racket surface friction sensing module sensor under different pressures according to the present invention. Figure 8 This is a diagram showing the output voltage of the racket surface friction sensing module sensor at different distances according to the present invention. Figure 9 The output voltage diagram of the racket surface friction sensing module sensor at different frequencies of the present invention is shown. Figure 10 This is a diagram showing the output voltage of the friction sensing module on the surface of the cricket racket when the balls strike the center of the sensor in sequence. Figure 11 This is a diagram showing the output voltage of the piezoelectric sensing module on the racket handle when using the standard grip posture in this invention. Figure 12 This is a schematic diagram of the three basic movements in cricket in this invention; Figure 13 The output voltage diagram of the piezoelectric sensing module of the racket handle under the three basic actions of this invention is shown. Figure 14 The diagram shows the output voltage of the piezoelectric sensor module on the handle when different volunteers perform the three basic actions of cricket. Figure 15 This is a schematic diagram illustrating the convolutional neural network used for classification and recognition in image processing according to the present invention, along with three basic actions. Figure 16 This is a confusion matrix diagram for classifying the three basic actions of this invention; Figure 17 This is a graph showing the accuracy and loss of the model used in this invention for training. Figure 18 This is a schematic diagram illustrating the working principle of the cricket training and scoring system of the present invention; Figure 19 This is a diagram showing the output voltage of the ten-channel sensor used in this invention to monitor the cricket training process. Figure 20 This invention is based on the analysis of the output signal to identify the display results and scores of the hitting position and grip posture, as well as optical photographs taken during use on the court; Figure 21 This is an overall flowchart of the cricket racket trajectory optimization method for real-time cricket monitoring according to the present invention. Detailed Implementation

[0020] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0021] The present invention provides a self-driven cricket racket, as a self-driven cricket device, such as... Figure 1 and Figure 2As shown, it includes a cricket racket with a friction sensing module and a handle with a piezoelectric sensing module. The friction sensing module includes the cricket racket, six copper foils cut to a preset shape and size, and a cover layer. Adjacent copper foils are evenly spaced and adhered to the first end face of the cricket racket. A signal acquisition circuit board is located on the second end face of the cricket racket, and the first end of each copper foil is connected to the signal acquisition circuit board via a first wire. The cover layer is a thermally sensitive film cut to a preset shape, covering the copper foils and encasing the cricket racket. The piezoelectric sensing module includes a handle, four piezoelectric films cut to a preset shape and size, and an insulating layer. The four piezoelectric films include a first piezoelectric film, a second piezoelectric film, a third piezoelectric film, and a fourth piezoelectric film. The first and third piezoelectric films are located on the first end face of the handle, and the second and fourth piezoelectric films are located on the first end face of the handle. The piezoelectric films are located on the second end face of the handle, with the first and second piezoelectric films perpendicularly corresponding to each other, as well as the third and fourth piezoelectric films. Each piezoelectric film has a silver-plated layer on its first and second end faces. The first end of each piezoelectric film is connected to the signal acquisition circuit board via a second wire. The insulating layer is insulating tape cut to a preset shape and size, which is wrapped around the four piezoelectric films and the handle. When a leather-surfaced cricket ball comes into contact with the friction sensing module, the friction sensing module generates an induced current. By comparing the electrical signals of other pathways, the actual position of the ball being struck is accurately determined. Simultaneously, when the user holds the handle and performs a hitting action, the system outputs a judgment result of the current hitting action category by comparing it with the trained machine learning model, thereby achieving automatic recognition and classification of the hitting action.

[0022] like Figure 3 and Figure 4 As shown, for the friction sensing module, when the leather-surfaced cricket ball comes into contact with the thermistor on the cricket paddle, the thermistor absorbs electrons from the leather surface of the cricket ball, at which point the thermistor carries a negative charge, while the leather surface of the cricket ball carries a positive charge. When the leather-surfaced cricket ball separates from the thermistor on the cricket paddle, the potential difference between the thermistor and the leather surface gradually increases, causing a transient electron flow from the copper foil to the ground terminal in the external circuit. This transient electron flow continues until the leather-surfaced cricket ball and the thermistor on the cricket paddle are completely separated. When the leather-surfaced cricket ball approaches the thermistor on the cricket paddle, under the action of an external load, a transient electron flow is generated from the ground terminal to the copper foil. This transient electron flow continues until the leather-surfaced cricket ball and the thermistor on the cricket paddle are in complete contact. Through the reciprocating motion between the leather-surfaced cricket ball and the thermistor on the cricket paddle, alternating current is generated. Figure 3 1 is a wire, 2 is a heat-sensitive material, and 3 is a conductive copper tape.

