Data interaction system and method of handheld tablet personal computer

Abstract

The invention discloses a data interaction system and method for a handheld tablet personal computer, and the system comprises a touch sensing module which is provided with a pressure-sensitive layer and an electromagnetic resonance antenna array, and is used for collecting touch handwriting data; the multi-modal data processing engine integrates an NLP processing unit and a handwriting tracking unit, and the data processing engine executes a dynamic weighting algorithm; the deformable tactile feedback layer comprises a plurality of micro magneto-rheological actuator arrays and responds to the output of the data processing engine to generate real-time force field distribution; and the dynamic rendering optimization module is used for mapping the force field distribution to an adjustment factor of a screen refresh rate based on a Bayesian optimization rendering strategy generator. The touch sensing module can effectively predict and eliminate phase lag of a pen point pressure signal at different writing speeds through a composite attenuation function; space-time alignment compensation of touch, vision and posture data is realized by adopting a multi-channel data processing engine of a bidirectional attention mechanism, so that the interaction misoperation rate is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of data interaction, and more specifically, particularly relates to a data interaction system for a handheld tablet computer. At the same time, the present invention also relates to a data interaction method for a handheld tablet computer. Background Art

[0002] In the data interaction of handheld tablet computers, the problem of multi-physical field coupling mismatch has always been an issue to be solved. In high-speed writing scenarios, the mechanical response delay of traditional pressure sensors will cause a phase deviation between the touch trajectory and the actual handwriting; when the ambient light changes suddenly, the existing tactile dynamic adjustment mechanism is difficult to maintain the perceptual consistency of tactile feedback; during the multi-modal data processing, the time-domain misalignment between the visual rendering cycle and the force field physical response will cause user dizziness. Especially in three-dimensional gesture scenarios, the tactile feedback module lacks an adaptive compensation function for the device's holding posture, resulting in a significant misalignment between the force field distribution and the ergonomic requirements of the human body; In addition, the standard optimization algorithm for rendering parameters lacks the ability to dynamically correct in the face of personalized usage habits, making it difficult to balance touch accuracy and display effects among different user groups. Therefore, we propose a data interaction system and method for a handheld tablet computer. Summary of the Invention

[0003] The purpose of the present invention is to solve the shortcomings existing in the prior art, and to propose a data interaction system and method for a handheld tablet computer, realizing the three-dimensional collaborative optimization of high-resolution touch acquisition, low-latency data processing, and realistic physical feedback, thereby enhancing the experience of mobile interaction devices.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: A data interaction system for a handheld tablet computer, comprising: A touch perception module, configured with a pressure-sensitive layer and an electromagnetic resonance antenna array, for collecting touch handwriting data, and the expression is: ; where P(t) is the contact pressure change rate function, k is the dynamic attenuation factor, is the user's historical writing habit parameter, β is an adaptation coefficient of 0.8 - 1.2, is the dynamic attenuation term, used to eliminate the phase delay during high-speed writing; A multi-modal data processing engine, integrating an NLP processing unit and a handwriting tracking unit, and the data processing engine executes a dynamic weighting algorithm, and the expression is: ; where λ is the ambient light adaptation coefficient, S(x,y) is the continuous feature function of the touch point, ΔH is the change amount of the nib height, θx and θy represent the trajectory parameters in the spatio-temporal coordinate system, and σ is the empirical constraint value; A deformable tactile feedback layer, comprising a plurality of micro magnetorheological actuator arrays, generates a real-time force field distribution in response to the output of the data processing engine, and the expression is: ; where G v represents the spatial gradient value of the user's visual focus, is the current holding posture matrix of the device, represents the associated user visual attention distribution, represents the fusion of device attitude sensor data; A dynamic rendering optimization module, a rendering strategy generator based on Bayesian optimization, maps the force field distribution to an adjustment factor of the screen refresh rate, and the expression is: , where τ is the tactile sensitivity coefficient, F0 is the force field perception threshold, and F(x,y) represents the force field distribution.

