Self-Powered Non-Contact Interaction System and Method Based on Turbulence-Modulated Wet Electricity Effect

Abstract

This invention provides a self-powered non-contact interactive system and method based on turbulence-modulated wet electrostatic effect, comprising: a signal acquisition module for acquiring dynamic electrical signals from a self-powered non-contact interactive device; a signal preprocessing module for preprocessing the dynamic electrical signals to obtain preprocessing results; a recognition module for recognizing the signals from the preprocessing results to obtain recognition results corresponding to the motion causing the dynamic electrical signals; and a control output module for outputting corresponding control commands based on the recognition results. The interaction process provided by this invention requires no physical contact, avoiding mechanical wear and cross-contamination risks, and is particularly suitable for public or medical settings with high hygiene requirements.

Description

Technical Field

[0001] This invention belongs to the field of human-computer interaction technology, specifically relating to a self-powered non-contact interaction system and method based on turbulence-modulated wet electrostatic effect. More specifically, it is a self-powered non-contact human-computer interaction system, device, and method based on air turbulence-modulated wet electrostatic effect. Background Technology

[0002] Human-machine interfaces (HMIs) are a key component in expanding the functionality of augmented reality (AR) and virtual reality (VR) systems. They provide users with a variety of interactive experiences through feedback from hand or fingertip movements. These interfaces are not only an important foundation for user-oriented entertainment applications but also show broad application prospects in fields such as surgical robots, rehabilitation training, information interaction, and military training. By enabling users to execute commands using simple gestures, HMIs can significantly improve operational efficiency and user experience.

[0003] Currently, HMI interactive feedback technologies can be mainly divided into two categories: contact devices and non-contact devices. Contact systems, such as those utilizing force feedback tactile effects, are convenient to use, but due to the physical contact involved, they are prone to hygiene risks and mechanical wear problems.

[0004] In contrast, contactless systems based on mechanisms such as capacitance changes, photonic effects, and thermoelectric effects can avoid direct contact, thus mitigating the aforementioned problems. However, a continuous and reliable power supply is still required for these devices to operate stably over extended periods. Traditional battery systems, such as lithium-ion batteries, are typically rigid, bulky, and require frequent replacements, increasing system integration complexity and potentially posing safety hazards such as leaks or explosions. In recent years, environmental energy harvesting technologies such as solar, wind, and mechanical energy have gained attention as sustainable alternatives, but they often face challenges such as insufficient device configuration compatibility and complex system architecture.

[0005] Therefore, developing self-powered energy harvesting systems with non-contact sensing capabilities is crucial. This can improve system compatibility, reduce structural complexity, decrease maintenance costs, and eliminate safety risks. Such systems promise to utilize ambient energy for interactive feedback while avoiding the drawbacks of traditional power supply methods.

[0006] Among renewable energy sources, ambient moisture is an abundant and recyclable energy source, and one of the largest natural energy reservoirs on Earth. Moisture-driven electricity generation (MEG) technology offers a promising solution for next-generation self-powered electronics. This technology utilizes the interaction between hygroscopic materials and moisture to generate energy through the migration of free carriers, which can then be harvested and converted into electricity.

[0007] Current wet electrical technologies primarily focus on generating stable, continuous direct current (DC) output, aiming to serve as a miniature power source. Researchers mainly enhance electrical output by modulating functional groups to improve water absorption capacity, increase ion migration efficiency, and optimize interfacial charge migration characteristics. These moisture-induced electrical systems can be used to power external devices or as self-powered sensors. However, under stable moisture flow conditions, these devices typically output humidity-related DC signals, posing a challenge in sensing and recognizing different feedbacks corresponding to various complex commands in HMI (Human-Machine Interface) scenarios. Specifically, the signal response of these technologies is usually slow, and the output signal characteristics are singular, making it difficult to effectively encode and recognize dynamic information generated by rapid and complex user movements, such as gestures and air writing. This results in existing wet electrical technologies failing to meet the requirements of high dynamic response and high signal discrimination in advanced human-computer interaction scenarios.

[0008] Therefore, there is an urgent need for a new type of interactive device and method that can operate in an air environment, requires no external power supply, can perform non-contact sensing, and has output signals with sufficient discriminative power to support intelligent algorithm recognition. Summary of the Invention

[0009] To address the shortcomings of existing technologies, the purpose of this invention is to provide a self-powered non-contact interactive system and method based on turbulence-modulated wet electrostatic effect.

[0010] A self-powered non-contact interactive system based on turbulence-modulated wet electrostatic effect, according to the present invention, comprises: Signal acquisition module: Acquires dynamic electrical signals from self-powered contactless interactive devices; Signal preprocessing module: preprocesses the dynamic electrical signal to obtain the preprocessing result; Recognition module: Recognizes the signal of the preprocessed result, and then obtains the recognition result corresponding to the motion that caused the dynamic electrical signal; Control output module: Based on the recognition results, output corresponding control commands.

[0011] Preferably, the self-powered non-contact interactive device includes: a first electrode, a second electrode, a wet electrostatic functional layer disposed between the first electrode and the second electrode, and an airflow disturbance structure; the airflow disturbance structure is disposed on the surfaces of the first electrode and the second electrode; when an external object moves non-contactly within a predetermined distance from the surface of the device, local air turbulence is formed on the surface of the wet electrostatic functional layer, thereby outputting a dynamic electrical signal related to the object's movement.