[0023] like Figure 5 and Figure 6As shown, for the piezoelectric sensing module, when the user holds the handle, the external pressure causes the piezoelectric film on the handle to deform, resulting in bound charges on the first and second end faces of the piezoelectric film. This disrupts the electrostatic balance between the silver plating layers on the first and second end faces of the piezoelectric film. At this time, electrons are redistributed between the silver plating layers on the first and second end faces of the piezoelectric film, creating an instantaneous potential difference between the silver plating layers, which generates current in the external circuit. When the user does not hold the handle, the piezoelectric film returns to its initial shape, generating a reverse current, thereby realizing the conversion of the mechanical pressure signal into an alternating current signal.

[0024] Specifically, it converts mechanical energy into electrical energy based on the piezoelectric effect. Its core is a piezoelectric thin film with a β-crystalline phase, which has undergone polarization treatment. The CF2 and CH2 groups in the molecular chain of the piezoelectric thin film form an inherent dipole moment. External pressure causes deformation of the piezoelectric thin film, which alters the arrangement and spacing of the internal dipole moments, leading to a change in the spontaneous polarization intensity of the material. This, in turn, generates bound charges on the upper and lower surfaces of the piezoelectric thin film, disrupting the electrostatic balance of the Ag coating on the electrodes. To balance the bound charges, electrons redistribute between the upper and lower electrodes, thus generating a transient potential difference between them. Connecting an external circuit drives the electron flow to form a current. When the external force is removed, the thin film recovers and generates a reverse current pulse, ultimately achieving the conversion of mechanical pressure signals into alternating current signals.

[0025] like Figure 21 As shown, a second aspect of the present invention provides a method for optimizing the trajectory of a self-driven racket, comprising the following steps: S1. Collect the actual position of the hitting point on the cricket racket at the moment of impact, the induced current generated by the friction sensor module when the leather surface of the cricket ball comes into contact with the thermal film on the cricket racket, and the induced current generated synchronously by the piezoelectric sensor module on the handle when the user holds the handle to hit the ball. S11. The triboelectric signal transmitted to the host computer is in the form of a six-channel data that is updated in real time, e.g., P1: 0.12V, P2: 0.34V, P3: 1.35V, P4: 0.01V, P5: 0.04V, P6: 0.13V. A cricket racket with six hitting positions P1, P2, P3, P4, P5, and P6 is set on the host computer interface. S12. In the host computer, set a column of white rectangular blocks of the same size at positions P1, P2, P3, P4, P5, and P6; S13. By setting a comparison function in the software, the six-channel data is analyzed in real time, the largest one in each group of data is selected, and the corresponding rectangular block is turned red. S14. By setting a critical condition in the software, the rectangular block at the corresponding position turns red only when the maximum voltage in each set of data exceeds 0.5V, thus controlling the detection error and effectively reducing the impact of other interferences on the detection results.

[0026] S2. Transmit the signals collected in real time from the friction sensing module and the piezoelectric sensing module to the host computer in the computer system. Includes the following sub-steps: S21. The induced current generated by the tribological sensing module and the piezoelectric sensing module flows through the connected external load, thereby generating a corresponding voltage signal. S22. The voltage signals output by each sensor are synchronously acquired through the signal acquisition module; S23. The analog voltage signal is converted into a digital signal by the analog-to-digital converter built into the signal acquisition module; S24. Transmit the collected signals to the host computer in the computer system through the communication interface.

[0027] S3. Input multiple sets of sample data of the hitting action into the hitting action learning model in the host computer, establish the mapping relationship between the temporal characteristics of the multi-channel voltage signal and the hitting action category, and output the real-time running trajectory after the ball is hit.