[0005] Preferably, the piezovariable capacitance units of the pressure-sensitive layer are distributed in an 8×8 matrix, and each unit is configured with a self-calibration compensation circuit, and its real-time compensation value satisfies: ; where C0 is the reference capacitance value, and C b is a pair of temperature compensation coefficients, ΔT is the environmental temperature drift amount, τ c is the material thermal response time constant, represents the target capacitance value after dynamic non-linear temperature compensation correction; The dynamic attenuation factor k is dynamically generated by a double-layer LSTM network, and its iterative formula is: ; In the formula, represents the dynamic adjustment coefficient, σ is the Sigmoid activation function, W k and b k are parameter matrices, is the average value of the force field intensity at the previous moment.

[0006] Preferably, the micro magnetorheological actuator arrays form 8 groups of annular topological structures in three-dimensional space, and the driving voltages of each group of actuators satisfy an asymmetric constraint, and the expression is: ; In the formula, V i represents the actual driving voltage of the i-th group of actuators, V base is the basic parameter for adaptive adjustment of the real-time voltage of the battery, θ i is the polar angle of the i-th group of actuators relative to the center of the device, and is a spatial gradient field, reflecting the rate of change of the force field of the controlled object in the x and y directions.

[0007] A data interaction method for a palm tablet computer, which is implemented by the above system, and includes the following steps: S1. Scan to obtain touch input stream data, and the acquisition frequency is adjusted in real time according to the user's palm contact area S h to meet is the curvature eigenvalue of the touch trajectory, where is the dynamic sampling frequency, S max is the maximum contact area, t c is the curvature eigenvalue of the touch trajectory; S2. Use an improved convolution kernel to perform spatio-temporal normalization processing on the input data, where Γ is the handwriting coherence metric matrix, ρ is the data noise suppression factor, and H(u) is the Hermite polynomial basis function; S3. Construct a multi-channel attention matrix where W Q is the query parameter matrix, K T is the input key vector matrix, d is the feature dimension, and M haptic is the tactile feedback mask matrix; S4. Optimize the output timing and calculate the rendering delay compensation amount where δ max and δ min respectively represent the maximum and minimum rendering delay boundary values, t d is the current data processing time, t0 is the timing reference threshold, and η is the adaptive adjustment coefficient.

[0008] Preferably, in step S2, an adaptive noise suppression mechanism is introduced during the construction of the improved convolution kernel, and the real-time update rule expression is: ; In the formula, ρ 0 is the initial noise suppression factor, represents the Frobenius norm of the handwriting coherence metric matrix at the previous moment, is the spatio-temporal gradient tensor of the current touch data stream, κ is the critical threshold calculated according to the reading of the device temperature sensor, and Δ ρ is dynamically adjusted according to the user's grip pressure variance, ω and are the angular frequency and phase compensation amount associated with the sensor sampling rate, respectively.

[0009] Preferably, sequence prediction optimization is integrated in the calculation step of the rendering delay compensation amount, specifically: A motion vector prediction unit is installed at the front end of the GPU rendering pipeline. The change trend of the touch trajectory in five consecutive frames is analyzed through an optical flow algorithm to generate an inter-frame prediction compensation grid. The priority allocation strategy of the rendering thread is dynamically adjusted by monitoring the data of the device memory bandwidth. When the video memory occupancy rate exceeds the critical value, the multi-level cache compression mechanism is forcibly started to reduce the data throughput delay. A user behavior pattern atlas library is established. The pre-trained rendering parameter template is called according to the classification label of the current operation application, and the model deviation value is corrected in real time through a residual network.

[0010] Preferably, the calculation process of the rendering delay compensation amount in step S4 is embedded in the user behavior model, and the expression is: ; In the formula, μ is the user habit adjustment factor, D habit represents the KL divergence feature vector of the user usage pattern, ν is the scaling coefficient related to the device battery life status, G ref is the pre-calibrated device standard attitude quaternion matrix, F max is the maximum attitude deviation allowed by the system.