[0012] Preferably, the airflow disturbance structure is a ventilation structure disposed on the first electrode and the second electrode; the ventilation structure includes a through-hole structure or a mesh structure.

[0013] Preferably, the wet electrostatic functional layer is a porous polymer network material; the porous polymer network material comprises a main-chain polymer, additives with dissociable groups, and a migratable ionic salt; The main chain polymer includes polyglycolic acid, polyvinyl alcohol, polyacrylic acid, cellulose-based polymers, alginate, gelatin, or any combination thereof; The dissociable groups include one or more of carboxyl, phosphate, phenolic hydroxyl, and sulfonic acid groups; The components with the dissociable groups include: tannic acid, citric acid, phytic acid, or any combination thereof; The migratable ionic salt comprises a metal cation and a corresponding anion; the metal cation includes Na. + K + Li + Ca² + Mg² + One or more of them.

[0014] Preferably, the predetermined range is 1 mm to 120 mm.

[0015] Preferably, in the signal preprocessing module, the preprocessing includes: filtering, resampling, normalization, differential operation, or frequency domain transformation; In the recognition module, the signal of the preprocessed result is input into a preset machine learning model, and then the corresponding recognition result is output; the machine learning model includes a support vector machine model or a convolutional neural network model; the recognition result includes one or more of the following: numbers, characters, words, gesture categories or control commands.

[0016] Preferably, the self-powered contactless interactive device is arranged in a multi-channel array with 2 to 16 channels.

[0017] The present invention provides a self-powered non-contact interaction method based on turbulence-modulated wet electrostatic effect, implemented based on a self-powered non-contact interaction system based on turbulence-modulated wet electrostatic effect, comprising: The signal acquisition module acquires dynamic electrical signals from the self-powered contactless interactive device. The signal preprocessing module preprocesses the dynamic electrical signal to obtain the preprocessing result. The recognition module identifies the signal of the preprocessed result, thereby obtaining the recognition result corresponding to the motion that caused the dynamic electrical signal; The control output module outputs corresponding control commands based on the recognition results.

[0018] Preferably, the preprocessing includes: filtering, resampling, normalization, differential operation, or frequency domain transformation.

[0019] Preferably, the identification result is a classification of the motion.

[0020] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention utilizes the spontaneous generation of electricity from water, which is widely present in the environment, to drive the entire sensing and interaction process. No external power supply or built-in battery is required, which significantly reduces the complexity, size and maintenance cost of the system.

[0021] 2. The interactive process provided by this invention does not require physical contact, avoiding mechanical wear and cross-contamination risks of equipment, and is particularly suitable for public or medical scenarios with high hygiene requirements.

[0022] 3. This invention uses air turbulence induced by external motion to dynamically modulate the wet electrical signal through the dual physical factors of "local humidity change" and "local air pressure disturbance". Combined with the design of the airflow disturbance structure, it generates dynamic signal waveforms with rich features and high repeatability, fundamentally solving the problem of traditional wet electrical technology having a single signal and being unable to recognize complex dynamic commands.

[0023] 4. The dynamic electrical signal output by this invention has rich temporal characteristics, making it suitable as input for machine learning models. It can achieve high-precision recognition of complex commands such as gestures and air writing. Furthermore, this invention has a simple structure, is easy to integrate, and its device is a simple sandwich structure with low manufacturing cost. It is easy to miniaturize and make flexible, and can be easily expanded into a multi-channel array to improve recognition performance and robustness. Attached Figure Description

[0024] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 A schematic diagram of the overall framework of the self-powered contactless interaction system provided by the present invention; Figure 2 This is a schematic diagram of the interactive device structure provided by the present invention; Figure 3This is a schematic diagram of the top electrode through-hole / mesh structure and its turbulence effect provided by the present invention, wherein a represents the top electrode through-hole / mesh structure and b represents a schematic diagram of the turbulence effect of the top electrode through-hole / mesh structure. Figure 4 The diagram shows the dynamic electrical signal waveform output by the device provided by the present invention, as well as the flowchart for distinguishing different gestures and signal preprocessing and machine learning recognition, where a~f represent the labels of the overall process sequence. Figure 5 This is an application diagram provided by the present invention, where a~g represent the numbers of the application process sequence. Detailed Implementation

[0025] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the scope of protection of the present invention.

[0026] This invention relates to the fields of self-powered sensing, wet power generation and intelligent recognition technology, specifically to a human-computer interaction system, device and method that uses local humidity and air pressure disturbances caused by air turbulence to modulate the wet power output, thereby realizing non-contact motion / gesture / air writing recognition without external power supply.

[0027] The device includes a hygroscopic ion-absorbing porous polymer wet electrostatic functional layer disposed between two electrodes, the electrodes comprising an array of through-holes or a mesh structure. The device generates a basic potential output under the influence of air humidity; when an external object moves non-contactly, it induces local air turbulence, causing local humidity changes and air pressure disturbances on the device surface, thereby modulating the ion dissociation and migration process and outputting a dynamic electrical signal related to the motion trajectory. The system, through signal acquisition and preprocessing, inputs the dynamic electrical signal into a machine learning model to obtain gesture / writing / command recognition results, and outputs control commands to achieve information input, encrypted interaction, or remote control.