[0028] Let the spatial coordinate system be a rectangular coordinate system of xyz, where the horizontal plane is represented by the xz plane and the positive y-axis is represented by the vertical upward direction; The launch position is determined based on the impact position data collected by the friction sensor module on the cricket racket. The friction sensor module contains six copper foil areas P1-P6. When the cricket hits a certain area, the local coordinates of the impact point, such as P3: 1.35V, are identified through multi-channel voltage signals. The real-time attitude of the cricket racket is captured by an inertial measurement unit or camera, and combined with the real-time attitude data of the cricket racket, it is converted into global coordinates. ).

[0029] Based on the four-channel voltage signals output by the piezoelectric sensing module, such as P1: 0.24V, P2: 0.32V, P3: 1.46V…, and the ball-hitting motion learning model, the velocity of the cricket ball upon release is estimated. and velocity direction, including angle, etc. and The motion model uses ResNet50 machine learning to map voltage timing features to different speed ranges corresponding to typical shooting speed patterns such as forward defense, slam, and lateral pressure.

[0030] Estimate the angular velocity of the ball's spin upon launch based on the type of striking action and the characteristics of the piezoelectric signal. Data on spin direction and spin axis attitude. Cricket spin mainly originates from the tangential force of the striking motion, and this data can be obtained through linear fitting of the motion model.

[0031] Includes the following sub-steps: S31: Estimate the velocity of the ball upon release based on the four-channel voltage signal output from the piezoelectric sensing module and the ball-hitting motion learning model. and the direction of velocity; S32: The motion learning model maps voltage timing features to different speed ranges corresponding to different types of hitting actions through machine learning; S33: Based on the type of striking action and the characteristics of piezoelectric signals, the tangential force of the striking action is obtained through linear fitting of the action model, and the spin angular velocity of the cricket ball at the time of launch is estimated based on the tangential force of the striking action. Rotation direction and rotation axis attitude data; S34: Calculate the real-time trajectory of the cricket ball after it is struck. : S341: The gravitational force acting on the cricket ball at the next ball-collision event is obtained according to the following expression. : ; in, For cricket quality, This is the acceleration due to gravity.

[0032] The gravity acting on the cricket ball at the next ball-collision event is calculated. air resistance And Magnus : air resistance The calculation method is as follows: ; in, air density, The cross-sectional area of ​​the cricket ball. The instantaneous velocity of a cricket ball. The drag coefficient is calculated based on the Reynolds number. calculate: ; Among them, Reynolds number : ; In the formula, air density, The cross-sectional area of ​​the cricket ball. For cricket speed, is the air viscosity coefficient.

[0033] S342: Calculate the cricket velocity components before the next collision event. : ; S343: By integrating the velocity components in S342, the real-time trajectory of the cricket ball after being struck is obtained. ; The final real-time running trajectory is obtained: : ; In the formula, At the moment of launch, The time from launch to the next collision; Let be the component of the launch velocity in the coordinate system. Let be the angle between the velocity direction and the xz plane. Let be the angle between the projection of the velocity onto the xz plane and the positive z-axis. For Magnus force The angle with the xz plane, The coordinates of the launch location in space; S4. By comparing the real-time trajectory of the ball after being hit with the preset ideal trajectory through the signal analysis and comparison unit in the host computer, the result of the judgment of the current hitting action category is output.

[0034] Includes the following sub-steps: S41. The piezoelectric signal transmitted to the host computer is in the format of real-time updated four-channel data, and a ball-hitting action display module is set on the host computer interface. S42. By comparing real-time data features with the feature patterns stored in the learning model of the hitting action, output the judgment result of the current hitting action category.

[0035] S5. Based on the deviation between the real-time trajectory and the preset ideal trajectory, optimize the control parameters of the hitting action in reverse, and output adjustment suggestions for the control parameters. These control parameters include: racket face angle of incidence. , velocity vector of the shot Tangential force of the racket face , contact point of the racket face , grip strength .