[0011] Preferably, the construction steps of the multi-channel attention matrix are provided with a hierarchical fusion mechanism, which specifically includes: During the generation stage of the tactile feedback mask matrix, the eye movement tracking data collected by the binocular camera is introduced to establish a dynamic mapping model between the visual focus heat map and the tactile sensitive area; When it is detected that the horizontal movement speed of the user's line of sight exceeds the preset threshold, a function constraint is imposed on the tactile feedback intensity of the edge area; Synchronously fuse the device tilt angle data collected by the gyroscope, calculate the physical coordinate system conversion factor of the touch trajectory using the quaternion interpolation method, and input the converted feature vector into the multi-layer perceptron to generate the spatial attention correction coefficient.

[0012] Preferably, the hierarchical fusion mechanism has the following expression: ; In the formula, σ represents the update gate activation function unique to the gated recurrent unit, W tactile is the tactile feature projection matrix, F accum is the time integral feature quantity of the historical force field distribution. ⊕ is defined as a mixed operation of the Hadamard product and the convolution operation, Z delay is the delay tensor for compensating the attention weight.

[0013] Technical effects and advantages of the present invention: A data interaction system and method for a palm-sized tablet computer provided by the present invention, compared with the prior art, the touch perception module of the present invention can effectively predict and eliminate the phase lag of the pen tip pressure signal at different writing speeds through a composite attenuation function; a multi-channel data processing engine adopting a bidirectional attention mechanism realizes spatio-temporal alignment compensation of tactile, visual and posture data, reducing the interactive misoperation rate; Secondly, the multi-modal data processing engine introduces autoregressive modeling of the environmental light adaptation coefficient, and can generate a personalized touch response curve according to the muscle movement characteristics of different users. Description of the Drawings

[0014] Figure 1 It is the architecture diagram of the data interaction system of the palm-sized tablet computer of the present invention; Figure 2 It is the flowchart of the data interaction method of the palm-sized tablet computer of the present invention. Detailed Embodiments

[0015] In order to make the objectives, technical solutions and advantages of the present invention clearer, the following further describes the present invention in detail with reference to specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0016] The present invention provides a data interaction system and method for a palm-sized tablet computer. Through touch perception optimization driven by a dynamic attenuation factor and multi-modal semantic weighted fusion technology, it systematically solves the key technical bottlenecks of traditional tablet devices in high-speed writing delay, environmental interference misrecognition, and tactile feedback lag; at the touch perception level, it utilizes the multi-physical field collaboration of a pressure-sensitive layer and an electromagnetic resonance antenna array, combines the dynamic attenuation factor with the user's historical writing habit parameters, significantly reduces the phase delay error, improves the touch response speed to the millisecond level, and adaptively matches the writing style differences of different users. In multi-modal data processing, through the dynamic weighting algorithm of the environmental light adaptation coefficient, the continuous feature function of the touch point and the spatio-temporal trajectory parameters, the misrecognition rate of handwriting in the scene of sudden light change is reduced, and at the same time, the semantic parsing accuracy of handwritten text is enhanced; As Figure 1 shown, the data interaction system of the palm-sized tablet computer includes: A touch perception module, configured with a pressure-sensitive layer and an electromagnetic resonance antenna array, for collecting touch handwriting data, and the expression is: ; where P(t) is the contact pressure change rate function, k is the dynamic attenuation factor, is the parameter of the user's historical writing habit, and β is the adaptation coefficient between 0.8 and 1.2. is the dynamic attenuation term, which is used to eliminate the phase delay during high-speed writing; Further, the piezoresistive capacitance units of the pressure-sensitive layer are distributed in an 8×8 matrix, and each unit is configured with a self-calibration compensation circuit, and its real-time compensation value satisfies: ; where C0 is the reference capacitance value, and C b are the temperature compensation coefficient pair, ΔT is the environmental temperature drift, and τ c is the material thermal response time constant, represents the target capacitance value after dynamic non-linear temperature compensation correction; The dynamic attenuation factor k is dynamically generated by a double-layer LSTM network, and its iterative formula is: ; In the formula, represents the dynamic adjustment coefficient, σ is the Sigmoid activation function, W k and b k are the parameter matrices, is the average value of the force field intensity at the previous moment.