[0028] Specifically, the device provided by this invention includes a wet electrostatic functional layer disposed between two electrodes and an airflow disturbance structure. Its working principle is as follows: when an external object moves without contact, the local air turbulence it induces modulates the ion migration process within the wet electrostatic functional layer through the dual physical factors of causing local humidity changes and local air pressure disturbances, thereby outputting a dynamic electrical signal with rich motion-related characteristics.

[0029] This invention comprises a device, a system for signal acquisition, preprocessing, recognition, and control output modules, and a corresponding recognition method. Utilizing ambient humidity for self-powered operation, this invention requires no external power source. Through an innovative two-factor modulation principle, it generates a highly discriminative signal, enabling high-precision, non-contact intelligent recognition of complex commands such as air writing. The structure is simple, safe, and hygienic. This invention requires no external power source, operates non-contactly, and provides highly discriminative output signals, making it suitable for scenarios such as VR / AR interaction, intelligent medical care, and robot control.

[0030] Example 1: This example provides a basic implementation scheme for a self-powered non-contact interactive device based on air turbulence modulation, and elaborates on its structure, preparation process and working principle in detail.

[0031] Users interact with a self-powered contactless interactive device (hereinafter referred to as the interactive device) through non-contact limb movements, such as handwriting in the air. The user's movements induce localized air turbulence near the surface of the interactive device. This air turbulence dynamically modulates the electrostatic effect within the interactive device by simultaneously altering two physical factors: local pressure P and local relative humidity RH, thereby generating a dynamic electrical signal highly correlated with the user's movement trajectory. This signal is then acquired, analyzed, and identified by a signal acquisition and processing unit, ultimately enabling human-computer interaction. This embodiment will focus on describing the specific structure of the interactive device.

[0032] Specifically, the interactive device adopts a sandwich design in terms of structure. Its core includes a wet-electric functional layer, and a first electrode and a second electrode respectively disposed on both sides of the wet-electric functional layer.

[0033] In this embodiment, the preparation process of the wet electrical functional layer may include the following steps: First, a mixed solution is prepared. Specifically, polyglycolic acid is used as the host polymer, which provides a porous polymer network framework. Then, tannic acid is added to the solution. Tannic acid molecules are rich in dissociable functional groups, such as carboxyl or phenolic hydroxyl groups, which can dissociate into cations in the presence of moisture, serving as a source of charge. Furthermore, sodium chloride is incorporated into the solution as a migratory ionic salt to provide additional mobile charge carriers and enhance ionic conductivity. To further optimize ion migration channels and increase the hygroscopic properties of the material, a small amount of cellulose nanofibers can be selectively added as a functional filler to help construct a more regular porous network structure. The above components are then fully dispersed in deionized water and subjected to stirring or ultrasonic treatment to form a homogeneous mixed solution.

[0034] Subsequently, a film-forming step is performed. The prepared mixed solution is drop-coated onto a substrate serving as the first electrode. As a specific implementation, the first electrode can be a piece of carbon cloth with good conductivity and porosity. After drop-coating, the carbon cloth with the attached solution is placed in an oven at 60°C for drying until the solvent completely evaporates, thereby forming a wet electrostatic functional layer film with a thickness of approximately 100 micrometers on the surface of the carbon cloth. It should be noted that this thickness range, for example, 50 to 150 micrometers, can maintain a relatively fast response speed while ensuring sufficient moisture absorption.

[0035] Next, the device is assembled. Another piece of carbon cloth is used as the second electrode. Unlike the flat first electrode, this second electrode has a perforated structure serving as an airflow disturbance mechanism. Specifically, a 3×3 array of perforated structures can be fabricated on the carbon cloth of the second electrode using laser cutting or precision stamping. The diameter of each perforation can be set to 2 mm, and the center-to-center spacing between the perforations can be set to 5 mm. This perforated array design effectively creates localized turbulence on the device surface. Finally, this second electrode with the perforated structure is placed over the already formed wet-electric functional layer, sandwiching the wet-electric functional layer between the first and second electrodes, forming a complete "carbon cloth-wet-electric functional layer-perforated carbon cloth" sandwich structure. To facilitate signal extraction, conductive silver paste is used to connect wires to the first and second electrodes, respectively, serving as the device's signal output terminals.

[0036] Under normal ambient humidity conditions, the wet electrostatic functional layer absorbs water molecules (H2O) from the air. These water molecules solubilize dissociable groups on the polymer backbone, such as the carboxyl groups of tannic acid, causing them to dissociate and release mobile cations, such as hydrogen ions (H+). + And leaves fixed dissociated groups -COO on the polymer backbone. - Due to ion concentration gradients that may form within the material itself or during the preparation process, these mobile cations undergo directional migration within the functional layer, thereby establishing a stable base potential between the first and second electrodes. When external airflow disturbances occur, such as a momentary drop in local humidity on the device surface caused by a user's finger tracing the surface (i.e., a dehumidification process), according to the principle of chemical equilibrium shift, this process will induce reverse ionization, releasing the mobile cations H+. + It will re-interact with the fixed, dissociated group -COO - These components combine to form unionized groups. This process causes drastic changes in the local concentration and distribution of mobile charge carriers, resulting in dynamic changes in the device's output potential.