[0036] Specifically, by analyzing the deviation between the cricket's trajectory after a collision event and its ideal trajectory, a method for reverse optimization of cricket racket control related to the striking action is proposed. The optimization parameters include: racket face incident angle. , velocity vector of the shot Tangential force of the racket face , contact point of the racket face , grip strength The aim is to optimize the cricket racket control parameters related to the striking action by minimizing the error function between the actual trajectory of the cricket ball and the ideal trajectory. ; in: The trajectory deviation value indicates the degree of deviation from the ball's trajectory. This represents the total time the cricket ball spends in the air. The actual cricket ball trajectory predicted based on sensors and physical models; For the ideal shot trajectory.

[0037] The coordinates of the point of impact on the cricket racket are calculated by weighting the voltages of the six triboelectric channels P1-P6: ; in, The coordinates of the center of the hitting point are calculated; For the first The induced voltage value of each copper foil channel; Let represent the position of the copper foil in the coordinate system of the cricket racket surface.

[0038] The cricket racket face posture is derived from the voltage gradient. Since the forces acting on different channels of the racket differ during a collision, a voltage difference exists, leading to the following calculations: ; in, Let x be the angle of the racket face in the x and y directions. The output voltage of the six triboelectric signal channels.

[0039] The racket's movement speed and rotation characteristics are calculated from the rate of change of the four-channel piezoelectric signal on the handle: ; in: The output voltage of the four piezoelectric films; The component representing the velocity of the racket face movement; This is the tangential force acting on the racket face.

[0040] By combining the ResNet50 machine learning model, the piezoelectric waveform is mapped to specific hitting action categories and launch velocity ranges.

[0041] Based on Magnus force, air resistance, and the ball's own weight, the following formula is derived for solving the ball's trajectory in the air after a collision event: ; in, For cricket ball quality; The instantaneous velocity vector of the cricket ball; The angular velocity vector of the cricket ball's spin; This is the acceleration due to gravity, directed perpendicularly downwards from the ground. This is the air drag coefficient term; This is the coefficient term of the Magnus force.

[0042] By integrating the above equations, the actual trajectory of the cricket ball can be obtained. .

[0043] The trajectory error of a cricket ball is calculated as follows: ; in, This refers to the positional deviation error; This refers to spin error; This refers to the key time points following the collision event.

[0044] Based on the calculated error gradient, the racket face parameters of the cricket racket at the time of the collision event are corrected: ; in, The amount of change in the motion parameters that need to be adjusted; The intensity of the adjustment of the action parameters is determined by the learning gain matrix; This represents the partial derivatives of the trajectory error with respect to each parameter.

[0045] Finally, maintain a ball-hitting feature—trajectory deviation—parameter adjustment mapping library in the host computer: ; Iterative updates are used: ; in, This is the parameter vector for each shot in the current collision event; Use the learning rate to control the magnitude of each adjustment.

[0046] In a preferred embodiment of the present invention, the optimized parameter set provided by the present invention includes: 5° forward tilt of the cricket racket, 3% increase in racket movement speed, and 10mm shift of the hitting point to the right, etc.

[0047] Specifically, in a preferred embodiment of a self-driven cricket ball device of the present invention, such as Figures 9-20 As shown, it includes a cricket racket equipped with a friction sensing module and a handle equipped with a piezoelectric sensing module.

[0048] like Figure 7As shown, the output voltage of the racket surface friction sensing module under different pressures is illustrated. Measured data indicates that as the pressure on the sensor increases, the open-circuit voltage gradually increases, subsequently exhibiting nonlinear saturation characteristics. This positive correlation between the output voltage and pressure, along with the saturation characteristic, demonstrates that the friction sensing module can effectively convert externally applied pressure stimuli into electrical signals with corresponding amplitudes. These analog electrical signals can be received and processed by a subsequent signal acquisition module, which preferably includes a signal conditioning circuit and an analog-to-digital converter to convert the analog electrical signals into digital signals. The converted digital signals are then transmitted to a remote data processing device via a communication unit, providing a reliable data foundation for determining the hitting position based on pressure distribution.