[0017] The multi-modal data processing engine integrates the NLP processing unit and the handwriting tracking unit. The data processing engine executes the dynamic weighting algorithm, and the expression is: ; where λ is the environmental light adaptation coefficient, S(x,y) is the continuous feature function of the touch point, ΔH is the change in the nib height, θx and θy represent the trajectory parameters in the spatio-temporal coordinate system, and σ is the empirical constraint value; The deformable tactile feedback layer includes multiple micro-magnetorheological actuator arrays, and generates a real-time force field distribution in response to the output of the data processing engine. The expression is: ; where G v represents the spatial gradient value of the user's visual focus, is the current holding posture matrix of the device, represents the associated user visual attention distribution, represents the fusion of device attitude sensor data; It should be noted that the micro-magnetorheological actuator arrays form 8 groups of annular topological structures in the three-dimensional space, and the driving voltages of each group of actuators satisfy the asymmetric constraint. The expression is: ; In the formula, V i represents the actual driving voltage of the i-th group of actuators, V baseis the basic parameter for the adaptive adjustment of the real-time voltage of the battery, θ i is the polar angle of the i-th actuator relative to the device center, and is the spatial gradient field, reflecting the rate of change of the force field of the controlled object in the x and y directions.

[0018] The dynamic rendering optimization module, a rendering strategy generator based on Bayesian optimization, maps the force field distribution to the adjustment factor of the screen refresh rate, and the expression is: , where τ is the tactile sensitivity coefficient, F0 is the force field perception threshold, and F(x, y) represents the force field distribution.

[0019] This embodiment also proposes a data interaction method for a palm-sized tablet computer, which is implemented by using the above system, as Figure 2 shown, including the following steps: S1. Scan to obtain touch input stream data, and the acquisition frequency is adjusted in real time according to the user's palm contact area S h to satisfy is the curvature eigenvalue of the touch trajectory, where, is the dynamic sampling frequency, S max is the maximum contact area, t c is the curvature eigenvalue of the touch trajectory; S2. Use the improved convolution kernel to perform spatio-temporal normalization processing on the input data, where Γ is the handwriting coherence metric matrix, ρ is the data noise suppression factor, and H(u) is the Hermite polynomial basis function; Among them, in step S2, an adaptive noise suppression mechanism is introduced during the construction of the improved convolution kernel, and the real-time update rule expression is: ; In the formula, ρ 0 is the initial noise suppression factor, represents the Frobenius norm of the handwriting coherence metric matrix at the previous moment, is the spatio-temporal gradient tensor of the current touch data stream, κ is the critical threshold calculated according to the readings of the device temperature sensor, Δ ρ is dynamically adjusted according to the variance of the user's grip pressure, ω and are the angular frequency and phase compensation amount associated with the sensor sampling rate, respectively.

[0020] S3. Construct a multi-channel attention matrix where, W Q is the query parameter matrix, K Tis the input key vector matrix, d is the feature dimension, and M haptic is the tactile feedback mask matrix; Specifically, the construction steps of the multi-channel attention matrix add a hierarchical fusion mechanism, which specifically includes: Introduce the eye movement tracking data collected by the binocular camera in the tactile feedback mask matrix generation stage, and establish a dynamic mapping model between the visual focus heat map and the tactile sensitive area; When it is detected that the horizontal movement speed of the user's line of sight exceeds the preset threshold, apply a function constraint to the tactile feedback intensity of the edge area; Synchronously fuse the device tilt angle data collected by the gyroscope, calculate the physical coordinate system conversion factor of the touch trajectory using the quaternion interpolation method, and input the converted feature vector into the multi-layer perceptron to generate the spatial attention correction coefficient; The hierarchical fusion mechanism, and its expression is: ; In the formula, σ represents the update gate activation function unique to the gated recurrent unit, and W tactile is the tactile feature projection matrix, and F accum is the time integral feature quantity of the historical force field distribution. ⊕ is defined as a mixed operation of the Hadamard product and the convolution operation, and Z delay is the delay tensor for compensating the attention weight.