[0037] When a smooth external airflow passes over an electrode with through-holes, significant vortices and turbulence zones form below the through-holes, on the surface of the wetted electrostatic functional layer. This demonstrates that the through-hole structure can effectively transform macroscopic, smooth external airflow into microscopic, complex local turbulence. This turbulence greatly enhances the interaction between the airflow and the functional layer surface, forming the basis for achieving efficient signal modulation.

[0038] The core working principle of this device, namely the "two-factor modulation principle," is illustrated in the following diagram. Figure 4 As shown. When an external object, such as a user's finger, moves within a certain range from the device surface—for example, 10 to 80 millimeters, a range that supports non-contact interactive movement over relatively long distances—its movement stirs up the surrounding air, inducing localized air turbulence on the device surface. This single physical phenomenon, turbulence, simultaneously triggers two parallel physical effects that contribute to the signal: Localized humidity changes: The occurrence of turbulence accelerates the evaporation of moisture on the surface of the electro-hydraulic functional layer, resulting in a momentary, concave drop in localized relative humidity. This change in localized moisture content directly modulates the ion dissociation equilibrium within the functional layer.

[0039] Local pressure disturbances: The formation, development, and dissipation of turbulence are inherently accompanied by local air pressure fluctuations. Typically, when an object passes by, the surface of the device experiences a momentary, pulse-like pressure disturbance. This pressure change may cause minute mechanical deformations in the porous polymer network, or affect the ion migration process through piezoelectric effects and by influencing the effective cross-section of ion migration channels.

[0040] Ultimately, the local humidity changes and air pressure disturbances caused by local air turbulence work together to dynamically modulate the ion migration process within the wet electro-functional layer of the interactive device. This dual modulation mechanism enables the waveform characteristics of the output dynamic electrical signal, such as peak value, valley value, width, and slope, to establish a highly repeatable correlation with parameters such as the trajectory, speed, and distance of external objects.

[0041] In practical testing, the interactive device prepared in this embodiment can be placed in a normal air environment with a relative humidity of 40% to 80%, without any external power supply. A voltage acquisition device with high input impedance, such as an electrometer or source meter, is connected to the two electrodes of the device. When a tester's finger moves across the perforated electrode surface of the device at a speed of approximately 0.5 m / s within a range of 10 mm to 80 mm, a dynamic voltage pulse signal with a peak value of tens of millivolts can be clearly recorded.

[0042] Experimental results show that different trajectories, such as straight lines, circles, S-shapes, or Z-shapes, can all produce voltage waveforms with different shapes but high repeatability. This fully demonstrates that the device provided in this embodiment can successfully encode macroscopic non-contact motion into unique and highly distinguishable electrical signal patterns, laying the foundation for subsequent intelligent recognition.

[0043] Example 2 illustrates that the technical solution proposed in this invention has broad material applicability, and its core working principle is not limited to the specific chemical material system in Example 1. By replacing the main polymer of the wet electrofunctional layer, the source of the dissociable groups and the ionic salt, as well as the electrode material, the same technical effect can still be achieved, and a flexible interactive device can be prepared.

[0044] As an alternative implementation, this embodiment modifies the material system of the wet electrostatic functional layer. Specifically, a mixture of polyvinyl alcohol and sodium alginate is used as the main polymer, both of which have good hydrophilicity and film-forming properties, and are inexpensive. Citric acid is chosen as the source of dissociable groups, as the citric acid molecule contains multiple carboxyl groups, which can provide abundant dissociable protons. Potassium chloride is chosen as a migratable ionic salt. The selection of these materials all fall within the preferred material range disclosed in this invention. The preparation process is similar to that of Example 1, where the above components are dissolved in water to form a mixed solution.

[0045] Accordingly, the electrode materials are also modified to achieve device flexibility. In this embodiment, flexible conductive fabrics, such as nylon fabrics with silver or carbon nanotube coatings, are selected as the first and second electrodes. This electrode material not only has good conductivity but also excellent bending and stretching properties.

[0046] As an airflow disturbance structure, this embodiment fabricates a mesh structure on the flexible conductive fabric serving as the second electrode, replacing the through-hole array in Embodiment 1. For example, a conductive mesh with a line width of 1 mm and a mesh size of 3 mm × 3 mm can be formed on the fabric by screen printing conductive paste or laser selective ablation. It is understood that this mesh structure can also effectively generate local turbulence when airflow passes through it, and its function is equivalent to that of the through-hole array.

[0047] The assembly process of the device may include: uniformly coating a mixed solution of the wet electrostatic functional layer onto a flexible conductive fabric serving as the first electrode, and drying it into a film at a suitable temperature, such as 60°C. Then, a second electrode with a mesh structure, i.e., the flexible conductive fabric, is placed over the functional layer, and the electrode is led out using flexible conductive tape or conductive silver paste. Thus, a non-contact interactive device entirely composed of flexible materials is obtained.

[0048] In performance testing, the flexible interactive device can be attached to various curved surfaces, such as a tester's arm, the wall of a water glass, or irregular object models. Even when bent, the device operates stably and outputs motion-related, clearly characteristic dynamic electrical signals when an object, such as a finger, moves in its vicinity. For example, swiping left and right in front of the device attached to the arm produces approximately mirror-symmetrical voltage waveforms from two swipes in opposite directions, demonstrating the distinguishability of its signals.