[0049] like Figure 8 As shown, the output voltage of the friction sensing module sensor on the racket surface is displayed at different distances. The measured data shows that the amplitude of the open-circuit voltage increases significantly with the increase of the relative displacement between the friction sensing module sensor and the triggering object. This correlation between the output voltage and the movement distance proves that the friction sensing module sensor can effectively convert external dynamic mechanical stimuli into electrical signals with specific amplitudes. This analog electrical signal can be received and processed by a subsequent signal acquisition module, which preferably includes a signal conditioning circuit and an analog-to-digital converter to convert the analog electrical signal into a digital signal. The converted digital signal is transmitted to a remote data processing device through a communication unit, thus providing a reliable data basis for determining the hitting position based on pressure distribution.

[0050] like Figure 9 The figure shows the output voltage of the friction sensing module sensor on the racket surface at different frequencies. The measured data demonstrates the frequency insensitivity and high stability of this output voltage, indicating that the friction sensing module sensor can still generate a reliable and consistent voltage signal when the external excitation frequency changes. This analog electrical signal can be received and processed by a subsequent signal acquisition module. The signal acquisition module preferably includes a signal conditioning circuit and an analog-to-digital converter to convert the analog electrical signal into a digital signal. The converted digital signal is transmitted to a remote data processing device through a communication unit, thus providing a reliable data basis for determining the hitting position based on pressure distribution.

[0051] like Figure 10The figure shows the output voltage when a cricket ball sequentially strikes the center of multiple friction sensor modules on the surface of the cricket racket. As can be seen from the figure, under the same experimental conditions, the signal characteristics of channels striking different areas are significantly higher than those of other channels. This analog electrical signal can be received and processed by a subsequent signal acquisition module. The signal acquisition module preferably includes a signal conditioning circuit and an analog-to-digital converter to convert the analog electrical signal into a digital signal. The converted digital signal is transmitted to a remote data processing device through a communication unit, thereby providing a reliable data basis for determining the striking position based on pressure distribution.

[0052] like Figure 11 The figure shows the output voltage of the piezoelectric sensor module on the cricket racket handle when using the standard grip.

[0053] like Figure 12 The diagram shown illustrates three basic movements in cricket: forward guarding (FD), batting (D), and cross-court press (S).

[0054] like Figure 13 As shown, the output voltage of the piezoelectric sensing module of the racket handle under three basic actions is shown. There are obvious differences and characteristics between each signal waveform, indicating that different hitting actions can be identified based on the differences in the signals.

[0055] like Figure 14 As shown, the output voltage of the piezoelectric sensor module on the cricket racket handle is displayed when different volunteers perform the three basic cricket movements. This shows the differences and similarities in the voltage signals of the same hitting action, and different hitting actions can be identified based on the differences and similarities in the signals.

[0056] like Figure 15 The diagram illustrates the working principle of the ResNET50 model, a convolutional neural network based on image processing. This convolutional neural network is used to classify and recognize telecommunications signal waveforms related to these three basic actions. This invention constructs a network-based voltage waveform fusion analysis model. This model, designed for feature fusion and pattern recognition of four-channel voltage waveform data, employs a deep residual network architecture and includes three core modules: a multimodal input layer, a feature fusion backbone network, and a classification head network. Training optimization was performed using a momentum adaptive optimizer with an initial learning rate of 0.001, coupled with a cosine annealing strategy. The batch size was set to 32. Specifically, each action contains 80 samples, and each sample is mapped to a 4-channel voltage waveform. The acquired waveforms are divided into three groups in a 7:2:1 ratio, for example, 70% for training, 20% for testing, and 10% for validation.

[0057] like Figure 16As shown, the confusion matrix for the three basic motion classifications is presented. The voltage signal generated by the piezoelectric sensing module achieves a high classification accuracy of 95%, indicating that the piezoelectric sensing array is effective in capturing body posture features.

[0058] like Figure 17 As shown, the accuracy and loss of the model used for training are displayed, indicating that as the number of samples increases, the accuracy gradually increases to 95%, and the training loss gradually decreases, demonstrating high accuracy.

[0059] like Figure 18 The diagram shown illustrates the working principle of a cricket training scoring system. This invention integrates intelligent cricket, machine learning algorithms, and feedback applications into a wirelessly transmitted training platform that can provide corresponding scores based on real-time monitoring of the user's athletic performance.

[0060] like Figure 19 As shown in the figure, the output voltage of the ten-channel sensor is used to monitor the cricket training process. The figure shows the voltage signals of different channels of the ten-channel sensor in practical applications.