[0021] S4. Optimize the output timing and calculate the rendering delay compensation amount Among them, δ max and δ min respectively represent the maximum and minimum rendering delay boundary values, t d is the current data processing time consumption, t0 is the timing reference threshold, and η is the adaptive adjustment coefficient; The calculation steps of the rendering delay compensation amount integrate sequence prediction optimization, specifically: Set up a motion vector prediction unit at the front end of the GPU rendering pipeline, analyze the change trend of the continuous five-frame touch trajectory through the optical flow algorithm, and generate an inter-frame prediction compensation grid; Dynamically adjust the priority allocation strategy of the rendering thread using the device memory bandwidth monitoring data. When the video memory occupancy rate exceeds the critical value, force the start of the multi-level cache compression mechanism to reduce the data throughput delay; Establish a user behavior pattern atlas library, call the pre-trained rendering parameter template according to the classification label of the current operation application, and correct the model deviation value in real time through the residual network; The calculation process of the rendering delay compensation amount embeds the user behavior model, and the expression is: ; In the formula, μ is the user habit adjustment factor, Dhabit KL divergence feature vector characterizing the user usage pattern ν is the scaling coefficient related to the device battery life status, G ref is the pre-calibrated device standard attitude quaternion matrix, F max is the maximum attitude deviation allowed by the system.

[0022] In summary, through the real-time calculation formula of the palm contact area and the trajectory curvature, the data acquisition frequency is dynamically adjusted to double the sampling density of high-curvature strokes; An improved convolution kernel introducing Hermite basis functions is adopted, combined with a noise suppression factor and a handwriting coherence matrix to effectively separate the writing action and noise, and the signal-to-noise ratio optimization rate is improved; through the collaborative control of the tactile feedback mask matrix and the predictive rendering delay compensation amount, the tactile-visual interaction delay is compressed and the comprehensive energy consumption is reduced.

[0023] This solution has achieved innovative breakthroughs in touch accuracy, interaction response, energy consumption efficiency, and multimodal adaptability, providing a highly reliable and low-latency software and hardware integrated solution for intelligent interaction of mobile terminals.

[0024] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data interaction system for a handheld tablet computer, characterized in that, Including: A touch sensing module, configured with a pressure - sensitive layer and an electromagnetic resonance antenna array, for collecting touch handwriting data; A multi - modal data processing engine, integrating an NLP processing unit and a handwriting tracking unit, and the data processing engine executes a dynamic weighting algorithm; A deformable haptic feedback layer, containing multiple micro - magnetorheological actuator arrays, generating a real - time force field distribution in response to the output of the data processing engine; A dynamic rendering optimization module, a rendering strategy generator based on Bayesian optimization, mapping the force field distribution to an adjustment factor of the screen refresh rate.

2. The data interaction system of a handheld tablet computer according to claim 1, characterized in that, The piezovariable capacitance units of the pressure - sensitive layer are distributed in an 8×8 matrix, and each unit is configured with a self - calibration compensation circuit.

3. The data interaction system of a handheld tablet computer according to claim 1, characterized in that, The micro - magnetorheological actuator arrays form 8 groups of annular topologies in three - dimensional space, and the driving voltages of each group of actuators satisfy an asymmetric constraint, and the expression is: ； Where, V i represents the actual driving voltage of the i-th group of actuators, V base is the basic parameter for the adaptive adjustment of the real-time battery voltage, θ i is the polar angle of the i-th group of actuators relative to the center of the device, and is the spatial gradient field, reflecting the rate of change of the force field of the controlled object in the x and y directions.