[0049] The results of this embodiment confirm that the technical solution proposed in this invention has a very wide range of material choices, including but not limited to various main-chain polymers such as polyglycolic acid, polyvinyl alcohol, polyacrylic acid, cellulose-based polymers, alginate, and gelatin; various additives containing dissociable groups such as tannic acid, citric acid, and phytic acid; and migratable ionic salts containing various cations such as sodium ions, potassium ions, and lithium ions. Furthermore, the device is easy to design flexibly, which greatly expands its application scenarios in wearable devices, flexible electronic skin, smart textiles, and other fields.

[0050] Example 3 provides a variant structure of an interactive device, which aims to further clarify the technical concept of "airflow disturbance structure", that is, to prove that the structure used to generate turbulence can be separated from the electrode itself and exist as an independent component.

[0051] In this embodiment, the preparation method of the core wet electrical part of the interactive device is basically the same as that in Example 1, that is, the wet electrical functional layer is prepared by using materials such as polyglycolic acid, tannic acid and sodium chloride.

[0052] However, unlike Embodiment 1 and Embodiment 2, the first and second electrodes used in this embodiment are both flat carbon cloth electrodes without any openings or grids.

[0053] The device assembly also employs a sandwich structure of "first electrode - wet-electric functional layer - second electrode". The key difference lies in the additional layer of independent, non-conductive plastic mesh covering the second electrode of the assembled sandwich device, i.e., the side facing the user. This plastic mesh can be a mesh cut from a commercially available ordinary window screen and has no electrical function itself. During installation, tiny pads are placed at the edges to maintain a small gap of approximately 0.5 mm between the plastic mesh and the surface of the second electrode below. This plastic mesh constitutes the "airflow disturbance structure" in this embodiment.

[0054] The device operates as follows: when a user's finger makes a non-contact movement above the uppermost plastic mesh, the induced airflow, upon passing through the mesh, generates complex turbulence due to the mesh's obstruction and guiding effect. This turbulence passes through the tiny gap between the plastic mesh and the second electrode, acting on the surface of the underlying wet-electric device—the outer surface of the second electrode—and is transmitted to the wet-electric functional layer through the porous carbon cloth electrode. This localized turbulence acting on the device surface also causes dynamic changes in local humidity and pressure, thereby modulating the ion migration process within the wet-electric functional layer, ultimately outputting a dynamic electrical signal related to finger movement between the first and second electrodes.

[0055] Experimental tests show that the device using this split airflow disturbance structure can also effectively convert non-contact motion into a distinguishable electrical signal. Although the signal amplitude may be slightly lower than that of the electrode-integrated structure due to energy loss during turbulence propagation, the correspondence between the signal waveform and the motion trajectory is still clearly discernible.

[0056] The results of this embodiment demonstrate that the core of the technical solution of this invention lies in the physical process of "using local air turbulence for two-factor modulation," rather than the mandatory requirement to physically integrate the turbulence-generating structure onto the electrodes. The airflow disturbance structure can be a standalone, or even a non-functional, additional component. This conclusion provides a broader scope of protection for the technical solution of this invention, covering all structural designs capable of achieving the function of "forming local air turbulence on the surface of the wet electrostatic functional layer."

[0057] Example 4 demonstrates how a single interactive device can be expanded into a multi-channel array and combined with signal processing and machine learning algorithms to construct a complete self-powered contactless human-computer interaction system, enabling high-precision recognition of complex commands such as handwritten digits in the air. This example will be described in detail from both system and method perspectives.

[0058] The system in this embodiment includes a multi-channel self-powered contactless interactive device, a signal acquisition module, a signal preprocessing module, an identification module, and a control output module.

[0059] First, a device array is constructed. Three individual interactive device units prepared according to the method in Example 1 are arranged side by side to form a 1×3 linear channel array. Each device unit serves as an independent sensing channel, and the spacing between them can be set to, for example, 1 cm. As a preferred embodiment, this array design containing multiple sensing channels can synchronously capture signals generated by the same motion from different locations in space, providing richer spatial dimensional information, thereby greatly improving the accuracy and robustness of recognition.

[0060] Secondly, the system hardware is integrated. The signal output terminals of the three channels (i.e., the three device units), each connected by two wires, are connected to a multi-channel signal acquisition module. The core of this module is a high-impedance analog-to-digital converter (ADC) with a high input impedance (e.g., greater than 1 gigaohm) and high resolution (e.g., 16-bit).

[0061] The high input impedance ensures accurate measurement of the weak voltage signals generated by the wet electrical device without excessive current shunting, while the high resolution guarantees precise capture of signal details. The analog-to-digital converter is controlled by a microcontroller, such as an STM32 series microcontroller, which periodically acquires data from each channel and transmits the data in real-time to a computer via a serial interface, such as USB.

[0062] Next, the software program running on the computer implemented the functions of signal preprocessing, identification, and control output.

[0063] Step S1: Acquire multi-channel dynamic electrical signals. The software program receives a series of multi-channel voltage data points from the microcontroller via the serial port.