[0061] like Figure 20 As shown, this invention identifies the hitting position and grip posture based on the analysis of the DMSP-CR output signal, as well as the optical photographs taken on the court when using the DMSP-CR. Based on the signal strength of the six channels at the racket, the invention can distinguish and identify the hitting point; based on the signal strength of the four channels at the handle, it can distinguish and identify the hitting action, and finally give the corresponding score.

[0062] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for optimizing cricket racket trajectory for real-time cricket monitoring, characterized in that: It includes the following steps: S1. Real-time acquisition of the actual position of the hitting point on the cricket racket at the moment of impact, the friction sensing current generated by the friction sensing module on the cricket racket when the leather surface of the cricket ball contacts the thermal film on the cricket racket, and the pressure sensing current generated synchronously by the piezoelectric sensing module on the handle. S2. Convert the collected actual position, friction induced current and pressure induced current into digital signals and transmit them to the host computer; S3. Input multiple sets of sample data of the hitting action into the hitting action learning model in the host computer, establish the mapping relationship between the temporal characteristics of the multi-channel voltage signal and the hitting action category, and output the real-time trajectory of the cricket ball after being hit: ： ； In the formula, For the moment of cricket kicking, The time from launch to the next collision; Let be the component of the launch velocity in the coordinate system. Let be the angle between the velocity direction and the xz plane. Let be the angle between the projection of the velocity onto the xz plane and the positive z-axis. For Magnus force The angle with the xz plane, The coordinates of the cricket's launch position in space; Where m is air resistance and m is the mass of the cricket ball. Where D is the air density and D is the diameter of the cricket ball; S4. Through the signal analysis and comparison unit in the host computer, the type of hitting action is determined according to the mapping relationship in S3, and the real-time trajectory after the ball is hit is compared with the preset ideal trajectory of the corresponding hitting action type to obtain the trajectory deviation value. S5. Based on the trajectory deviation value, optimize the control parameters of the hitting action in reverse and output the optimized parameter set. The control parameters include: racket face incident angle. , velocity vector of the shot Tangential force of the racket face , contact point of the racket face and grip strength .

2. The cricket racket trajectory optimization method for real-time cricket monitoring as described in claim 1, characterized in that: In S5, the set of optimized parameters output includes: By minimizing the error function between the actual trajectory of the cricket ball and the ideal trajectory, the cricket racket control parameters related to the hitting action are optimized in reverse. The coordinates of the cricket racket's striking point are obtained by weighting the voltages of multiple triboelectric channels. The cricket racket face posture is calculated by voltage gradient; By combining machine learning models, piezoelectric waveforms are mapped to the type of hitting action and the range of launch velocities. Based on the trajectory deviation value obtained in S4, and using the error gradient, the racket face parameters at the time of the collision event are corrected: ； in, The amount of change in the motion parameters that need to be adjusted; To learn the gain matrix and adjust the intensity of the action parameters; This represents the partial derivatives of the trajectory error with respect to each parameter; Adjust the mapping library in the host computer: ； Iterative updates are used: ； in, This is the parameter vector for each shot in the current collision event; The learning rate is used to control the adjustment range each time and output an optimized solution. This represents the trajectory deviation value.

3. The cricket racket trajectory optimization method for real-time cricket monitoring as described in claim 1, characterized in that: Step S2 includes the following sub-steps: S21. Friction-induced current and pressure-induced current flow through the connected external load, thereby generating corresponding voltage signals; S22. The voltage signals output by each sensor are synchronously acquired through the signal acquisition module; S23. The analog voltage signal is converted into a digital signal by the analog-to-digital converter built into the signal acquisition module; S24. Transmit the acquired digital signals to the host computer in the computer system through the communication interface.