4. A data interaction method for a handheld tablet computer, characterized in that, The method is implemented by using the system according to any one of claims 1 - 3, and includes the following steps: S1. Scan to obtain touch input stream data, and the acquisition frequency is adjusted in real time according to the user's palm contact area S to meet h where is the curvature eigenvalue of the touch trajectory, and among them, is the dynamic sampling frequency, S max is the maximum contact area, and t c is the curvature eigenvalue of the touch trajectory; S2. Perform spatio - temporal normalization processing on the input data by using an improved convolutional kernel; S3. Construct a multi - channel attention matrix; S4. Optimize the output timing and calculate the rendering delay compensation amount.

5. A data interaction method for a handheld tablet computer according to claim 4, characterized in that, In step S2, during the construction process of the improved convolutional kernel, an adaptive noise suppression mechanism is introduced, and the real - time update rule expression is: ； wherein, ρ 0 is the initial noise suppression factor, denotes the Frobenius norm of the handwriting coherence metric matrix at the previous moment, is the spatio-temporal gradient tensor of the current touch data stream, κ is the critical threshold calculated based on the device temperature sensor reading, Δ ρ is dynamically adjusted according to the user's holding pressure variance, ω and are the angular frequency and phase compensation amount associated with the sensor sampling rate, respectively.

6. The data interaction method of a handheld tablet computer according to claim 4, characterized in that In the calculation step of the rendering delay compensation amount, sequence prediction optimization is integrated, specifically: A motion vector prediction unit is installed at the front end of the GPU rendering pipeline, and the change trend of the touch trajectory of five consecutive frames is analyzed by using an optical flow algorithm to generate an inter - frame prediction compensation grid; The priority allocation strategy of the rendering thread is dynamically adjusted by using the device memory bandwidth monitoring data. When the video memory occupancy rate exceeds the critical value, a multi - level cache compression mechanism is forcibly started to reduce the data throughput delay; A user behavior pattern atlas library is established, a pre - trained rendering parameter template is called according to the classification label of the current operation application, and the model deviation value is corrected in real time through a residual network.

7. A data interaction method for a handheld tablet computer according to claim 6, characterized in that In step S4, the user behavior model is embedded in the calculation process of the rendering delay compensation amount, and the expression is: ； In the formula, μ is the user habit adjustment factor, D habit represents the KL divergence feature vector of the user usage pattern, ν is the scaling factor related to the device battery life status, G ref is the pre-calibrated device standard attitude quaternion matrix, F max is the maximum attitude deviation allowed by the system.

8. A data interaction method for a handheld tablet computer according to claim 4, characterized in that In the construction step of the multi - channel attention matrix, a hierarchical fusion mechanism is added, specifically including: During the generation stage of the haptic feedback mask matrix, the eye movement tracking data collected by a binocular camera is introduced to establish a dynamic mapping model between the visual focus heat map and the haptic sensitive area; When it is detected that the horizontal movement speed of the user's line of sight exceeds the preset threshold, a function constraint is imposed on the haptic feedback intensity of the edge area; Synchronously fuse the device tilt angle data collected by a gyroscope, calculate the physical coordinate system conversion factor of the touch trajectory by using a quaternion interpolation method, and input the converted feature vector into a multi - layer perceptron to generate a spatial attention correction coefficient.

9. A data interaction method for a handheld tablet computer according to claim 8, characterized in that, The hierarchical fusion mechanism, its expression is: ； In the formula, σ represents the update gate activation function unique to the gated recurrent unit, W tactile is the tactile feature projection matrix, F accum is the time-integrated feature quantity of the historical force field distribution, ⊕ is defined as a mixed operation of the Hadamard product and the convolution operation, Z delay is the delay tensor for compensating the attention weight.

Images