[0064] Step S2: Perform signal preprocessing. This step can be performed by the signal preprocessing module in the system. The original signal usually contains noise and varies in length, requiring preprocessing to extract effective features. Specifically, preprocessing may include the following operations: Filtering, i.e., applying a low-pass filter, such as a Butterworth filter, to remove high-frequency noise. Resampling, i.e., because the original signal duration varies due to different writing speeds by the user each time. By using methods such as interpolation or downsampling, the signal sequence of each channel is resampled to a fixed length, such as 200 data points, so that the subsequent machine learning model can process standardized input. Normalization, i.e., using the Z-score normalization method, which subtracts the mean of the signal sequence of each channel and divides it by its standard deviation. This step aims to eliminate the differences in the overall amplitude of the signal between different acquisitions or between different channels, making the model focus more on the morphological characteristics of the signal waveform rather than absolute values. Differentiation, i.e., performing a first-order difference operation on the normalized signal sequence. This operation can highlight the changing trend and dynamic characteristics of the signal, i.e., the slope information of the signal, which is very helpful in distinguishing the details of fast motion. It should be noted that normalization and difference operations are preferred preprocessing steps in the method of this invention.

[0065] Step S3: Construct the feature vector. The preprocessed data sequences from the three channels are concatenated to form a single, long-dimensional feature vector, for example, 3 × 200 = 600 dimensions. This vector completely represents the electrical signal characteristics of a single non-contact motion.

[0066] Step S4: Input the trained machine learning model for recognition. This step can be performed by the recognition module. In this embodiment, a large amount of sample data of users writing the numbers "0" to "9" above the device array is collected in advance. After each sample undergoes the above preprocessing process, it is used together with its corresponding real digit label (0~9) to train a one-dimensional convolutional neural network model. One-dimensional convolutional neural networks are particularly good at automatically extracting local and global pattern features from time series data.

[0067] As an alternative, a support vector machine (SVM) model can also be used. In real-time operation, the feature vectors obtained after real-time acquisition and processing are input into this pre-trained one-dimensional convolutional neural network model.

[0068] Step S5: Output the recognition result. The model analyzes the input feature vector and outputs a probability distribution indicating the likelihood of the input signal corresponding to each category from 0 to 9. The category with the highest probability is selected as the final recognition result. For example, if the model outputs that "3" has the highest probability, the recognition result is the digit "3".

[0069] Step S6: Generate and output control commands. This step can be performed by the control output module. The system generates corresponding control commands based on the recognition result obtained in the previous step. For example, if the recognition result is the number "3", the system can use it as a password input or display the character "3" on the screen.

[0070] To verify the system's performance, an aerial handwritten digit recognition experiment was conducted. The user, positioned approximately 5 cm above the array, used their finger to sequentially "write" the numbers "0" to "9". The raw voltage waveforms acquired by the multi-channel array when the user wrote "3", "4", and "5" clearly show that different digits produced different signal pattern combinations that remained consistent across multiple repetitions. After classification using a machine learning model, a confusion matrix was obtained, which quantitatively evaluated the recognition accuracy. Whether using a support vector machine or a one-dimensional convolutional neural network model, the category recognition accuracy reached 100%. Overall statistics indicate that for the classification task of the 10 digits 0-9, the system's average recognition accuracy can consistently reach over 98%.

[0071] This result strongly demonstrates the high efficiency of the system and method provided by this invention. The high recognition accuracy directly verifies the core advantage of the technical solution of this invention: the dynamic electrical signal generated by dual-factor modulation of air turbulence has a sufficiently high signal-to-noise ratio and discrimination, which can support complex intelligent human-computer interaction applications.

[0072] Finally, some application scenarios of the interactive system in the embodiments are illustrated. For example, the recognized handwritten numbers or characters can be used as passwords or keys for highly secure contactless login or information encryption, as shown in the RSA encryption application scenario. Different gestures, such as swiping left, right, up, and down, can be recognized as different command inputs for controlling character movement or object interaction in virtual reality or augmented reality games.

[0073] In addition, this system can also be used in the fields of Internet of Things and robotics, remotely controlling a physical device, such as a smart car, through gesture control. By recognizing left or right swipe gestures, the system outputs corresponding control signals to the smart car's motors, such as motor1 / motor2, thereby precisely controlling it to complete actions such as turning and going straight.

[0074] Example 5: Fabrication of the interactive device.

[0075] The wet-electric functional layer formulation uses a wettable material that can form a porous polymer network as the main polymer, such as polyglycolic acid, i.e., PGA or an equivalent polymer; introduces organic acids containing dissociable groups, such as tannic acid, citric acid, phytic acid or combinations thereof; adds migratable salt ions, such as NaCl or an equivalent electrolyte; and optionally adds cellulose nanofibers, i.e., CNF, to form ion migration channels.

[0076] Film formation involves coating or drop-coating the above solution / dispersion system onto the surface of the first electrode, and then drying and curing it to form a wet electrical functional layer. The thickness can be selected in the range of 10 μm to 500 μm, preferably about 50 μm to 150 μm.

[0077] Device assembly involves covering the wetted electrical functional layer with a second electrode to form a sandwich structure. The second electrode can be designed to include an array of through-holes or a mesh of holes, such as a 3×3 hole array with a hole diameter of 0.5 mm to 5 mm and a hole spacing of 1 mm to 20 mm, to enhance the repeatability of turbulence modulation. The electrode material can be a carbon electrode, conductive fabric, metal thin film, or conductive composite material, etc.