4. The cricket racket trajectory optimization method for real-time cricket monitoring as described in claim 3, characterized in that: Step S3, the process of transmitting the signal generated by the real-time hitting action to the hitting action learning model for calculation and analysis, includes: S31: Estimate the velocity of the ball upon release based on the four-channel voltage signal output from the piezoelectric sensing module and the ball-hitting motion learning model. and the direction of velocity; S32: The motion learning model maps voltage timing features to different speed ranges corresponding to different types of hitting actions through machine learning; S33: Based on the type of striking action and the characteristics of piezoelectric signals, the tangential force of the striking action is obtained through linear fitting of the action model, and the spin angular velocity of the cricket ball at the time of launch is estimated based on the tangential force of the striking action. Rotation direction and rotation axis attitude data; S34: Calculate the real-time trajectory of the cricket ball after it is struck. .

5. The cricket racket trajectory optimization method for real-time cricket monitoring as described in claim 4, characterized in that: In step S34, the trajectory of the cricket ball after being struck is calculated. Includes the following sub-steps: S341: Calculate the gravitational force acting on the cricket ball during the next ball-collision event. air resistance And Magnus : S342: Calculate the cricket velocity components before the next collision event. : ； S343: By integrating the velocity components in S342, the real-time trajectory of the cricket ball after being struck is obtained.

6. The cricket racket trajectory optimization method for real-time cricket monitoring as described in claim 5, characterized in that: In step S341, air resistance The calculation method is as follows: ； in, air density, The cross-sectional area of ​​the cricket ball. The instantaneous velocity of a cricket ball. The drag coefficient is calculated based on the Reynolds number. calculate: ； Among them, Reynolds number : ； In the formula, The cross-sectional area of ​​the cricket ball. is the air viscosity coefficient.

7. The cricket racket trajectory optimization method for real-time cricket monitoring as described in claim 1, characterized in that: Step S1 involves acquiring the actual position of the striking point on the cricket racket at the moment of impact, including the following sub-steps: S11. The triboelectric signal transmitted to the host computer is in the form of real-time updated multi-channel data, and the cricket racket with multiple hitting positions is set on the host computer interface. S12. Set a column of feature rectangles of the same size in the host computer; S13. Set up a comparison function in the host computer to analyze multi-channel data in real time, select the largest one in each group of data, and change the color of the rectangle block at its corresponding position. S14. Set a critical condition in the host computer: when the maximum voltage in each set of data exceeds 0.5V, the rectangular block at the corresponding position will change color, thereby controlling the detection error.

8. The cricket racket trajectory optimization method for real-time cricket monitoring as described in claim 7, characterized in that: Step S4 includes the following sub-steps: S41. The piezoelectric signal transmitted to the host computer is in the format of real-time updated four-channel data, and a ball-hitting action display module is set on the host computer interface. S42. By comparing real-time data features with the feature patterns stored in the learning model of the hitting action, output the judgment result of the current hitting action category.

9. A self-driven cricket racket for implementing the cricket racket trajectory optimization method for real-time monitoring of cricket as described in any one of claims 1-8, characterized in that: It includes a racket body equipped with a friction sensing module and a handle equipped with a piezoelectric sensing module and connected to the racket body; The friction sensing module includes multiple copper foils and a cover layer disposed on the first end face of the cricket racket. The copper foils are evenly spaced apart. A signal acquisition circuit board is provided on the second end face of the cricket racket. Each copper foil is connected to the signal acquisition circuit board through a first wire. The cover layer is a thermal film, which covers the copper foils and wraps around the cricket racket. Multiple copper foils cover the striking area of ​​the cricket racket. When the cricket ball hits one of the copper foils, the signal acquisition circuit board identifies the local coordinates of the striking point through multi-channel voltage signals. Combined with the real-time posture data of the cricket racket captured by the measurement unit, the local coordinates are converted into global coordinates. ); The piezoelectric sensing module includes piezoelectric films and an insulating layer. The piezoelectric films include a first piezoelectric film, a second piezoelectric film, a third piezoelectric film, and a fourth piezoelectric film. The first and third piezoelectric films are located on the first end face of the handle, and the second and fourth piezoelectric films are located on the second end face of the handle. The first and second piezoelectric films are perpendicularly aligned, as are the third and fourth piezoelectric films. Each piezoelectric film has a silver-plated layer on both ends. Each piezoelectric film is connected to a signal acquisition circuit board via a second wire. The insulating layer is insulating tape cut to a preset shape and size. The insulating tape is wrapped around the four piezoelectric films and then wrapped around the handle.

Images