[0078] The output terminal is connected by connecting leads through conductive adhesive or soldering to form a voltage / current output terminal.

[0079] The working mechanism of air turbulence-modulated wet electrostatic output includes: the device is placed in ambient air, the wet electrostatic functional layer adsorbs water molecules from the air and undergoes group dissociation, generating migratable ions. Due to the ion concentration gradient, the confinement of fixed negative ions, and the migration of mobile cations, a potential difference is formed between the electrodes. When an external object, such as a finger, stylus, or rod-shaped object, moves within a range of 1–80 mm from the device surface, the object's movement induces local airflow and turbulence, resulting in: a decrease in local humidity on the device surface, i.e., a dehydration effect, which alters the ion dissociation and migration capabilities; and local air pressure disturbance, i.e., a pressure effect, which changes the interfacial charge migration and internal electric field distribution. The superposition of these two factors forms a dynamic voltage waveform related to the motion trajectory, thereby achieving non-contact recognition input.

[0080] For multi-channel signal acquisition and handwritten digit recognition, a single device is expanded into a multi-channel array, such as 2-channel, 3-channel, or 5-channel arrays, to acquire output signals from different regions, thereby improving distinguishability. The acquisition circuit can employ a high-input-impedance ADC and be configured with a low-pass filter to suppress noise. The acquired waveforms are resampled, normalized, and subjected to differential / FFT to construct features, which are then fed into an SVM or one-dimensional convolutional neural network for classification and recognition. This enables the recognition of digits 0-9 or letters, words, and other handwritten characters.

[0081] For encrypted input and remote control applications, the recognition result is mapped to a password key, encryption key, or control command to achieve contactless input. The control command is output via a microcontroller to further drive game control, smart car movement, or robot execution, etc.

[0082] Specifically, according to the present invention, a self-powered non-contact interactive device based on the air turbulence-modulated wet electrostatic effect includes: a wet electrostatic functional layer, a first electrode, and a second electrode, wherein the wet electrostatic functional layer is disposed between the first electrode and the second electrode to form a sandwich structure; the wet electrostatic functional layer is a porous polymer network material that can absorb moisture and generate ion dissociation and ion migration; the first electrode or the second electrode includes a through-hole structure or a mesh structure for guiding or enhancing local air disturbance; when an external object moves within a predetermined range from the surface of the device, the local air turbulence induced by the external object causes local humidity changes and air pressure disturbances on the surface of the device, thereby modulating the ion migration process of the wet electrostatic functional layer and outputting a dynamic electrical signal related to the movement.

[0083] Specifically, in the device, the porous polymer network material comprises a main-chain polymer, an additive component containing dissociable groups, and a migratable ionic salt. The main-chain polymer is selected from polyglycolic acid, polyvinyl alcohol, polyacrylic acid, cellulose-based polymers, alginate, gelatin, or any combination thereof. The dissociable groups include one or more of carboxyl groups, phosphate groups, phenolic hydroxyl groups, and sulfonic acid groups. The additive component containing dissociable groups is an organic acid or polyphenol compound, selected from tannic acid, citric acid, phytic acid, or combinations thereof. The migratable ionic salt comprises a metal cation and a corresponding anion, wherein the metal cation is selected from Na+. + K + Li + Ca² + Mg² + One or more of them.

[0084] The wet electro-functional layer contains cellulose nanofibers, nanocellulose, or nanoporous fillers to provide ion migration channels or enhance the moisture absorption area.

[0085] The thickness of the wet electrostatic functional layer is 10 μm to 500 μm, preferably 30 μm to 150 μm.

[0086] Specifically, the through-hole structure is an array of through holes with a diameter of 0.2 mm to 10 mm; preferably 0.5 mm to 5 mm.

[0087] The through-hole structure is a two-dimensional array, and the array form is 2×2, 3×3, 4×4 or higher order array.

[0088] Specifically, the first or second electrode is a carbon electrode, a conductive fabric electrode, a metal thin film electrode, or a conductive polymer composite electrode. The dynamic electrical signal includes one or more of the following: open-circuit voltage signal, short-circuit current signal, or charge signal.

[0089] According to the present invention, a self-powered non-contact human-computer interaction system, based on the self-powered non-contact interaction device, includes: a signal acquisition module for acquiring dynamic electrical signals output by the interaction device; a signal preprocessing module for filtering, resampling, normalizing, differentially operating, and / or frequency domain transforming the dynamic electrical signals; a recognition module for inputting the preprocessed signals into a machine learning model and outputting recognition results; and a control output module for converting the recognition results into control commands to drive external devices or virtual systems.

[0090] The machine learning model is a support vector machine, convolutional neural network, one-dimensional convolutional neural network, random forest, K-nearest neighbors, or a combination thereof. The recognition result includes one or more of the following: numbers, characters, words, gesture categories, or control commands. The signal acquisition module includes an analog-to-digital converter (ADC) with an input impedance greater than or equal to 100 kΩ and a resolution greater than or equal to 12 bits, preferably greater than or equal to 16 bits.

[0091] The system includes a multi-channel interactive device array with 2 to 16 channels to obtain modulation signals from different spatial regions to improve recognition accuracy.

[0092] The control output module is used to control virtual reality / augmented reality interaction, encrypted information input, intelligent vehicle / robot motion control, or remote operating system.

[0093] According to the present invention, a self-powered non-contact interactive identification method based on air turbulence modulated wet electrostatic effect includes: S1: Place the wet electrical interaction device in the air environment to allow it to absorb moisture and form a basic potential output; S2: Collect the dynamic electrical signal caused by the movement of an external object in a non-contact state; S3: Preprocess the dynamic electrical signal to obtain standardized signal features; S4: Input the standardized signal features into a trained machine learning model and output the category result corresponding to the movement of the external object; S5: Output control commands or information input results based on the category result.

[0094] The preprocessing includes resampling the original signal into a fixed-length vector, performing z-score normalization, and first-order difference operations. The preprocessing further includes Fast Fourier Transform (FFT) to generate frequency domain features, which are then concatenated with time domain features to form the model input.

[0095] The machine learning model is trained under supervised supervision using a training sample library containing dynamic electrical signals collected under different distances, relative humidity levels, and writing speeds. The classification results are either classifications of Arabic numerals 0-9 or classifications of characters / words composed of multiple strokes. The control commands are used to trigger actions on the remote device, including one or more of the following: moving straight, turning, stopping, accelerating, or decelerating.

[0096] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0097] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0098] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features of the present invention can be arbitrarily combined with each other.

Claims

1. A self-powered non-contact interactive system based on turbulence-modulated wet electrostatic effect, characterized in that, include: Signal acquisition module: Acquires dynamic electrical signals from self-powered contactless interactive devices; Signal preprocessing module: preprocesses the dynamic electrical signal to obtain the preprocessing result; Recognition module: Recognizes the signal of the preprocessed result, and then obtains the recognition result corresponding to the motion that caused the dynamic electrical signal; Control output module: Based on the recognition results, output corresponding control commands.

2. The self-powered non-contact interactive system based on turbulence-modulated wet electrostatic effect according to claim 1, characterized in that, The self-powered non-contact interactive device includes: a first electrode, a second electrode, a wet electrostatic functional layer disposed between the first electrode and the second electrode, and an airflow disturbance structure; the airflow disturbance structure is disposed on the surfaces of the first electrode and the second electrode; when an external object moves non-contactly within a predetermined distance from the surface of the device, local air turbulence is formed on the surface of the wet electrostatic functional layer, thereby outputting a dynamic electrical signal related to the object's movement.

3. The self-powered non-contact interactive system based on turbulence-modulated wet electrostatic effect according to claim 2, characterized in that, The airflow disturbance structure is a ventilation structure disposed on the first electrode and the second electrode; the ventilation structure includes a through-hole structure or a mesh structure.

4. The self-powered non-contact interactive system based on turbulence-modulated wet electrostatic effect according to claim 2, characterized in that, The wet electrostatic functional layer is a porous polymer network material; the porous polymer network material includes a main chain polymer, additive components with dissociable groups, and a migratable ionic salt; The main chain polymer includes polyglycolic acid, polyvinyl alcohol, polyacrylic acid, cellulose-based polymers, alginate, gelatin, or any combination thereof; The dissociable groups include one or more of carboxyl, phosphate, phenolic hydroxyl, and sulfonic acid groups; The components with the dissociable groups include: tannic acid, citric acid, phytic acid, or any combination thereof; The migratable ionic salt comprises a metal cation and a corresponding anion; the metal cation includes Na. + K + Li + Ca² + Mg² + One or more of them.

5. The self-powered non-contact interactive system based on turbulence-modulated wet electrostatic effect according to claim 2, characterized in that, The predetermined range is 1 mm to 120 mm.

6. The self-powered non-contact interactive system based on turbulent modulation wet electrostatic effect according to claim 1, characterized in that, In the signal preprocessing module, the preprocessing includes: filtering, resampling, normalization, differential operation, or frequency domain transformation; In the recognition module, the signal of the preprocessed result is input into a preset machine learning model, and then the corresponding recognition result is output; the machine learning model includes a support vector machine model or a convolutional neural network model; the recognition result includes one or more of the following: numbers, characters, words, gesture categories or control commands.

7. The self-powered non-contact interactive system based on turbulence-modulated wet electrostatic effect according to claim 1, characterized in that, The self-powered non-contact interactive device is arranged in a multi-channel array with 2 to 16 channels.

8. A self-powered non-contact interaction method based on turbulence-modulated wet electrostatic effect, implemented based on the self-powered non-contact interaction system based on turbulence-modulated wet electrostatic effect as described in any one of claims 1 to 7, characterized in that, include: The signal acquisition module acquires dynamic electrical signals from the self-powered contactless interactive device. The signal preprocessing module preprocesses the dynamic electrical signal to obtain the preprocessing result. The recognition module identifies the signal of the preprocessed result, thereby obtaining the recognition result corresponding to the motion that caused the dynamic electrical signal; The control output module outputs corresponding control commands based on the recognition results.

9. The self-powered non-contact interaction method based on turbulence-modulated wet electrostatic effect according to claim 8, characterized in that, The preprocessing includes filtering, resampling, normalization, differential operations, or frequency domain transformation.

10. The self-powered non-contact interaction method based on turbulence-modulated wet electrostatic effect according to claim 9, characterized in that, The identification result is a classification of the motion.

Images